diff --git "a/checkpoint-6000/trainer_state.json" "b/checkpoint-6000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6000/trainer_state.json"
@@ -0,0 +1,42033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8188890405350076,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001364815067558346,
+      "grad_norm": 5.499263763427734,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 2.2291,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002729630135116692,
+      "grad_norm": 4.660139560699463,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 2.16,
+      "step": 2
+    },
+    {
+      "epoch": 0.00040944452026750374,
+      "grad_norm": 4.5329132080078125,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.2355,
+      "step": 3
+    },
+    {
+      "epoch": 0.0005459260270233384,
+      "grad_norm": 4.985339641571045,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 2.1168,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006824075337791729,
+      "grad_norm": 4.656001567840576,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 2.1413,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008188890405350075,
+      "grad_norm": 3.8532602787017822,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.0836,
+      "step": 6
+    },
+    {
+      "epoch": 0.0009553705472908421,
+      "grad_norm": 5.339655876159668,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 2.1081,
+      "step": 7
+    },
+    {
+      "epoch": 0.0010918520540466768,
+      "grad_norm": 5.364240646362305,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 2.1882,
+      "step": 8
+    },
+    {
+      "epoch": 0.0012283335608025113,
+      "grad_norm": 4.23295783996582,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.1973,
+      "step": 9
+    },
+    {
+      "epoch": 0.0013648150675583458,
+      "grad_norm": 3.9411795139312744,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 2.0814,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015012965743141805,
+      "grad_norm": 3.924015760421753,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 2.046,
+      "step": 11
+    },
+    {
+      "epoch": 0.001637778081070015,
+      "grad_norm": 3.311981201171875,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.1344,
+      "step": 12
+    },
+    {
+      "epoch": 0.0017742595878258497,
+      "grad_norm": 3.360774040222168,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 2.0675,
+      "step": 13
+    },
+    {
+      "epoch": 0.0019107410945816842,
+      "grad_norm": 2.448817253112793,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 2.1028,
+      "step": 14
+    },
+    {
+      "epoch": 0.002047222601337519,
+      "grad_norm": 2.544743299484253,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.1193,
+      "step": 15
+    },
+    {
+      "epoch": 0.0021837041080933536,
+      "grad_norm": 2.36195969581604,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 2.0871,
+      "step": 16
+    },
+    {
+      "epoch": 0.002320185614849188,
+      "grad_norm": 1.793238878250122,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.8511,
+      "step": 17
+    },
+    {
+      "epoch": 0.0024566671216050225,
+      "grad_norm": 2.0045647621154785,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.1472,
+      "step": 18
+    },
+    {
+      "epoch": 0.0025931486283608573,
+      "grad_norm": 1.3631515502929688,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.9385,
+      "step": 19
+    },
+    {
+      "epoch": 0.0027296301351166915,
+      "grad_norm": 1.7018771171569824,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.9948,
+      "step": 20
+    },
+    {
+      "epoch": 0.0028661116418725262,
+      "grad_norm": 1.1106555461883545,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.9604,
+      "step": 21
+    },
+    {
+      "epoch": 0.003002593148628361,
+      "grad_norm": 1.2149659395217896,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 2.0106,
+      "step": 22
+    },
+    {
+      "epoch": 0.0031390746553841956,
+      "grad_norm": 0.8920502066612244,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.7102,
+      "step": 23
+    },
+    {
+      "epoch": 0.00327555616214003,
+      "grad_norm": 0.8421320915222168,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.9706,
+      "step": 24
+    },
+    {
+      "epoch": 0.0034120376688958646,
+      "grad_norm": 0.8898251056671143,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.9256,
+      "step": 25
+    },
+    {
+      "epoch": 0.0035485191756516993,
+      "grad_norm": 0.8477884531021118,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 2.0185,
+      "step": 26
+    },
+    {
+      "epoch": 0.0036850006824075336,
+      "grad_norm": 0.6731047630310059,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.8752,
+      "step": 27
+    },
+    {
+      "epoch": 0.0038214821891633683,
+      "grad_norm": 0.8444198369979858,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 1.9105,
+      "step": 28
+    },
+    {
+      "epoch": 0.003957963695919203,
+      "grad_norm": 0.6602134704589844,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 1.888,
+      "step": 29
+    },
+    {
+      "epoch": 0.004094445202675038,
+      "grad_norm": 0.5404819846153259,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.9087,
+      "step": 30
+    },
+    {
+      "epoch": 0.004230926709430872,
+      "grad_norm": 0.5722994804382324,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 1.8945,
+      "step": 31
+    },
+    {
+      "epoch": 0.004367408216186707,
+      "grad_norm": 0.5562062859535217,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 1.8278,
+      "step": 32
+    },
+    {
+      "epoch": 0.004503889722942541,
+      "grad_norm": 0.5615333914756775,
+      "learning_rate": 8.8e-06,
+      "loss": 1.956,
+      "step": 33
+    },
+    {
+      "epoch": 0.004640371229698376,
+      "grad_norm": 0.5378018021583557,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 1.8295,
+      "step": 34
+    },
+    {
+      "epoch": 0.004776852736454211,
+      "grad_norm": 0.47099870443344116,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.8913,
+      "step": 35
+    },
+    {
+      "epoch": 0.004913334243210045,
+      "grad_norm": 0.4328675866127014,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.8578,
+      "step": 36
+    },
+    {
+      "epoch": 0.005049815749965879,
+      "grad_norm": 0.5030441284179688,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 1.8735,
+      "step": 37
+    },
+    {
+      "epoch": 0.0051862972567217145,
+      "grad_norm": 0.43279239535331726,
+      "learning_rate": 1.0133333333333335e-05,
+      "loss": 1.8315,
+      "step": 38
+    },
+    {
+      "epoch": 0.005322778763477549,
+      "grad_norm": 0.4185083210468292,
+      "learning_rate": 1.04e-05,
+      "loss": 1.8237,
+      "step": 39
+    },
+    {
+      "epoch": 0.005459260270233383,
+      "grad_norm": 0.4204055368900299,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.8742,
+      "step": 40
+    },
+    {
+      "epoch": 0.005595741776989218,
+      "grad_norm": 0.32618430256843567,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 1.809,
+      "step": 41
+    },
+    {
+      "epoch": 0.0057322232837450525,
+      "grad_norm": 0.35566413402557373,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.7725,
+      "step": 42
+    },
+    {
+      "epoch": 0.005868704790500887,
+      "grad_norm": 0.29983246326446533,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 1.9016,
+      "step": 43
+    },
+    {
+      "epoch": 0.006005186297256722,
+      "grad_norm": 0.2848670482635498,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 1.8058,
+      "step": 44
+    },
+    {
+      "epoch": 0.006141667804012556,
+      "grad_norm": 0.2734982967376709,
+      "learning_rate": 1.2e-05,
+      "loss": 1.7248,
+      "step": 45
+    },
+    {
+      "epoch": 0.006278149310768391,
+      "grad_norm": 0.27347004413604736,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 1.8156,
+      "step": 46
+    },
+    {
+      "epoch": 0.0064146308175242256,
+      "grad_norm": 0.2598482668399811,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 1.7959,
+      "step": 47
+    },
+    {
+      "epoch": 0.00655111232428006,
+      "grad_norm": 0.23635509610176086,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.7623,
+      "step": 48
+    },
+    {
+      "epoch": 0.006687593831035895,
+      "grad_norm": 0.24684293568134308,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 1.845,
+      "step": 49
+    },
+    {
+      "epoch": 0.006824075337791729,
+      "grad_norm": 0.23189635574817657,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.8112,
+      "step": 50
+    },
+    {
+      "epoch": 0.0069605568445475635,
+      "grad_norm": 0.2241699993610382,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.7468,
+      "step": 51
+    },
+    {
+      "epoch": 0.007097038351303399,
+      "grad_norm": 0.24107544124126434,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 1.8191,
+      "step": 52
+    },
+    {
+      "epoch": 0.007233519858059233,
+      "grad_norm": 0.22175419330596924,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 1.7544,
+      "step": 53
+    },
+    {
+      "epoch": 0.007370001364815067,
+      "grad_norm": 0.2288978397846222,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.7702,
+      "step": 54
+    },
+    {
+      "epoch": 0.007506482871570902,
+      "grad_norm": 0.22638070583343506,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 1.7895,
+      "step": 55
+    },
+    {
+      "epoch": 0.007642964378326737,
+      "grad_norm": 0.248519629240036,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 1.8377,
+      "step": 56
+    },
+    {
+      "epoch": 0.007779445885082571,
+      "grad_norm": 0.2243175208568573,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.7965,
+      "step": 57
+    },
+    {
+      "epoch": 0.007915927391838405,
+      "grad_norm": 0.2314777821302414,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 1.7646,
+      "step": 58
+    },
+    {
+      "epoch": 0.008052408898594241,
+      "grad_norm": 0.22540760040283203,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 1.791,
+      "step": 59
+    },
+    {
+      "epoch": 0.008188890405350075,
+      "grad_norm": 0.24623164534568787,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7741,
+      "step": 60
+    },
+    {
+      "epoch": 0.00832537191210591,
+      "grad_norm": 0.24007107317447662,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 1.7735,
+      "step": 61
+    },
+    {
+      "epoch": 0.008461853418861744,
+      "grad_norm": 0.2574295699596405,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 1.7881,
+      "step": 62
+    },
+    {
+      "epoch": 0.008598334925617578,
+      "grad_norm": 0.2421531081199646,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.7364,
+      "step": 63
+    },
+    {
+      "epoch": 0.008734816432373414,
+      "grad_norm": 0.2307588905096054,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 1.727,
+      "step": 64
+    },
+    {
+      "epoch": 0.008871297939129249,
+      "grad_norm": 0.238815039396286,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 1.7139,
+      "step": 65
+    },
+    {
+      "epoch": 0.009007779445885083,
+      "grad_norm": 0.22701887786388397,
+      "learning_rate": 1.76e-05,
+      "loss": 1.773,
+      "step": 66
+    },
+    {
+      "epoch": 0.009144260952640917,
+      "grad_norm": 0.22581444680690765,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 1.7459,
+      "step": 67
+    },
+    {
+      "epoch": 0.009280742459396751,
+      "grad_norm": 0.21733753383159637,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 1.7576,
+      "step": 68
+    },
+    {
+      "epoch": 0.009417223966152586,
+      "grad_norm": 0.24274277687072754,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.7336,
+      "step": 69
+    },
+    {
+      "epoch": 0.009553705472908422,
+      "grad_norm": 0.20616181194782257,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.7586,
+      "step": 70
+    },
+    {
+      "epoch": 0.009690186979664256,
+      "grad_norm": 0.205729678273201,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 1.7318,
+      "step": 71
+    },
+    {
+      "epoch": 0.00982666848642009,
+      "grad_norm": 0.20966161787509918,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.7637,
+      "step": 72
+    },
+    {
+      "epoch": 0.009963149993175924,
+      "grad_norm": 0.20736359059810638,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 1.7353,
+      "step": 73
+    },
+    {
+      "epoch": 0.010099631499931759,
+      "grad_norm": 0.20281726121902466,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 1.7322,
+      "step": 74
+    },
+    {
+      "epoch": 0.010236113006687593,
+      "grad_norm": 0.20042139291763306,
+      "learning_rate": 2e-05,
+      "loss": 1.7195,
+      "step": 75
+    },
+    {
+      "epoch": 0.010372594513443429,
+      "grad_norm": 0.21634063124656677,
+      "learning_rate": 2.026666666666667e-05,
+      "loss": 1.7943,
+      "step": 76
+    },
+    {
+      "epoch": 0.010509076020199263,
+      "grad_norm": 0.2153872698545456,
+      "learning_rate": 2.0533333333333336e-05,
+      "loss": 1.763,
+      "step": 77
+    },
+    {
+      "epoch": 0.010645557526955098,
+      "grad_norm": 0.2081228494644165,
+      "learning_rate": 2.08e-05,
+      "loss": 1.7792,
+      "step": 78
+    },
+    {
+      "epoch": 0.010782039033710932,
+      "grad_norm": 0.2098853439092636,
+      "learning_rate": 2.1066666666666666e-05,
+      "loss": 1.7918,
+      "step": 79
+    },
+    {
+      "epoch": 0.010918520540466766,
+      "grad_norm": 0.20721176266670227,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 1.7064,
+      "step": 80
+    },
+    {
+      "epoch": 0.011055002047222602,
+      "grad_norm": 0.2197636067867279,
+      "learning_rate": 2.1600000000000003e-05,
+      "loss": 1.7002,
+      "step": 81
+    },
+    {
+      "epoch": 0.011191483553978436,
+      "grad_norm": 0.19892142713069916,
+      "learning_rate": 2.186666666666667e-05,
+      "loss": 1.6049,
+      "step": 82
+    },
+    {
+      "epoch": 0.01132796506073427,
+      "grad_norm": 0.20634345710277557,
+      "learning_rate": 2.2133333333333337e-05,
+      "loss": 1.7028,
+      "step": 83
+    },
+    {
+      "epoch": 0.011464446567490105,
+      "grad_norm": 0.213628351688385,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.711,
+      "step": 84
+    },
+    {
+      "epoch": 0.01160092807424594,
+      "grad_norm": 0.20134010910987854,
+      "learning_rate": 2.2666666666666668e-05,
+      "loss": 1.672,
+      "step": 85
+    },
+    {
+      "epoch": 0.011737409581001773,
+      "grad_norm": 0.21955366432666779,
+      "learning_rate": 2.2933333333333336e-05,
+      "loss": 1.8046,
+      "step": 86
+    },
+    {
+      "epoch": 0.01187389108775761,
+      "grad_norm": 0.22468940913677216,
+      "learning_rate": 2.32e-05,
+      "loss": 1.763,
+      "step": 87
+    },
+    {
+      "epoch": 0.012010372594513444,
+      "grad_norm": 0.20560674369335175,
+      "learning_rate": 2.346666666666667e-05,
+      "loss": 1.7117,
+      "step": 88
+    },
+    {
+      "epoch": 0.012146854101269278,
+      "grad_norm": 0.21587899327278137,
+      "learning_rate": 2.373333333333334e-05,
+      "loss": 1.8043,
+      "step": 89
+    },
+    {
+      "epoch": 0.012283335608025112,
+      "grad_norm": 0.210346519947052,
+      "learning_rate": 2.4e-05,
+      "loss": 1.6873,
+      "step": 90
+    },
+    {
+      "epoch": 0.012419817114780947,
+      "grad_norm": 0.22895421087741852,
+      "learning_rate": 2.426666666666667e-05,
+      "loss": 1.7526,
+      "step": 91
+    },
+    {
+      "epoch": 0.012556298621536783,
+      "grad_norm": 0.21360772848129272,
+      "learning_rate": 2.4533333333333334e-05,
+      "loss": 1.7118,
+      "step": 92
+    },
+    {
+      "epoch": 0.012692780128292617,
+      "grad_norm": 0.23913121223449707,
+      "learning_rate": 2.4800000000000003e-05,
+      "loss": 1.7234,
+      "step": 93
+    },
+    {
+      "epoch": 0.012829261635048451,
+      "grad_norm": 0.2257775068283081,
+      "learning_rate": 2.5066666666666672e-05,
+      "loss": 1.7052,
+      "step": 94
+    },
+    {
+      "epoch": 0.012965743141804285,
+      "grad_norm": 0.22123000025749207,
+      "learning_rate": 2.5333333333333334e-05,
+      "loss": 1.7659,
+      "step": 95
+    },
+    {
+      "epoch": 0.01310222464856012,
+      "grad_norm": 0.25808942317962646,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.7974,
+      "step": 96
+    },
+    {
+      "epoch": 0.013238706155315954,
+      "grad_norm": 0.22531723976135254,
+      "learning_rate": 2.5866666666666667e-05,
+      "loss": 1.7703,
+      "step": 97
+    },
+    {
+      "epoch": 0.01337518766207179,
+      "grad_norm": 0.22541013360023499,
+      "learning_rate": 2.6133333333333336e-05,
+      "loss": 1.6882,
+      "step": 98
+    },
+    {
+      "epoch": 0.013511669168827624,
+      "grad_norm": 0.24326053261756897,
+      "learning_rate": 2.6400000000000005e-05,
+      "loss": 1.7636,
+      "step": 99
+    },
+    {
+      "epoch": 0.013648150675583458,
+      "grad_norm": 0.22722408175468445,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.7421,
+      "step": 100
+    },
+    {
+      "epoch": 0.013784632182339293,
+      "grad_norm": 0.23277103900909424,
+      "learning_rate": 2.6933333333333335e-05,
+      "loss": 1.7752,
+      "step": 101
+    },
+    {
+      "epoch": 0.013921113689095127,
+      "grad_norm": 0.2395719438791275,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.762,
+      "step": 102
+    },
+    {
+      "epoch": 0.014057595195850963,
+      "grad_norm": 0.21451711654663086,
+      "learning_rate": 2.746666666666667e-05,
+      "loss": 1.7415,
+      "step": 103
+    },
+    {
+      "epoch": 0.014194076702606797,
+      "grad_norm": 0.24826586246490479,
+      "learning_rate": 2.7733333333333338e-05,
+      "loss": 1.8362,
+      "step": 104
+    },
+    {
+      "epoch": 0.014330558209362632,
+      "grad_norm": 0.24056163430213928,
+      "learning_rate": 2.8e-05,
+      "loss": 1.801,
+      "step": 105
+    },
+    {
+      "epoch": 0.014467039716118466,
+      "grad_norm": 0.24722066521644592,
+      "learning_rate": 2.8266666666666668e-05,
+      "loss": 1.727,
+      "step": 106
+    },
+    {
+      "epoch": 0.0146035212228743,
+      "grad_norm": 0.23071065545082092,
+      "learning_rate": 2.8533333333333337e-05,
+      "loss": 1.7135,
+      "step": 107
+    },
+    {
+      "epoch": 0.014740002729630134,
+      "grad_norm": 0.234844371676445,
+      "learning_rate": 2.8800000000000002e-05,
+      "loss": 1.8161,
+      "step": 108
+    },
+    {
+      "epoch": 0.01487648423638597,
+      "grad_norm": 0.2340114414691925,
+      "learning_rate": 2.906666666666667e-05,
+      "loss": 1.7293,
+      "step": 109
+    },
+    {
+      "epoch": 0.015012965743141805,
+      "grad_norm": 0.22716818749904633,
+      "learning_rate": 2.9333333333333333e-05,
+      "loss": 1.7176,
+      "step": 110
+    },
+    {
+      "epoch": 0.015149447249897639,
+      "grad_norm": 0.2477683275938034,
+      "learning_rate": 2.96e-05,
+      "loss": 1.6629,
+      "step": 111
+    },
+    {
+      "epoch": 0.015285928756653473,
+      "grad_norm": 0.2254948765039444,
+      "learning_rate": 2.986666666666667e-05,
+      "loss": 1.6964,
+      "step": 112
+    },
+    {
+      "epoch": 0.015422410263409307,
+      "grad_norm": 0.23815442621707916,
+      "learning_rate": 3.0133333333333335e-05,
+      "loss": 1.7543,
+      "step": 113
+    },
+    {
+      "epoch": 0.015558891770165142,
+      "grad_norm": 0.22915583848953247,
+      "learning_rate": 3.0400000000000004e-05,
+      "loss": 1.7556,
+      "step": 114
+    },
+    {
+      "epoch": 0.015695373276920978,
+      "grad_norm": 0.24844741821289062,
+      "learning_rate": 3.066666666666667e-05,
+      "loss": 1.7621,
+      "step": 115
+    },
+    {
+      "epoch": 0.01583185478367681,
+      "grad_norm": 0.22810101509094238,
+      "learning_rate": 3.093333333333334e-05,
+      "loss": 1.7511,
+      "step": 116
+    },
+    {
+      "epoch": 0.015968336290432646,
+      "grad_norm": 0.2491244375705719,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 1.7414,
+      "step": 117
+    },
+    {
+      "epoch": 0.016104817797188482,
+      "grad_norm": 0.22312700748443604,
+      "learning_rate": 3.146666666666667e-05,
+      "loss": 1.7253,
+      "step": 118
+    },
+    {
+      "epoch": 0.016241299303944315,
+      "grad_norm": 0.22729012370109558,
+      "learning_rate": 3.173333333333334e-05,
+      "loss": 1.6379,
+      "step": 119
+    },
+    {
+      "epoch": 0.01637778081070015,
+      "grad_norm": 0.2469879686832428,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.7275,
+      "step": 120
+    },
+    {
+      "epoch": 0.016514262317455983,
+      "grad_norm": 0.23909714818000793,
+      "learning_rate": 3.226666666666667e-05,
+      "loss": 1.692,
+      "step": 121
+    },
+    {
+      "epoch": 0.01665074382421182,
+      "grad_norm": 0.25035664439201355,
+      "learning_rate": 3.2533333333333336e-05,
+      "loss": 1.7314,
+      "step": 122
+    },
+    {
+      "epoch": 0.016787225330967655,
+      "grad_norm": 0.23064374923706055,
+      "learning_rate": 3.28e-05,
+      "loss": 1.8229,
+      "step": 123
+    },
+    {
+      "epoch": 0.016923706837723488,
+      "grad_norm": 0.22864677011966705,
+      "learning_rate": 3.3066666666666666e-05,
+      "loss": 1.6719,
+      "step": 124
+    },
+    {
+      "epoch": 0.017060188344479324,
+      "grad_norm": 0.23785266280174255,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.7707,
+      "step": 125
+    },
+    {
+      "epoch": 0.017196669851235157,
+      "grad_norm": 0.2404264658689499,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.7706,
+      "step": 126
+    },
+    {
+      "epoch": 0.017333151357990993,
+      "grad_norm": 0.24038049578666687,
+      "learning_rate": 3.386666666666667e-05,
+      "loss": 1.8063,
+      "step": 127
+    },
+    {
+      "epoch": 0.01746963286474683,
+      "grad_norm": 0.26162853837013245,
+      "learning_rate": 3.413333333333334e-05,
+      "loss": 1.7752,
+      "step": 128
+    },
+    {
+      "epoch": 0.01760611437150266,
+      "grad_norm": 0.22979912161827087,
+      "learning_rate": 3.44e-05,
+      "loss": 1.7798,
+      "step": 129
+    },
+    {
+      "epoch": 0.017742595878258497,
+      "grad_norm": 0.23631471395492554,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 1.8091,
+      "step": 130
+    },
+    {
+      "epoch": 0.01787907738501433,
+      "grad_norm": 0.230385422706604,
+      "learning_rate": 3.493333333333333e-05,
+      "loss": 1.6328,
+      "step": 131
+    },
+    {
+      "epoch": 0.018015558891770166,
+      "grad_norm": 0.24715574085712433,
+      "learning_rate": 3.52e-05,
+      "loss": 1.6432,
+      "step": 132
+    },
+    {
+      "epoch": 0.018152040398525998,
+      "grad_norm": 0.25098592042922974,
+      "learning_rate": 3.546666666666667e-05,
+      "loss": 1.6562,
+      "step": 133
+    },
+    {
+      "epoch": 0.018288521905281834,
+      "grad_norm": 0.2581266760826111,
+      "learning_rate": 3.573333333333333e-05,
+      "loss": 1.7382,
+      "step": 134
+    },
+    {
+      "epoch": 0.01842500341203767,
+      "grad_norm": 0.24009093642234802,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7733,
+      "step": 135
+    },
+    {
+      "epoch": 0.018561484918793503,
+      "grad_norm": 0.23203186690807343,
+      "learning_rate": 3.626666666666667e-05,
+      "loss": 1.672,
+      "step": 136
+    },
+    {
+      "epoch": 0.01869796642554934,
+      "grad_norm": 0.24924978613853455,
+      "learning_rate": 3.653333333333334e-05,
+      "loss": 1.7233,
+      "step": 137
+    },
+    {
+      "epoch": 0.01883444793230517,
+      "grad_norm": 0.2568417489528656,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.6306,
+      "step": 138
+    },
+    {
+      "epoch": 0.018970929439061007,
+      "grad_norm": 0.25758302211761475,
+      "learning_rate": 3.706666666666667e-05,
+      "loss": 1.7771,
+      "step": 139
+    },
+    {
+      "epoch": 0.019107410945816843,
+      "grad_norm": 0.23922881484031677,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 1.768,
+      "step": 140
+    },
+    {
+      "epoch": 0.019243892452572676,
+      "grad_norm": 0.22760292887687683,
+      "learning_rate": 3.76e-05,
+      "loss": 1.6672,
+      "step": 141
+    },
+    {
+      "epoch": 0.019380373959328512,
+      "grad_norm": 0.25242963433265686,
+      "learning_rate": 3.786666666666667e-05,
+      "loss": 1.7868,
+      "step": 142
+    },
+    {
+      "epoch": 0.019516855466084344,
+      "grad_norm": 0.2641962766647339,
+      "learning_rate": 3.8133333333333336e-05,
+      "loss": 1.6787,
+      "step": 143
+    },
+    {
+      "epoch": 0.01965333697284018,
+      "grad_norm": 0.2402966320514679,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.6341,
+      "step": 144
+    },
+    {
+      "epoch": 0.019789818479596016,
+      "grad_norm": 0.3131767809391022,
+      "learning_rate": 3.866666666666667e-05,
+      "loss": 1.7194,
+      "step": 145
+    },
+    {
+      "epoch": 0.01992629998635185,
+      "grad_norm": 0.2589942514896393,
+      "learning_rate": 3.8933333333333336e-05,
+      "loss": 1.7461,
+      "step": 146
+    },
+    {
+      "epoch": 0.020062781493107685,
+      "grad_norm": 0.30445966124534607,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.7237,
+      "step": 147
+    },
+    {
+      "epoch": 0.020199262999863517,
+      "grad_norm": 0.2719636559486389,
+      "learning_rate": 3.946666666666667e-05,
+      "loss": 1.71,
+      "step": 148
+    },
+    {
+      "epoch": 0.020335744506619353,
+      "grad_norm": 0.23952992260456085,
+      "learning_rate": 3.9733333333333335e-05,
+      "loss": 1.8032,
+      "step": 149
+    },
+    {
+      "epoch": 0.020472226013375186,
+      "grad_norm": 0.2376592457294464,
+      "learning_rate": 4e-05,
+      "loss": 1.7425,
+      "step": 150
+    },
+    {
+      "epoch": 0.020608707520131022,
+      "grad_norm": 0.24733857810497284,
+      "learning_rate": 4.0266666666666665e-05,
+      "loss": 1.7476,
+      "step": 151
+    },
+    {
+      "epoch": 0.020745189026886858,
+      "grad_norm": 0.23061153292655945,
+      "learning_rate": 4.053333333333334e-05,
+      "loss": 1.6733,
+      "step": 152
+    },
+    {
+      "epoch": 0.02088167053364269,
+      "grad_norm": 0.2736116051673889,
+      "learning_rate": 4.08e-05,
+      "loss": 1.6989,
+      "step": 153
+    },
+    {
+      "epoch": 0.021018152040398527,
+      "grad_norm": 0.2521119713783264,
+      "learning_rate": 4.106666666666667e-05,
+      "loss": 1.6959,
+      "step": 154
+    },
+    {
+      "epoch": 0.02115463354715436,
+      "grad_norm": 0.25929921865463257,
+      "learning_rate": 4.133333333333334e-05,
+      "loss": 1.6956,
+      "step": 155
+    },
+    {
+      "epoch": 0.021291115053910195,
+      "grad_norm": 0.2587325870990753,
+      "learning_rate": 4.16e-05,
+      "loss": 1.7962,
+      "step": 156
+    },
+    {
+      "epoch": 0.02142759656066603,
+      "grad_norm": 0.23272058367729187,
+      "learning_rate": 4.186666666666667e-05,
+      "loss": 1.6473,
+      "step": 157
+    },
+    {
+      "epoch": 0.021564078067421864,
+      "grad_norm": 0.2319646179676056,
+      "learning_rate": 4.213333333333333e-05,
+      "loss": 1.6893,
+      "step": 158
+    },
+    {
+      "epoch": 0.0217005595741777,
+      "grad_norm": 0.23524560034275055,
+      "learning_rate": 4.240000000000001e-05,
+      "loss": 1.718,
+      "step": 159
+    },
+    {
+      "epoch": 0.021837041080933532,
+      "grad_norm": 0.23512865602970123,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 1.7011,
+      "step": 160
+    },
+    {
+      "epoch": 0.021973522587689368,
+      "grad_norm": 0.21862873435020447,
+      "learning_rate": 4.293333333333333e-05,
+      "loss": 1.6088,
+      "step": 161
+    },
+    {
+      "epoch": 0.022110004094445204,
+      "grad_norm": 0.22855187952518463,
+      "learning_rate": 4.3200000000000007e-05,
+      "loss": 1.7111,
+      "step": 162
+    },
+    {
+      "epoch": 0.022246485601201037,
+      "grad_norm": 0.23267142474651337,
+      "learning_rate": 4.346666666666667e-05,
+      "loss": 1.6642,
+      "step": 163
+    },
+    {
+      "epoch": 0.022382967107956873,
+      "grad_norm": 0.244045227766037,
+      "learning_rate": 4.373333333333334e-05,
+      "loss": 1.6515,
+      "step": 164
+    },
+    {
+      "epoch": 0.022519448614712705,
+      "grad_norm": 0.2407236099243164,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.7264,
+      "step": 165
+    },
+    {
+      "epoch": 0.02265593012146854,
+      "grad_norm": 0.26442793011665344,
+      "learning_rate": 4.4266666666666674e-05,
+      "loss": 1.7056,
+      "step": 166
+    },
+    {
+      "epoch": 0.022792411628224377,
+      "grad_norm": 0.21957506239414215,
+      "learning_rate": 4.4533333333333336e-05,
+      "loss": 1.6498,
+      "step": 167
+    },
+    {
+      "epoch": 0.02292889313498021,
+      "grad_norm": 0.2429881989955902,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.7154,
+      "step": 168
+    },
+    {
+      "epoch": 0.023065374641736046,
+      "grad_norm": 0.2509872615337372,
+      "learning_rate": 4.506666666666667e-05,
+      "loss": 1.7395,
+      "step": 169
+    },
+    {
+      "epoch": 0.02320185614849188,
+      "grad_norm": 0.2709554433822632,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 1.7147,
+      "step": 170
+    },
+    {
+      "epoch": 0.023338337655247714,
+      "grad_norm": 0.22771143913269043,
+      "learning_rate": 4.56e-05,
+      "loss": 1.6264,
+      "step": 171
+    },
+    {
+      "epoch": 0.023474819162003547,
+      "grad_norm": 0.2200254499912262,
+      "learning_rate": 4.586666666666667e-05,
+      "loss": 1.6607,
+      "step": 172
+    },
+    {
+      "epoch": 0.023611300668759383,
+      "grad_norm": 0.2651680111885071,
+      "learning_rate": 4.6133333333333334e-05,
+      "loss": 1.6919,
+      "step": 173
+    },
+    {
+      "epoch": 0.02374778217551522,
+      "grad_norm": 0.2326844334602356,
+      "learning_rate": 4.64e-05,
+      "loss": 1.6351,
+      "step": 174
+    },
+    {
+      "epoch": 0.02388426368227105,
+      "grad_norm": 0.26194167137145996,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.7389,
+      "step": 175
+    },
+    {
+      "epoch": 0.024020745189026887,
+      "grad_norm": 0.22865556180477142,
+      "learning_rate": 4.693333333333334e-05,
+      "loss": 1.7029,
+      "step": 176
+    },
+    {
+      "epoch": 0.02415722669578272,
+      "grad_norm": 0.27646172046661377,
+      "learning_rate": 4.72e-05,
+      "loss": 1.6935,
+      "step": 177
+    },
+    {
+      "epoch": 0.024293708202538556,
+      "grad_norm": 0.24295304715633392,
+      "learning_rate": 4.746666666666668e-05,
+      "loss": 1.6951,
+      "step": 178
+    },
+    {
+      "epoch": 0.024430189709294392,
+      "grad_norm": 0.2156314253807068,
+      "learning_rate": 4.773333333333334e-05,
+      "loss": 1.648,
+      "step": 179
+    },
+    {
+      "epoch": 0.024566671216050225,
+      "grad_norm": 0.22737029194831848,
+      "learning_rate": 4.8e-05,
+      "loss": 1.6679,
+      "step": 180
+    },
+    {
+      "epoch": 0.02470315272280606,
+      "grad_norm": 0.248259037733078,
+      "learning_rate": 4.826666666666668e-05,
+      "loss": 1.6591,
+      "step": 181
+    },
+    {
+      "epoch": 0.024839634229561893,
+      "grad_norm": 0.22868888080120087,
+      "learning_rate": 4.853333333333334e-05,
+      "loss": 1.7137,
+      "step": 182
+    },
+    {
+      "epoch": 0.02497611573631773,
+      "grad_norm": 0.25324422121047974,
+      "learning_rate": 4.88e-05,
+      "loss": 1.7322,
+      "step": 183
+    },
+    {
+      "epoch": 0.025112597243073565,
+      "grad_norm": 0.24201399087905884,
+      "learning_rate": 4.906666666666667e-05,
+      "loss": 1.651,
+      "step": 184
+    },
+    {
+      "epoch": 0.025249078749829398,
+      "grad_norm": 0.2308133840560913,
+      "learning_rate": 4.933333333333334e-05,
+      "loss": 1.7445,
+      "step": 185
+    },
+    {
+      "epoch": 0.025385560256585234,
+      "grad_norm": 0.2358778715133667,
+      "learning_rate": 4.9600000000000006e-05,
+      "loss": 1.6765,
+      "step": 186
+    },
+    {
+      "epoch": 0.025522041763341066,
+      "grad_norm": 0.2414613515138626,
+      "learning_rate": 4.986666666666667e-05,
+      "loss": 1.6198,
+      "step": 187
+    },
+    {
+      "epoch": 0.025658523270096902,
+      "grad_norm": 0.26769500970840454,
+      "learning_rate": 5.0133333333333343e-05,
+      "loss": 1.7696,
+      "step": 188
+    },
+    {
+      "epoch": 0.025795004776852735,
+      "grad_norm": 0.2725449502468109,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 1.7405,
+      "step": 189
+    },
+    {
+      "epoch": 0.02593148628360857,
+      "grad_norm": 0.2680228054523468,
+      "learning_rate": 5.066666666666667e-05,
+      "loss": 1.6777,
+      "step": 190
+    },
+    {
+      "epoch": 0.026067967790364407,
+      "grad_norm": 0.2760530412197113,
+      "learning_rate": 5.093333333333334e-05,
+      "loss": 1.651,
+      "step": 191
+    },
+    {
+      "epoch": 0.02620444929712024,
+      "grad_norm": 0.24931032955646515,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 1.7195,
+      "step": 192
+    },
+    {
+      "epoch": 0.026340930803876075,
+      "grad_norm": 0.28645068407058716,
+      "learning_rate": 5.1466666666666666e-05,
+      "loss": 1.6904,
+      "step": 193
+    },
+    {
+      "epoch": 0.026477412310631908,
+      "grad_norm": 0.2453315705060959,
+      "learning_rate": 5.1733333333333335e-05,
+      "loss": 1.6433,
+      "step": 194
+    },
+    {
+      "epoch": 0.026613893817387744,
+      "grad_norm": 0.2835664451122284,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.6562,
+      "step": 195
+    },
+    {
+      "epoch": 0.02675037532414358,
+      "grad_norm": 0.26578426361083984,
+      "learning_rate": 5.226666666666667e-05,
+      "loss": 1.6577,
+      "step": 196
+    },
+    {
+      "epoch": 0.026886856830899412,
+      "grad_norm": 0.32209882140159607,
+      "learning_rate": 5.2533333333333334e-05,
+      "loss": 1.7582,
+      "step": 197
+    },
+    {
+      "epoch": 0.02702333833765525,
+      "grad_norm": 0.2664225399494171,
+      "learning_rate": 5.280000000000001e-05,
+      "loss": 1.687,
+      "step": 198
+    },
+    {
+      "epoch": 0.02715981984441108,
+      "grad_norm": 0.29573750495910645,
+      "learning_rate": 5.306666666666667e-05,
+      "loss": 1.677,
+      "step": 199
+    },
+    {
+      "epoch": 0.027296301351166917,
+      "grad_norm": 0.2537699341773987,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 1.6774,
+      "step": 200
+    },
+    {
+      "epoch": 0.027432782857922753,
+      "grad_norm": 0.3485255539417267,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 1.7344,
+      "step": 201
+    },
+    {
+      "epoch": 0.027569264364678586,
+      "grad_norm": 0.21835796535015106,
+      "learning_rate": 5.386666666666667e-05,
+      "loss": 1.7329,
+      "step": 202
+    },
+    {
+      "epoch": 0.02770574587143442,
+      "grad_norm": 0.26980435848236084,
+      "learning_rate": 5.413333333333333e-05,
+      "loss": 1.7202,
+      "step": 203
+    },
+    {
+      "epoch": 0.027842227378190254,
+      "grad_norm": 0.24076232314109802,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 1.7953,
+      "step": 204
+    },
+    {
+      "epoch": 0.02797870888494609,
+      "grad_norm": 0.25215744972229004,
+      "learning_rate": 5.466666666666667e-05,
+      "loss": 1.7256,
+      "step": 205
+    },
+    {
+      "epoch": 0.028115190391701926,
+      "grad_norm": 0.24413655698299408,
+      "learning_rate": 5.493333333333334e-05,
+      "loss": 1.6929,
+      "step": 206
+    },
+    {
+      "epoch": 0.02825167189845776,
+      "grad_norm": 0.22124704718589783,
+      "learning_rate": 5.52e-05,
+      "loss": 1.7886,
+      "step": 207
+    },
+    {
+      "epoch": 0.028388153405213595,
+      "grad_norm": 0.21877169609069824,
+      "learning_rate": 5.5466666666666675e-05,
+      "loss": 1.6658,
+      "step": 208
+    },
+    {
+      "epoch": 0.028524634911969427,
+      "grad_norm": 0.2238980531692505,
+      "learning_rate": 5.573333333333334e-05,
+      "loss": 1.7004,
+      "step": 209
+    },
+    {
+      "epoch": 0.028661116418725263,
+      "grad_norm": 0.21076123416423798,
+      "learning_rate": 5.6e-05,
+      "loss": 1.7027,
+      "step": 210
+    },
+    {
+      "epoch": 0.028797597925481096,
+      "grad_norm": 0.21802055835723877,
+      "learning_rate": 5.6266666666666675e-05,
+      "loss": 1.7392,
+      "step": 211
+    },
+    {
+      "epoch": 0.02893407943223693,
+      "grad_norm": 0.212049663066864,
+      "learning_rate": 5.6533333333333336e-05,
+      "loss": 1.6786,
+      "step": 212
+    },
+    {
+      "epoch": 0.029070560938992768,
+      "grad_norm": 0.22911016643047333,
+      "learning_rate": 5.6800000000000005e-05,
+      "loss": 1.6352,
+      "step": 213
+    },
+    {
+      "epoch": 0.0292070424457486,
+      "grad_norm": 0.2185114473104477,
+      "learning_rate": 5.7066666666666674e-05,
+      "loss": 1.5794,
+      "step": 214
+    },
+    {
+      "epoch": 0.029343523952504436,
+      "grad_norm": 0.253623366355896,
+      "learning_rate": 5.7333333333333336e-05,
+      "loss": 1.6806,
+      "step": 215
+    },
+    {
+      "epoch": 0.02948000545926027,
+      "grad_norm": 0.251220703125,
+      "learning_rate": 5.7600000000000004e-05,
+      "loss": 1.648,
+      "step": 216
+    },
+    {
+      "epoch": 0.029616486966016105,
+      "grad_norm": 0.24054937064647675,
+      "learning_rate": 5.786666666666667e-05,
+      "loss": 1.7466,
+      "step": 217
+    },
+    {
+      "epoch": 0.02975296847277194,
+      "grad_norm": 0.23784852027893066,
+      "learning_rate": 5.813333333333334e-05,
+      "loss": 1.6687,
+      "step": 218
+    },
+    {
+      "epoch": 0.029889449979527773,
+      "grad_norm": 0.25286269187927246,
+      "learning_rate": 5.84e-05,
+      "loss": 1.6895,
+      "step": 219
+    },
+    {
+      "epoch": 0.03002593148628361,
+      "grad_norm": 0.22769373655319214,
+      "learning_rate": 5.8666666666666665e-05,
+      "loss": 1.6349,
+      "step": 220
+    },
+    {
+      "epoch": 0.030162412993039442,
+      "grad_norm": 0.24668146669864655,
+      "learning_rate": 5.893333333333334e-05,
+      "loss": 1.7242,
+      "step": 221
+    },
+    {
+      "epoch": 0.030298894499795278,
+      "grad_norm": 0.2401885688304901,
+      "learning_rate": 5.92e-05,
+      "loss": 1.8042,
+      "step": 222
+    },
+    {
+      "epoch": 0.030435376006551114,
+      "grad_norm": 0.25869131088256836,
+      "learning_rate": 5.946666666666667e-05,
+      "loss": 1.7239,
+      "step": 223
+    },
+    {
+      "epoch": 0.030571857513306946,
+      "grad_norm": 0.23416532576084137,
+      "learning_rate": 5.973333333333334e-05,
+      "loss": 1.6702,
+      "step": 224
+    },
+    {
+      "epoch": 0.030708339020062782,
+      "grad_norm": 0.2530698776245117,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 1.5992,
+      "step": 225
+    },
+    {
+      "epoch": 0.030844820526818615,
+      "grad_norm": 0.24807973206043243,
+      "learning_rate": 6.026666666666667e-05,
+      "loss": 1.6091,
+      "step": 226
+    },
+    {
+      "epoch": 0.03098130203357445,
+      "grad_norm": 0.2626461088657379,
+      "learning_rate": 6.053333333333334e-05,
+      "loss": 1.7514,
+      "step": 227
+    },
+    {
+      "epoch": 0.031117783540330284,
+      "grad_norm": 0.25457173585891724,
+      "learning_rate": 6.080000000000001e-05,
+      "loss": 1.6573,
+      "step": 228
+    },
+    {
+      "epoch": 0.03125426504708612,
+      "grad_norm": 0.21489518880844116,
+      "learning_rate": 6.106666666666667e-05,
+      "loss": 1.6021,
+      "step": 229
+    },
+    {
+      "epoch": 0.031390746553841956,
+      "grad_norm": 0.2610913813114166,
+      "learning_rate": 6.133333333333334e-05,
+      "loss": 1.7039,
+      "step": 230
+    },
+    {
+      "epoch": 0.03152722806059779,
+      "grad_norm": 0.21213537454605103,
+      "learning_rate": 6.16e-05,
+      "loss": 1.6823,
+      "step": 231
+    },
+    {
+      "epoch": 0.03166370956735362,
+      "grad_norm": 0.20221945643424988,
+      "learning_rate": 6.186666666666668e-05,
+      "loss": 1.6794,
+      "step": 232
+    },
+    {
+      "epoch": 0.03180019107410946,
+      "grad_norm": 0.22009679675102234,
+      "learning_rate": 6.213333333333333e-05,
+      "loss": 1.7155,
+      "step": 233
+    },
+    {
+      "epoch": 0.03193667258086529,
+      "grad_norm": 0.20505869388580322,
+      "learning_rate": 6.240000000000001e-05,
+      "loss": 1.6534,
+      "step": 234
+    },
+    {
+      "epoch": 0.032073154087621125,
+      "grad_norm": 0.2528541684150696,
+      "learning_rate": 6.266666666666667e-05,
+      "loss": 1.6833,
+      "step": 235
+    },
+    {
+      "epoch": 0.032209635594376965,
+      "grad_norm": 0.2703731954097748,
+      "learning_rate": 6.293333333333334e-05,
+      "loss": 1.68,
+      "step": 236
+    },
+    {
+      "epoch": 0.0323461171011328,
+      "grad_norm": 0.20772017538547516,
+      "learning_rate": 6.32e-05,
+      "loss": 1.7331,
+      "step": 237
+    },
+    {
+      "epoch": 0.03248259860788863,
+      "grad_norm": 0.21209511160850525,
+      "learning_rate": 6.346666666666667e-05,
+      "loss": 1.5974,
+      "step": 238
+    },
+    {
+      "epoch": 0.03261908011464446,
+      "grad_norm": 0.25466233491897583,
+      "learning_rate": 6.373333333333334e-05,
+      "loss": 1.7179,
+      "step": 239
+    },
+    {
+      "epoch": 0.0327555616214003,
+      "grad_norm": 0.19708532094955444,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.6266,
+      "step": 240
+    },
+    {
+      "epoch": 0.032892043128156134,
+      "grad_norm": 0.2628290057182312,
+      "learning_rate": 6.426666666666668e-05,
+      "loss": 1.7026,
+      "step": 241
+    },
+    {
+      "epoch": 0.03302852463491197,
+      "grad_norm": 0.22539164125919342,
+      "learning_rate": 6.453333333333333e-05,
+      "loss": 1.6611,
+      "step": 242
+    },
+    {
+      "epoch": 0.033165006141667806,
+      "grad_norm": 0.21600398421287537,
+      "learning_rate": 6.48e-05,
+      "loss": 1.6971,
+      "step": 243
+    },
+    {
+      "epoch": 0.03330148764842364,
+      "grad_norm": 0.2313871830701828,
+      "learning_rate": 6.506666666666667e-05,
+      "loss": 1.6648,
+      "step": 244
+    },
+    {
+      "epoch": 0.03343796915517947,
+      "grad_norm": 0.22528918087482452,
+      "learning_rate": 6.533333333333334e-05,
+      "loss": 1.6178,
+      "step": 245
+    },
+    {
+      "epoch": 0.03357445066193531,
+      "grad_norm": 0.2156699150800705,
+      "learning_rate": 6.56e-05,
+      "loss": 1.7208,
+      "step": 246
+    },
+    {
+      "epoch": 0.03371093216869114,
+      "grad_norm": 0.2433416247367859,
+      "learning_rate": 6.586666666666668e-05,
+      "loss": 1.7256,
+      "step": 247
+    },
+    {
+      "epoch": 0.033847413675446976,
+      "grad_norm": 0.2248915582895279,
+      "learning_rate": 6.613333333333333e-05,
+      "loss": 1.6664,
+      "step": 248
+    },
+    {
+      "epoch": 0.03398389518220281,
+      "grad_norm": 0.23145483434200287,
+      "learning_rate": 6.64e-05,
+      "loss": 1.6347,
+      "step": 249
+    },
+    {
+      "epoch": 0.03412037668895865,
+      "grad_norm": 0.19734616577625275,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.5771,
+      "step": 250
+    },
+    {
+      "epoch": 0.03425685819571448,
+      "grad_norm": 0.2143116295337677,
+      "learning_rate": 6.693333333333334e-05,
+      "loss": 1.6928,
+      "step": 251
+    },
+    {
+      "epoch": 0.03439333970247031,
+      "grad_norm": 0.22231534123420715,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 1.6467,
+      "step": 252
+    },
+    {
+      "epoch": 0.03452982120922615,
+      "grad_norm": 0.20797906816005707,
+      "learning_rate": 6.746666666666668e-05,
+      "loss": 1.5702,
+      "step": 253
+    },
+    {
+      "epoch": 0.034666302715981985,
+      "grad_norm": 0.24098797142505646,
+      "learning_rate": 6.773333333333334e-05,
+      "loss": 1.6262,
+      "step": 254
+    },
+    {
+      "epoch": 0.03480278422273782,
+      "grad_norm": 0.2452200949192047,
+      "learning_rate": 6.8e-05,
+      "loss": 1.7232,
+      "step": 255
+    },
+    {
+      "epoch": 0.03493926572949366,
+      "grad_norm": 0.2327650785446167,
+      "learning_rate": 6.826666666666668e-05,
+      "loss": 1.7224,
+      "step": 256
+    },
+    {
+      "epoch": 0.03507574723624949,
+      "grad_norm": 0.2117166817188263,
+      "learning_rate": 6.853333333333334e-05,
+      "loss": 1.6533,
+      "step": 257
+    },
+    {
+      "epoch": 0.03521222874300532,
+      "grad_norm": 0.24029532074928284,
+      "learning_rate": 6.88e-05,
+      "loss": 1.6022,
+      "step": 258
+    },
+    {
+      "epoch": 0.035348710249761155,
+      "grad_norm": 0.24144768714904785,
+      "learning_rate": 6.906666666666667e-05,
+      "loss": 1.705,
+      "step": 259
+    },
+    {
+      "epoch": 0.035485191756516994,
+      "grad_norm": 0.2214532494544983,
+      "learning_rate": 6.933333333333334e-05,
+      "loss": 1.7441,
+      "step": 260
+    },
+    {
+      "epoch": 0.03562167326327283,
+      "grad_norm": 0.21449518203735352,
+      "learning_rate": 6.960000000000001e-05,
+      "loss": 1.6511,
+      "step": 261
+    },
+    {
+      "epoch": 0.03575815477002866,
+      "grad_norm": 0.22649604082107544,
+      "learning_rate": 6.986666666666667e-05,
+      "loss": 1.6649,
+      "step": 262
+    },
+    {
+      "epoch": 0.0358946362767845,
+      "grad_norm": 0.2125379592180252,
+      "learning_rate": 7.013333333333335e-05,
+      "loss": 1.6957,
+      "step": 263
+    },
+    {
+      "epoch": 0.03603111778354033,
+      "grad_norm": 0.23201030492782593,
+      "learning_rate": 7.04e-05,
+      "loss": 1.6779,
+      "step": 264
+    },
+    {
+      "epoch": 0.036167599290296164,
+      "grad_norm": 0.20203745365142822,
+      "learning_rate": 7.066666666666667e-05,
+      "loss": 1.6503,
+      "step": 265
+    },
+    {
+      "epoch": 0.036304080797051996,
+      "grad_norm": 0.23546801507472992,
+      "learning_rate": 7.093333333333334e-05,
+      "loss": 1.6396,
+      "step": 266
+    },
+    {
+      "epoch": 0.036440562303807836,
+      "grad_norm": 0.20308132469654083,
+      "learning_rate": 7.120000000000001e-05,
+      "loss": 1.7182,
+      "step": 267
+    },
+    {
+      "epoch": 0.03657704381056367,
+      "grad_norm": 0.20700064301490784,
+      "learning_rate": 7.146666666666666e-05,
+      "loss": 1.7149,
+      "step": 268
+    },
+    {
+      "epoch": 0.0367135253173195,
+      "grad_norm": 0.2265264242887497,
+      "learning_rate": 7.173333333333333e-05,
+      "loss": 1.7122,
+      "step": 269
+    },
+    {
+      "epoch": 0.03685000682407534,
+      "grad_norm": 0.21799160540103912,
+      "learning_rate": 7.2e-05,
+      "loss": 1.6572,
+      "step": 270
+    },
+    {
+      "epoch": 0.03698648833083117,
+      "grad_norm": 0.19743822515010834,
+      "learning_rate": 7.226666666666667e-05,
+      "loss": 1.6004,
+      "step": 271
+    },
+    {
+      "epoch": 0.037122969837587005,
+      "grad_norm": 0.19909922778606415,
+      "learning_rate": 7.253333333333334e-05,
+      "loss": 1.5925,
+      "step": 272
+    },
+    {
+      "epoch": 0.037259451344342845,
+      "grad_norm": 0.20454707741737366,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 1.7318,
+      "step": 273
+    },
+    {
+      "epoch": 0.03739593285109868,
+      "grad_norm": 0.19905130565166473,
+      "learning_rate": 7.306666666666668e-05,
+      "loss": 1.6399,
+      "step": 274
+    },
+    {
+      "epoch": 0.03753241435785451,
+      "grad_norm": 0.22123555839061737,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 1.6797,
+      "step": 275
+    },
+    {
+      "epoch": 0.03766889586461034,
+      "grad_norm": 0.2101968228816986,
+      "learning_rate": 7.360000000000001e-05,
+      "loss": 1.6601,
+      "step": 276
+    },
+    {
+      "epoch": 0.03780537737136618,
+      "grad_norm": 0.2130962312221527,
+      "learning_rate": 7.386666666666667e-05,
+      "loss": 1.734,
+      "step": 277
+    },
+    {
+      "epoch": 0.037941858878122015,
+      "grad_norm": 0.21536697447299957,
+      "learning_rate": 7.413333333333334e-05,
+      "loss": 1.7315,
+      "step": 278
+    },
+    {
+      "epoch": 0.03807834038487785,
+      "grad_norm": 0.21114717423915863,
+      "learning_rate": 7.44e-05,
+      "loss": 1.6886,
+      "step": 279
+    },
+    {
+      "epoch": 0.03821482189163369,
+      "grad_norm": 0.24319419264793396,
+      "learning_rate": 7.466666666666667e-05,
+      "loss": 1.7375,
+      "step": 280
+    },
+    {
+      "epoch": 0.03835130339838952,
+      "grad_norm": 0.19769835472106934,
+      "learning_rate": 7.493333333333334e-05,
+      "loss": 1.7083,
+      "step": 281
+    },
+    {
+      "epoch": 0.03848778490514535,
+      "grad_norm": 0.22592894732952118,
+      "learning_rate": 7.52e-05,
+      "loss": 1.6407,
+      "step": 282
+    },
+    {
+      "epoch": 0.038624266411901184,
+      "grad_norm": 0.20605631172657013,
+      "learning_rate": 7.546666666666668e-05,
+      "loss": 1.6803,
+      "step": 283
+    },
+    {
+      "epoch": 0.038760747918657024,
+      "grad_norm": 0.21297262609004974,
+      "learning_rate": 7.573333333333334e-05,
+      "loss": 1.6462,
+      "step": 284
+    },
+    {
+      "epoch": 0.038897229425412856,
+      "grad_norm": 0.20373934507369995,
+      "learning_rate": 7.6e-05,
+      "loss": 1.6578,
+      "step": 285
+    },
+    {
+      "epoch": 0.03903371093216869,
+      "grad_norm": 0.2054784893989563,
+      "learning_rate": 7.626666666666667e-05,
+      "loss": 1.7021,
+      "step": 286
+    },
+    {
+      "epoch": 0.03917019243892453,
+      "grad_norm": 0.20568403601646423,
+      "learning_rate": 7.653333333333334e-05,
+      "loss": 1.7562,
+      "step": 287
+    },
+    {
+      "epoch": 0.03930667394568036,
+      "grad_norm": 0.19960707426071167,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 1.8018,
+      "step": 288
+    },
+    {
+      "epoch": 0.03944315545243619,
+      "grad_norm": 0.2213381826877594,
+      "learning_rate": 7.706666666666668e-05,
+      "loss": 1.6819,
+      "step": 289
+    },
+    {
+      "epoch": 0.03957963695919203,
+      "grad_norm": 0.21227072179317474,
+      "learning_rate": 7.733333333333333e-05,
+      "loss": 1.718,
+      "step": 290
+    },
+    {
+      "epoch": 0.039716118465947865,
+      "grad_norm": 0.21815580129623413,
+      "learning_rate": 7.76e-05,
+      "loss": 1.7343,
+      "step": 291
+    },
+    {
+      "epoch": 0.0398525999727037,
+      "grad_norm": 0.2301357537508011,
+      "learning_rate": 7.786666666666667e-05,
+      "loss": 1.6573,
+      "step": 292
+    },
+    {
+      "epoch": 0.03998908147945953,
+      "grad_norm": 0.21771447360515594,
+      "learning_rate": 7.813333333333334e-05,
+      "loss": 1.7112,
+      "step": 293
+    },
+    {
+      "epoch": 0.04012556298621537,
+      "grad_norm": 0.196007639169693,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 1.6526,
+      "step": 294
+    },
+    {
+      "epoch": 0.0402620444929712,
+      "grad_norm": 0.21248257160186768,
+      "learning_rate": 7.866666666666666e-05,
+      "loss": 1.7492,
+      "step": 295
+    },
+    {
+      "epoch": 0.040398525999727035,
+      "grad_norm": 0.20819489657878876,
+      "learning_rate": 7.893333333333335e-05,
+      "loss": 1.6847,
+      "step": 296
+    },
+    {
+      "epoch": 0.040535007506482874,
+      "grad_norm": 0.20487208664417267,
+      "learning_rate": 7.92e-05,
+      "loss": 1.6649,
+      "step": 297
+    },
+    {
+      "epoch": 0.04067148901323871,
+      "grad_norm": 0.21150562167167664,
+      "learning_rate": 7.946666666666667e-05,
+      "loss": 1.6469,
+      "step": 298
+    },
+    {
+      "epoch": 0.04080797051999454,
+      "grad_norm": 0.24073141813278198,
+      "learning_rate": 7.973333333333334e-05,
+      "loss": 1.6623,
+      "step": 299
+    },
+    {
+      "epoch": 0.04094445202675037,
+      "grad_norm": 0.24552752077579498,
+      "learning_rate": 8e-05,
+      "loss": 1.6755,
+      "step": 300
+    },
+    {
+      "epoch": 0.04108093353350621,
+      "grad_norm": 0.21898119151592255,
+      "learning_rate": 8e-05,
+      "loss": 1.7248,
+      "step": 301
+    },
+    {
+      "epoch": 0.041217415040262044,
+      "grad_norm": 0.21144811809062958,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 302
+    },
+    {
+      "epoch": 0.04135389654701788,
+      "grad_norm": 0.20077790319919586,
+      "learning_rate": 8e-05,
+      "loss": 1.6616,
+      "step": 303
+    },
+    {
+      "epoch": 0.041490378053773716,
+      "grad_norm": 0.21370148658752441,
+      "learning_rate": 8e-05,
+      "loss": 1.6461,
+      "step": 304
+    },
+    {
+      "epoch": 0.04162685956052955,
+      "grad_norm": 0.19295500218868256,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 305
+    },
+    {
+      "epoch": 0.04176334106728538,
+      "grad_norm": 0.19603842496871948,
+      "learning_rate": 8e-05,
+      "loss": 1.5694,
+      "step": 306
+    },
+    {
+      "epoch": 0.04189982257404122,
+      "grad_norm": 0.22422997653484344,
+      "learning_rate": 8e-05,
+      "loss": 1.6848,
+      "step": 307
+    },
+    {
+      "epoch": 0.04203630408079705,
+      "grad_norm": 0.20466266572475433,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 308
+    },
+    {
+      "epoch": 0.042172785587552886,
+      "grad_norm": 0.2288111001253128,
+      "learning_rate": 8e-05,
+      "loss": 1.7092,
+      "step": 309
+    },
+    {
+      "epoch": 0.04230926709430872,
+      "grad_norm": 0.2034597247838974,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 310
+    },
+    {
+      "epoch": 0.04244574860106456,
+      "grad_norm": 0.20546230673789978,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 311
+    },
+    {
+      "epoch": 0.04258223010782039,
+      "grad_norm": 0.22445982694625854,
+      "learning_rate": 8e-05,
+      "loss": 1.7117,
+      "step": 312
+    },
+    {
+      "epoch": 0.04271871161457622,
+      "grad_norm": 0.20283575356006622,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 313
+    },
+    {
+      "epoch": 0.04285519312133206,
+      "grad_norm": 0.2159210741519928,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 314
+    },
+    {
+      "epoch": 0.042991674628087895,
+      "grad_norm": 0.20850765705108643,
+      "learning_rate": 8e-05,
+      "loss": 1.6698,
+      "step": 315
+    },
+    {
+      "epoch": 0.04312815613484373,
+      "grad_norm": 0.20244640111923218,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 316
+    },
+    {
+      "epoch": 0.04326463764159956,
+      "grad_norm": 0.1934514194726944,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 317
+    },
+    {
+      "epoch": 0.0434011191483554,
+      "grad_norm": 0.1918637752532959,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 318
+    },
+    {
+      "epoch": 0.04353760065511123,
+      "grad_norm": 0.20331163704395294,
+      "learning_rate": 8e-05,
+      "loss": 1.7058,
+      "step": 319
+    },
+    {
+      "epoch": 0.043674082161867064,
+      "grad_norm": 0.19439738988876343,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 320
+    },
+    {
+      "epoch": 0.043810563668622904,
+      "grad_norm": 0.20036499202251434,
+      "learning_rate": 8e-05,
+      "loss": 1.7089,
+      "step": 321
+    },
+    {
+      "epoch": 0.043947045175378736,
+      "grad_norm": 0.1999746859073639,
+      "learning_rate": 8e-05,
+      "loss": 1.6931,
+      "step": 322
+    },
+    {
+      "epoch": 0.04408352668213457,
+      "grad_norm": 0.18918439745903015,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 323
+    },
+    {
+      "epoch": 0.04422000818889041,
+      "grad_norm": 0.19035042822360992,
+      "learning_rate": 8e-05,
+      "loss": 1.6692,
+      "step": 324
+    },
+    {
+      "epoch": 0.04435648969564624,
+      "grad_norm": 0.21209511160850525,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 325
+    },
+    {
+      "epoch": 0.044492971202402073,
+      "grad_norm": 0.20302721858024597,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 326
+    },
+    {
+      "epoch": 0.044629452709157906,
+      "grad_norm": 0.21007706224918365,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 327
+    },
+    {
+      "epoch": 0.044765934215913745,
+      "grad_norm": 0.1938336044549942,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 328
+    },
+    {
+      "epoch": 0.04490241572266958,
+      "grad_norm": 0.2358740270137787,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 329
+    },
+    {
+      "epoch": 0.04503889722942541,
+      "grad_norm": 0.20879460871219635,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 330
+    },
+    {
+      "epoch": 0.04517537873618125,
+      "grad_norm": 0.19917111098766327,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 331
+    },
+    {
+      "epoch": 0.04531186024293708,
+      "grad_norm": 0.19454148411750793,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 332
+    },
+    {
+      "epoch": 0.045448341749692915,
+      "grad_norm": 0.2147936373949051,
+      "learning_rate": 8e-05,
+      "loss": 1.6701,
+      "step": 333
+    },
+    {
+      "epoch": 0.045584823256448755,
+      "grad_norm": 0.2066538780927658,
+      "learning_rate": 8e-05,
+      "loss": 1.6621,
+      "step": 334
+    },
+    {
+      "epoch": 0.04572130476320459,
+      "grad_norm": 0.1907842755317688,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 335
+    },
+    {
+      "epoch": 0.04585778626996042,
+      "grad_norm": 0.19874846935272217,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 336
+    },
+    {
+      "epoch": 0.04599426777671625,
+      "grad_norm": 0.18912295997142792,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 337
+    },
+    {
+      "epoch": 0.04613074928347209,
+      "grad_norm": 0.20593127608299255,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 338
+    },
+    {
+      "epoch": 0.046267230790227924,
+      "grad_norm": 0.22252845764160156,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 339
+    },
+    {
+      "epoch": 0.04640371229698376,
+      "grad_norm": 0.18263423442840576,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 340
+    },
+    {
+      "epoch": 0.046540193803739596,
+      "grad_norm": 0.23404349386692047,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 341
+    },
+    {
+      "epoch": 0.04667667531049543,
+      "grad_norm": 0.19697609543800354,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 342
+    },
+    {
+      "epoch": 0.04681315681725126,
+      "grad_norm": 0.22630178928375244,
+      "learning_rate": 8e-05,
+      "loss": 1.6312,
+      "step": 343
+    },
+    {
+      "epoch": 0.046949638324007094,
+      "grad_norm": 0.21530590951442719,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 344
+    },
+    {
+      "epoch": 0.04708611983076293,
+      "grad_norm": 0.2517217695713043,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 345
+    },
+    {
+      "epoch": 0.047222601337518766,
+      "grad_norm": 0.20627780258655548,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 346
+    },
+    {
+      "epoch": 0.0473590828442746,
+      "grad_norm": 0.25718629360198975,
+      "learning_rate": 8e-05,
+      "loss": 1.6864,
+      "step": 347
+    },
+    {
+      "epoch": 0.04749556435103044,
+      "grad_norm": 0.20818424224853516,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 348
+    },
+    {
+      "epoch": 0.04763204585778627,
+      "grad_norm": 0.24214956164360046,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 349
+    },
+    {
+      "epoch": 0.0477685273645421,
+      "grad_norm": 0.19357354938983917,
+      "learning_rate": 8e-05,
+      "loss": 1.5822,
+      "step": 350
+    },
+    {
+      "epoch": 0.04790500887129794,
+      "grad_norm": 0.24596109986305237,
+      "learning_rate": 8e-05,
+      "loss": 1.7334,
+      "step": 351
+    },
+    {
+      "epoch": 0.048041490378053775,
+      "grad_norm": 0.19316093623638153,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 352
+    },
+    {
+      "epoch": 0.04817797188480961,
+      "grad_norm": 0.21627448499202728,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 353
+    },
+    {
+      "epoch": 0.04831445339156544,
+      "grad_norm": 0.19135534763336182,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 354
+    },
+    {
+      "epoch": 0.04845093489832128,
+      "grad_norm": 0.22228562831878662,
+      "learning_rate": 8e-05,
+      "loss": 1.6874,
+      "step": 355
+    },
+    {
+      "epoch": 0.04858741640507711,
+      "grad_norm": 0.2129838615655899,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 356
+    },
+    {
+      "epoch": 0.048723897911832945,
+      "grad_norm": 0.20773866772651672,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 357
+    },
+    {
+      "epoch": 0.048860379418588784,
+      "grad_norm": 0.23375366628170013,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 358
+    },
+    {
+      "epoch": 0.04899686092534462,
+      "grad_norm": 0.20398132503032684,
+      "learning_rate": 8e-05,
+      "loss": 1.6864,
+      "step": 359
+    },
+    {
+      "epoch": 0.04913334243210045,
+      "grad_norm": 0.2301444411277771,
+      "learning_rate": 8e-05,
+      "loss": 1.5876,
+      "step": 360
+    },
+    {
+      "epoch": 0.04926982393885628,
+      "grad_norm": 0.19762130081653595,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 361
+    },
+    {
+      "epoch": 0.04940630544561212,
+      "grad_norm": 0.2054339498281479,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 362
+    },
+    {
+      "epoch": 0.049542786952367954,
+      "grad_norm": 0.20791882276535034,
+      "learning_rate": 8e-05,
+      "loss": 1.635,
+      "step": 363
+    },
+    {
+      "epoch": 0.049679268459123786,
+      "grad_norm": 0.2073269635438919,
+      "learning_rate": 8e-05,
+      "loss": 1.5745,
+      "step": 364
+    },
+    {
+      "epoch": 0.049815749965879626,
+      "grad_norm": 0.2151254564523697,
+      "learning_rate": 8e-05,
+      "loss": 1.7144,
+      "step": 365
+    },
+    {
+      "epoch": 0.04995223147263546,
+      "grad_norm": 0.22530192136764526,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 366
+    },
+    {
+      "epoch": 0.05008871297939129,
+      "grad_norm": 0.1957068145275116,
+      "learning_rate": 8e-05,
+      "loss": 1.6484,
+      "step": 367
+    },
+    {
+      "epoch": 0.05022519448614713,
+      "grad_norm": 0.18599842488765717,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 368
+    },
+    {
+      "epoch": 0.05036167599290296,
+      "grad_norm": 0.20138153433799744,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 369
+    },
+    {
+      "epoch": 0.050498157499658795,
+      "grad_norm": 0.18490669131278992,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 370
+    },
+    {
+      "epoch": 0.05063463900641463,
+      "grad_norm": 0.225370392203331,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 371
+    },
+    {
+      "epoch": 0.05077112051317047,
+      "grad_norm": 0.18147234618663788,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 372
+    },
+    {
+      "epoch": 0.0509076020199263,
+      "grad_norm": 0.26411259174346924,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 373
+    },
+    {
+      "epoch": 0.05104408352668213,
+      "grad_norm": 0.18885082006454468,
+      "learning_rate": 8e-05,
+      "loss": 1.6966,
+      "step": 374
+    },
+    {
+      "epoch": 0.05118056503343797,
+      "grad_norm": 0.23470209538936615,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 375
+    },
+    {
+      "epoch": 0.051317046540193804,
+      "grad_norm": 0.1856028288602829,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 376
+    },
+    {
+      "epoch": 0.05145352804694964,
+      "grad_norm": 0.1803678572177887,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 377
+    },
+    {
+      "epoch": 0.05159000955370547,
+      "grad_norm": 0.20237888395786285,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 378
+    },
+    {
+      "epoch": 0.05172649106046131,
+      "grad_norm": 0.18280954658985138,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 379
+    },
+    {
+      "epoch": 0.05186297256721714,
+      "grad_norm": 0.22067183256149292,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 380
+    },
+    {
+      "epoch": 0.051999454073972974,
+      "grad_norm": 0.2018761783838272,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 381
+    },
+    {
+      "epoch": 0.052135935580728814,
+      "grad_norm": 0.22056815028190613,
+      "learning_rate": 8e-05,
+      "loss": 1.6127,
+      "step": 382
+    },
+    {
+      "epoch": 0.052272417087484646,
+      "grad_norm": 0.18608184158802032,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 383
+    },
+    {
+      "epoch": 0.05240889859424048,
+      "grad_norm": 0.21474720537662506,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 384
+    },
+    {
+      "epoch": 0.05254538010099632,
+      "grad_norm": 0.1835620254278183,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 385
+    },
+    {
+      "epoch": 0.05268186160775215,
+      "grad_norm": 0.1864248812198639,
+      "learning_rate": 8e-05,
+      "loss": 1.709,
+      "step": 386
+    },
+    {
+      "epoch": 0.05281834311450798,
+      "grad_norm": 0.17684119939804077,
+      "learning_rate": 8e-05,
+      "loss": 1.652,
+      "step": 387
+    },
+    {
+      "epoch": 0.052954824621263816,
+      "grad_norm": 0.19974467158317566,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 388
+    },
+    {
+      "epoch": 0.053091306128019655,
+      "grad_norm": 0.22409911453723907,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 389
+    },
+    {
+      "epoch": 0.05322778763477549,
+      "grad_norm": 0.19997046887874603,
+      "learning_rate": 8e-05,
+      "loss": 1.6593,
+      "step": 390
+    },
+    {
+      "epoch": 0.05336426914153132,
+      "grad_norm": 0.24276980757713318,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 391
+    },
+    {
+      "epoch": 0.05350075064828716,
+      "grad_norm": 0.19641022384166718,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 392
+    },
+    {
+      "epoch": 0.05363723215504299,
+      "grad_norm": 0.2161017209291458,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 393
+    },
+    {
+      "epoch": 0.053773713661798825,
+      "grad_norm": 0.21501685678958893,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 394
+    },
+    {
+      "epoch": 0.05391019516855466,
+      "grad_norm": 0.19006696343421936,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 395
+    },
+    {
+      "epoch": 0.0540466766753105,
+      "grad_norm": 0.21929454803466797,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 396
+    },
+    {
+      "epoch": 0.05418315818206633,
+      "grad_norm": 0.1907109171152115,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 397
+    },
+    {
+      "epoch": 0.05431963968882216,
+      "grad_norm": 0.19140149652957916,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 398
+    },
+    {
+      "epoch": 0.054456121195578,
+      "grad_norm": 0.20610398054122925,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 399
+    },
+    {
+      "epoch": 0.054592602702333834,
+      "grad_norm": 0.19578817486763,
+      "learning_rate": 8e-05,
+      "loss": 1.5467,
+      "step": 400
+    },
+    {
+      "epoch": 0.054729084209089666,
+      "grad_norm": 0.18162508308887482,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 401
+    },
+    {
+      "epoch": 0.054865565715845506,
+      "grad_norm": 0.18938252329826355,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 402
+    },
+    {
+      "epoch": 0.05500204722260134,
+      "grad_norm": 0.18807187676429749,
+      "learning_rate": 8e-05,
+      "loss": 1.639,
+      "step": 403
+    },
+    {
+      "epoch": 0.05513852872935717,
+      "grad_norm": 0.19044460356235504,
+      "learning_rate": 8e-05,
+      "loss": 1.6816,
+      "step": 404
+    },
+    {
+      "epoch": 0.055275010236113004,
+      "grad_norm": 0.18897344172000885,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 405
+    },
+    {
+      "epoch": 0.05541149174286884,
+      "grad_norm": 0.187645822763443,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 406
+    },
+    {
+      "epoch": 0.055547973249624676,
+      "grad_norm": 0.192600280046463,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 407
+    },
+    {
+      "epoch": 0.05568445475638051,
+      "grad_norm": 0.1796467900276184,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 408
+    },
+    {
+      "epoch": 0.05582093626313635,
+      "grad_norm": 0.19574713706970215,
+      "learning_rate": 8e-05,
+      "loss": 1.602,
+      "step": 409
+    },
+    {
+      "epoch": 0.05595741776989218,
+      "grad_norm": 0.19001144170761108,
+      "learning_rate": 8e-05,
+      "loss": 1.6127,
+      "step": 410
+    },
+    {
+      "epoch": 0.05609389927664801,
+      "grad_norm": 0.18304583430290222,
+      "learning_rate": 8e-05,
+      "loss": 1.5418,
+      "step": 411
+    },
+    {
+      "epoch": 0.05623038078340385,
+      "grad_norm": 0.17753589153289795,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 412
+    },
+    {
+      "epoch": 0.056366862290159685,
+      "grad_norm": 0.19941185414791107,
+      "learning_rate": 8e-05,
+      "loss": 1.7627,
+      "step": 413
+    },
+    {
+      "epoch": 0.05650334379691552,
+      "grad_norm": 0.20377519726753235,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 414
+    },
+    {
+      "epoch": 0.05663982530367135,
+      "grad_norm": 0.1892452836036682,
+      "learning_rate": 8e-05,
+      "loss": 1.6369,
+      "step": 415
+    },
+    {
+      "epoch": 0.05677630681042719,
+      "grad_norm": 0.18668115139007568,
+      "learning_rate": 8e-05,
+      "loss": 1.7098,
+      "step": 416
+    },
+    {
+      "epoch": 0.05691278831718302,
+      "grad_norm": 0.19333577156066895,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 417
+    },
+    {
+      "epoch": 0.057049269823938854,
+      "grad_norm": 0.18673676252365112,
+      "learning_rate": 8e-05,
+      "loss": 1.673,
+      "step": 418
+    },
+    {
+      "epoch": 0.057185751330694694,
+      "grad_norm": 0.18598119914531708,
+      "learning_rate": 8e-05,
+      "loss": 1.7121,
+      "step": 419
+    },
+    {
+      "epoch": 0.057322232837450526,
+      "grad_norm": 0.18558545410633087,
+      "learning_rate": 8e-05,
+      "loss": 1.682,
+      "step": 420
+    },
+    {
+      "epoch": 0.05745871434420636,
+      "grad_norm": 0.18987078964710236,
+      "learning_rate": 8e-05,
+      "loss": 1.7683,
+      "step": 421
+    },
+    {
+      "epoch": 0.05759519585096219,
+      "grad_norm": 0.18435579538345337,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 422
+    },
+    {
+      "epoch": 0.05773167735771803,
+      "grad_norm": 0.182847261428833,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 423
+    },
+    {
+      "epoch": 0.05786815886447386,
+      "grad_norm": 0.1804504543542862,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 424
+    },
+    {
+      "epoch": 0.058004640371229696,
+      "grad_norm": 0.18960043787956238,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 425
+    },
+    {
+      "epoch": 0.058141121877985535,
+      "grad_norm": 0.19457489252090454,
+      "learning_rate": 8e-05,
+      "loss": 1.7394,
+      "step": 426
+    },
+    {
+      "epoch": 0.05827760338474137,
+      "grad_norm": 0.19392524659633636,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 427
+    },
+    {
+      "epoch": 0.0584140848914972,
+      "grad_norm": 0.18370689451694489,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 428
+    },
+    {
+      "epoch": 0.05855056639825304,
+      "grad_norm": 0.19954009354114532,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 429
+    },
+    {
+      "epoch": 0.05868704790500887,
+      "grad_norm": 0.18143226206302643,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 430
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 0.19837979972362518,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 431
+    },
+    {
+      "epoch": 0.05896001091852054,
+      "grad_norm": 0.1801684945821762,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 432
+    },
+    {
+      "epoch": 0.05909649242527638,
+      "grad_norm": 0.18215817213058472,
+      "learning_rate": 8e-05,
+      "loss": 1.6294,
+      "step": 433
+    },
+    {
+      "epoch": 0.05923297393203221,
+      "grad_norm": 0.1929459571838379,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 434
+    },
+    {
+      "epoch": 0.05936945543878804,
+      "grad_norm": 0.1846276819705963,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 435
+    },
+    {
+      "epoch": 0.05950593694554388,
+      "grad_norm": 0.18948525190353394,
+      "learning_rate": 8e-05,
+      "loss": 1.713,
+      "step": 436
+    },
+    {
+      "epoch": 0.059642418452299714,
+      "grad_norm": 0.18023772537708282,
+      "learning_rate": 8e-05,
+      "loss": 1.5822,
+      "step": 437
+    },
+    {
+      "epoch": 0.05977889995905555,
+      "grad_norm": 0.18806883692741394,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 438
+    },
+    {
+      "epoch": 0.05991538146581138,
+      "grad_norm": 0.18786229193210602,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 439
+    },
+    {
+      "epoch": 0.06005186297256722,
+      "grad_norm": 0.18303732573986053,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 440
+    },
+    {
+      "epoch": 0.06018834447932305,
+      "grad_norm": 0.1819573938846588,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 441
+    },
+    {
+      "epoch": 0.060324825986078884,
+      "grad_norm": 0.1820116937160492,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 442
+    },
+    {
+      "epoch": 0.06046130749283472,
+      "grad_norm": 0.1831539124250412,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 443
+    },
+    {
+      "epoch": 0.060597788999590556,
+      "grad_norm": 0.189011350274086,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 444
+    },
+    {
+      "epoch": 0.06073427050634639,
+      "grad_norm": 0.19471584260463715,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 445
+    },
+    {
+      "epoch": 0.06087075201310223,
+      "grad_norm": 0.18301334977149963,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 446
+    },
+    {
+      "epoch": 0.06100723351985806,
+      "grad_norm": 0.17497512698173523,
+      "learning_rate": 8e-05,
+      "loss": 1.5936,
+      "step": 447
+    },
+    {
+      "epoch": 0.06114371502661389,
+      "grad_norm": 0.19188492000102997,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 448
+    },
+    {
+      "epoch": 0.061280196533369725,
+      "grad_norm": 0.18681305646896362,
+      "learning_rate": 8e-05,
+      "loss": 1.7043,
+      "step": 449
+    },
+    {
+      "epoch": 0.061416678040125565,
+      "grad_norm": 0.18630671501159668,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 450
+    },
+    {
+      "epoch": 0.0615531595468814,
+      "grad_norm": 0.17838414013385773,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 451
+    },
+    {
+      "epoch": 0.06168964105363723,
+      "grad_norm": 0.1982869654893875,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 452
+    },
+    {
+      "epoch": 0.06182612256039307,
+      "grad_norm": 0.18733154237270355,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 453
+    },
+    {
+      "epoch": 0.0619626040671489,
+      "grad_norm": 0.1786513477563858,
+      "learning_rate": 8e-05,
+      "loss": 1.6243,
+      "step": 454
+    },
+    {
+      "epoch": 0.062099085573904735,
+      "grad_norm": 0.17923349142074585,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 455
+    },
+    {
+      "epoch": 0.06223556708066057,
+      "grad_norm": 0.18081489205360413,
+      "learning_rate": 8e-05,
+      "loss": 1.6823,
+      "step": 456
+    },
+    {
+      "epoch": 0.06237204858741641,
+      "grad_norm": 0.1793363094329834,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 457
+    },
+    {
+      "epoch": 0.06250853009417225,
+      "grad_norm": 0.17861542105674744,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 458
+    },
+    {
+      "epoch": 0.06264501160092807,
+      "grad_norm": 0.1745217889547348,
+      "learning_rate": 8e-05,
+      "loss": 1.6596,
+      "step": 459
+    },
+    {
+      "epoch": 0.06278149310768391,
+      "grad_norm": 0.19002461433410645,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 460
+    },
+    {
+      "epoch": 0.06291797461443974,
+      "grad_norm": 0.18763870000839233,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 461
+    },
+    {
+      "epoch": 0.06305445612119558,
+      "grad_norm": 0.18333010375499725,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 462
+    },
+    {
+      "epoch": 0.06319093762795142,
+      "grad_norm": 0.17597143352031708,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 463
+    },
+    {
+      "epoch": 0.06332741913470724,
+      "grad_norm": 0.18492625653743744,
+      "learning_rate": 8e-05,
+      "loss": 1.679,
+      "step": 464
+    },
+    {
+      "epoch": 0.06346390064146308,
+      "grad_norm": 0.17959477007389069,
+      "learning_rate": 8e-05,
+      "loss": 1.6545,
+      "step": 465
+    },
+    {
+      "epoch": 0.06360038214821892,
+      "grad_norm": 0.19852040708065033,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 466
+    },
+    {
+      "epoch": 0.06373686365497475,
+      "grad_norm": 0.17028813064098358,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 467
+    },
+    {
+      "epoch": 0.06387334516173059,
+      "grad_norm": 0.2159101963043213,
+      "learning_rate": 8e-05,
+      "loss": 1.6208,
+      "step": 468
+    },
+    {
+      "epoch": 0.06400982666848642,
+      "grad_norm": 0.18861988186836243,
+      "learning_rate": 8e-05,
+      "loss": 1.729,
+      "step": 469
+    },
+    {
+      "epoch": 0.06414630817524225,
+      "grad_norm": 0.19807402789592743,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 470
+    },
+    {
+      "epoch": 0.06428278968199809,
+      "grad_norm": 0.19683124125003815,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 471
+    },
+    {
+      "epoch": 0.06441927118875393,
+      "grad_norm": 0.22598282992839813,
+      "learning_rate": 8e-05,
+      "loss": 1.7043,
+      "step": 472
+    },
+    {
+      "epoch": 0.06455575269550975,
+      "grad_norm": 0.19435809552669525,
+      "learning_rate": 8e-05,
+      "loss": 1.6718,
+      "step": 473
+    },
+    {
+      "epoch": 0.0646922342022656,
+      "grad_norm": 0.19861872494220734,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 474
+    },
+    {
+      "epoch": 0.06482871570902143,
+      "grad_norm": 0.19188061356544495,
+      "learning_rate": 8e-05,
+      "loss": 1.6291,
+      "step": 475
+    },
+    {
+      "epoch": 0.06496519721577726,
+      "grad_norm": 0.18942494690418243,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 476
+    },
+    {
+      "epoch": 0.0651016787225331,
+      "grad_norm": 0.21525967121124268,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 477
+    },
+    {
+      "epoch": 0.06523816022928892,
+      "grad_norm": 0.19176410138607025,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 478
+    },
+    {
+      "epoch": 0.06537464173604476,
+      "grad_norm": 0.20893029868602753,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 479
+    },
+    {
+      "epoch": 0.0655111232428006,
+      "grad_norm": 0.22374695539474487,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 480
+    },
+    {
+      "epoch": 0.06564760474955643,
+      "grad_norm": 0.18589884042739868,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 481
+    },
+    {
+      "epoch": 0.06578408625631227,
+      "grad_norm": 0.23982296884059906,
+      "learning_rate": 8e-05,
+      "loss": 1.6712,
+      "step": 482
+    },
+    {
+      "epoch": 0.06592056776306811,
+      "grad_norm": 0.1724051833152771,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 483
+    },
+    {
+      "epoch": 0.06605704926982393,
+      "grad_norm": 0.2297288030385971,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 484
+    },
+    {
+      "epoch": 0.06619353077657977,
+      "grad_norm": 0.19763821363449097,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 485
+    },
+    {
+      "epoch": 0.06633001228333561,
+      "grad_norm": 0.21263867616653442,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 486
+    },
+    {
+      "epoch": 0.06646649379009144,
+      "grad_norm": 0.22302626073360443,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 487
+    },
+    {
+      "epoch": 0.06660297529684728,
+      "grad_norm": 0.18238939344882965,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 488
+    },
+    {
+      "epoch": 0.06673945680360312,
+      "grad_norm": 0.22283275425434113,
+      "learning_rate": 8e-05,
+      "loss": 1.554,
+      "step": 489
+    },
+    {
+      "epoch": 0.06687593831035894,
+      "grad_norm": 0.18292540311813354,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 490
+    },
+    {
+      "epoch": 0.06701241981711478,
+      "grad_norm": 0.21135155856609344,
+      "learning_rate": 8e-05,
+      "loss": 1.6997,
+      "step": 491
+    },
+    {
+      "epoch": 0.06714890132387062,
+      "grad_norm": 0.18980979919433594,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 492
+    },
+    {
+      "epoch": 0.06728538283062645,
+      "grad_norm": 0.17444701492786407,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 493
+    },
+    {
+      "epoch": 0.06742186433738229,
+      "grad_norm": 0.19433224201202393,
+      "learning_rate": 8e-05,
+      "loss": 1.6231,
+      "step": 494
+    },
+    {
+      "epoch": 0.06755834584413811,
+      "grad_norm": 0.18066327273845673,
+      "learning_rate": 8e-05,
+      "loss": 1.7068,
+      "step": 495
+    },
+    {
+      "epoch": 0.06769482735089395,
+      "grad_norm": 0.17841613292694092,
+      "learning_rate": 8e-05,
+      "loss": 1.5933,
+      "step": 496
+    },
+    {
+      "epoch": 0.06783130885764979,
+      "grad_norm": 0.1893261969089508,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 497
+    },
+    {
+      "epoch": 0.06796779036440562,
+      "grad_norm": 0.1816590577363968,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 498
+    },
+    {
+      "epoch": 0.06810427187116146,
+      "grad_norm": 0.18823958933353424,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 499
+    },
+    {
+      "epoch": 0.0682407533779173,
+      "grad_norm": 0.19860830903053284,
+      "learning_rate": 8e-05,
+      "loss": 1.6583,
+      "step": 500
+    },
+    {
+      "epoch": 0.06837723488467312,
+      "grad_norm": 0.18241938948631287,
+      "learning_rate": 8e-05,
+      "loss": 1.6137,
+      "step": 501
+    },
+    {
+      "epoch": 0.06851371639142896,
+      "grad_norm": 0.20241890847682953,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 502
+    },
+    {
+      "epoch": 0.0686501978981848,
+      "grad_norm": 0.1916690170764923,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 503
+    },
+    {
+      "epoch": 0.06878667940494063,
+      "grad_norm": 0.17784589529037476,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 504
+    },
+    {
+      "epoch": 0.06892316091169647,
+      "grad_norm": 0.18047195672988892,
+      "learning_rate": 8e-05,
+      "loss": 1.6797,
+      "step": 505
+    },
+    {
+      "epoch": 0.0690596424184523,
+      "grad_norm": 0.1743379384279251,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 506
+    },
+    {
+      "epoch": 0.06919612392520813,
+      "grad_norm": 0.1787814497947693,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 507
+    },
+    {
+      "epoch": 0.06933260543196397,
+      "grad_norm": 0.17678914964199066,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 508
+    },
+    {
+      "epoch": 0.06946908693871981,
+      "grad_norm": 0.19351589679718018,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 509
+    },
+    {
+      "epoch": 0.06960556844547564,
+      "grad_norm": 0.1810530722141266,
+      "learning_rate": 8e-05,
+      "loss": 1.6316,
+      "step": 510
+    },
+    {
+      "epoch": 0.06974204995223147,
+      "grad_norm": 0.1901681274175644,
+      "learning_rate": 8e-05,
+      "loss": 1.6253,
+      "step": 511
+    },
+    {
+      "epoch": 0.06987853145898731,
+      "grad_norm": 0.18215714395046234,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 512
+    },
+    {
+      "epoch": 0.07001501296574314,
+      "grad_norm": 0.1821824312210083,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 513
+    },
+    {
+      "epoch": 0.07015149447249898,
+      "grad_norm": 0.1817176342010498,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 514
+    },
+    {
+      "epoch": 0.0702879759792548,
+      "grad_norm": 0.189289852976799,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 515
+    },
+    {
+      "epoch": 0.07042445748601064,
+      "grad_norm": 0.19207283854484558,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 516
+    },
+    {
+      "epoch": 0.07056093899276648,
+      "grad_norm": 0.1871204674243927,
+      "learning_rate": 8e-05,
+      "loss": 1.5979,
+      "step": 517
+    },
+    {
+      "epoch": 0.07069742049952231,
+      "grad_norm": 0.2023966759443283,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 518
+    },
+    {
+      "epoch": 0.07083390200627815,
+      "grad_norm": 0.19608986377716064,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 519
+    },
+    {
+      "epoch": 0.07097038351303399,
+      "grad_norm": 0.2101517766714096,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 520
+    },
+    {
+      "epoch": 0.07110686501978981,
+      "grad_norm": 0.1760852038860321,
+      "learning_rate": 8e-05,
+      "loss": 1.5946,
+      "step": 521
+    },
+    {
+      "epoch": 0.07124334652654565,
+      "grad_norm": 0.2021082639694214,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 522
+    },
+    {
+      "epoch": 0.07137982803330149,
+      "grad_norm": 0.18128426373004913,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 523
+    },
+    {
+      "epoch": 0.07151630954005732,
+      "grad_norm": 0.18913501501083374,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 524
+    },
+    {
+      "epoch": 0.07165279104681316,
+      "grad_norm": 0.1942007690668106,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 525
+    },
+    {
+      "epoch": 0.071789272553569,
+      "grad_norm": 0.18365976214408875,
+      "learning_rate": 8e-05,
+      "loss": 1.6474,
+      "step": 526
+    },
+    {
+      "epoch": 0.07192575406032482,
+      "grad_norm": 0.19202566146850586,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 527
+    },
+    {
+      "epoch": 0.07206223556708066,
+      "grad_norm": 0.1858273595571518,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 528
+    },
+    {
+      "epoch": 0.0721987170738365,
+      "grad_norm": 0.17584769427776337,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 529
+    },
+    {
+      "epoch": 0.07233519858059233,
+      "grad_norm": 0.18194857239723206,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 530
+    },
+    {
+      "epoch": 0.07247168008734817,
+      "grad_norm": 0.18180221319198608,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 531
+    },
+    {
+      "epoch": 0.07260816159410399,
+      "grad_norm": 0.19153547286987305,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 532
+    },
+    {
+      "epoch": 0.07274464310085983,
+      "grad_norm": 0.1788415163755417,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 533
+    },
+    {
+      "epoch": 0.07288112460761567,
+      "grad_norm": 0.19157685339450836,
+      "learning_rate": 8e-05,
+      "loss": 1.6113,
+      "step": 534
+    },
+    {
+      "epoch": 0.0730176061143715,
+      "grad_norm": 0.18316860496997833,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 535
+    },
+    {
+      "epoch": 0.07315408762112734,
+      "grad_norm": 0.17753863334655762,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 536
+    },
+    {
+      "epoch": 0.07329056912788318,
+      "grad_norm": 0.19196350872516632,
+      "learning_rate": 8e-05,
+      "loss": 1.6801,
+      "step": 537
+    },
+    {
+      "epoch": 0.073427050634639,
+      "grad_norm": 0.18240304291248322,
+      "learning_rate": 8e-05,
+      "loss": 1.6567,
+      "step": 538
+    },
+    {
+      "epoch": 0.07356353214139484,
+      "grad_norm": 0.1779155284166336,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 539
+    },
+    {
+      "epoch": 0.07370001364815068,
+      "grad_norm": 0.17845813930034637,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 540
+    },
+    {
+      "epoch": 0.0738364951549065,
+      "grad_norm": 0.18534943461418152,
+      "learning_rate": 8e-05,
+      "loss": 1.6085,
+      "step": 541
+    },
+    {
+      "epoch": 0.07397297666166235,
+      "grad_norm": 0.17992828786373138,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 542
+    },
+    {
+      "epoch": 0.07410945816841819,
+      "grad_norm": 0.19209209084510803,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 543
+    },
+    {
+      "epoch": 0.07424593967517401,
+      "grad_norm": 0.18622608482837677,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 544
+    },
+    {
+      "epoch": 0.07438242118192985,
+      "grad_norm": 0.1822274625301361,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 545
+    },
+    {
+      "epoch": 0.07451890268868569,
+      "grad_norm": 0.19041560590267181,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 546
+    },
+    {
+      "epoch": 0.07465538419544152,
+      "grad_norm": 0.18942587077617645,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 547
+    },
+    {
+      "epoch": 0.07479186570219735,
+      "grad_norm": 0.1978999376296997,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 548
+    },
+    {
+      "epoch": 0.07492834720895318,
+      "grad_norm": 0.19729527831077576,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 549
+    },
+    {
+      "epoch": 0.07506482871570902,
+      "grad_norm": 0.20745132863521576,
+      "learning_rate": 8e-05,
+      "loss": 1.6587,
+      "step": 550
+    },
+    {
+      "epoch": 0.07520131022246486,
+      "grad_norm": 0.17871299386024475,
+      "learning_rate": 8e-05,
+      "loss": 1.6033,
+      "step": 551
+    },
+    {
+      "epoch": 0.07533779172922069,
+      "grad_norm": 0.2038271129131317,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 552
+    },
+    {
+      "epoch": 0.07547427323597652,
+      "grad_norm": 0.17945469915866852,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 553
+    },
+    {
+      "epoch": 0.07561075474273236,
+      "grad_norm": 0.19727687537670135,
+      "learning_rate": 8e-05,
+      "loss": 1.6736,
+      "step": 554
+    },
+    {
+      "epoch": 0.07574723624948819,
+      "grad_norm": 0.1985924392938614,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 555
+    },
+    {
+      "epoch": 0.07588371775624403,
+      "grad_norm": 0.18431121110916138,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 556
+    },
+    {
+      "epoch": 0.07602019926299987,
+      "grad_norm": 0.19703145325183868,
+      "learning_rate": 8e-05,
+      "loss": 1.6593,
+      "step": 557
+    },
+    {
+      "epoch": 0.0761566807697557,
+      "grad_norm": 0.1927427351474762,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 558
+    },
+    {
+      "epoch": 0.07629316227651153,
+      "grad_norm": 0.20637810230255127,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 559
+    },
+    {
+      "epoch": 0.07642964378326737,
+      "grad_norm": 0.19992417097091675,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 560
+    },
+    {
+      "epoch": 0.0765661252900232,
+      "grad_norm": 0.18012253940105438,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 561
+    },
+    {
+      "epoch": 0.07670260679677904,
+      "grad_norm": 0.194854736328125,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 562
+    },
+    {
+      "epoch": 0.07683908830353488,
+      "grad_norm": 0.18546807765960693,
+      "learning_rate": 8e-05,
+      "loss": 1.7072,
+      "step": 563
+    },
+    {
+      "epoch": 0.0769755698102907,
+      "grad_norm": 0.18692143261432648,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 564
+    },
+    {
+      "epoch": 0.07711205131704654,
+      "grad_norm": 0.20474644005298615,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 565
+    },
+    {
+      "epoch": 0.07724853282380237,
+      "grad_norm": 0.16820207238197327,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 566
+    },
+    {
+      "epoch": 0.07738501433055821,
+      "grad_norm": 0.19693829119205475,
+      "learning_rate": 8e-05,
+      "loss": 1.6249,
+      "step": 567
+    },
+    {
+      "epoch": 0.07752149583731405,
+      "grad_norm": 0.1874043047428131,
+      "learning_rate": 8e-05,
+      "loss": 1.6213,
+      "step": 568
+    },
+    {
+      "epoch": 0.07765797734406987,
+      "grad_norm": 0.18235719203948975,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 569
+    },
+    {
+      "epoch": 0.07779445885082571,
+      "grad_norm": 0.18116523325443268,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 570
+    },
+    {
+      "epoch": 0.07793094035758155,
+      "grad_norm": 0.18194203078746796,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 571
+    },
+    {
+      "epoch": 0.07806742186433738,
+      "grad_norm": 0.17890216410160065,
+      "learning_rate": 8e-05,
+      "loss": 1.5576,
+      "step": 572
+    },
+    {
+      "epoch": 0.07820390337109322,
+      "grad_norm": 0.21269476413726807,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 573
+    },
+    {
+      "epoch": 0.07834038487784906,
+      "grad_norm": 0.17812982201576233,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 574
+    },
+    {
+      "epoch": 0.07847686638460488,
+      "grad_norm": 0.20546594262123108,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 575
+    },
+    {
+      "epoch": 0.07861334789136072,
+      "grad_norm": 0.1874275803565979,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 576
+    },
+    {
+      "epoch": 0.07874982939811656,
+      "grad_norm": 0.17828841507434845,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 577
+    },
+    {
+      "epoch": 0.07888631090487239,
+      "grad_norm": 0.19493460655212402,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 578
+    },
+    {
+      "epoch": 0.07902279241162823,
+      "grad_norm": 0.17887650430202484,
+      "learning_rate": 8e-05,
+      "loss": 1.6464,
+      "step": 579
+    },
+    {
+      "epoch": 0.07915927391838407,
+      "grad_norm": 0.18878933787345886,
+      "learning_rate": 8e-05,
+      "loss": 1.6293,
+      "step": 580
+    },
+    {
+      "epoch": 0.07929575542513989,
+      "grad_norm": 0.17782963812351227,
+      "learning_rate": 8e-05,
+      "loss": 1.6023,
+      "step": 581
+    },
+    {
+      "epoch": 0.07943223693189573,
+      "grad_norm": 0.18788425624370575,
+      "learning_rate": 8e-05,
+      "loss": 1.6826,
+      "step": 582
+    },
+    {
+      "epoch": 0.07956871843865156,
+      "grad_norm": 0.1840958595275879,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 583
+    },
+    {
+      "epoch": 0.0797051999454074,
+      "grad_norm": 0.1809142827987671,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 584
+    },
+    {
+      "epoch": 0.07984168145216324,
+      "grad_norm": 0.17688719928264618,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 585
+    },
+    {
+      "epoch": 0.07997816295891906,
+      "grad_norm": 0.178343266248703,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 586
+    },
+    {
+      "epoch": 0.0801146444656749,
+      "grad_norm": 0.19121424853801727,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 587
+    },
+    {
+      "epoch": 0.08025112597243074,
+      "grad_norm": 0.1846960335969925,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 588
+    },
+    {
+      "epoch": 0.08038760747918657,
+      "grad_norm": 0.1875857263803482,
+      "learning_rate": 8e-05,
+      "loss": 1.7452,
+      "step": 589
+    },
+    {
+      "epoch": 0.0805240889859424,
+      "grad_norm": 0.18118062615394592,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 590
+    },
+    {
+      "epoch": 0.08066057049269824,
+      "grad_norm": 0.17859064042568207,
+      "learning_rate": 8e-05,
+      "loss": 1.5662,
+      "step": 591
+    },
+    {
+      "epoch": 0.08079705199945407,
+      "grad_norm": 0.19599363207817078,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 592
+    },
+    {
+      "epoch": 0.08093353350620991,
+      "grad_norm": 0.17053189873695374,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 593
+    },
+    {
+      "epoch": 0.08107001501296575,
+      "grad_norm": 0.17487818002700806,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 594
+    },
+    {
+      "epoch": 0.08120649651972157,
+      "grad_norm": 0.17549864947795868,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 595
+    },
+    {
+      "epoch": 0.08134297802647741,
+      "grad_norm": 0.17001301050186157,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 596
+    },
+    {
+      "epoch": 0.08147945953323325,
+      "grad_norm": 0.17742592096328735,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 597
+    },
+    {
+      "epoch": 0.08161594103998908,
+      "grad_norm": 0.18993912637233734,
+      "learning_rate": 8e-05,
+      "loss": 1.6296,
+      "step": 598
+    },
+    {
+      "epoch": 0.08175242254674492,
+      "grad_norm": 0.18070772290229797,
+      "learning_rate": 8e-05,
+      "loss": 1.7273,
+      "step": 599
+    },
+    {
+      "epoch": 0.08188890405350074,
+      "grad_norm": 0.17393943667411804,
+      "learning_rate": 8e-05,
+      "loss": 1.5863,
+      "step": 600
+    },
+    {
+      "epoch": 0.08202538556025658,
+      "grad_norm": 0.18577931821346283,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 601
+    },
+    {
+      "epoch": 0.08216186706701242,
+      "grad_norm": 0.1849592924118042,
+      "learning_rate": 8e-05,
+      "loss": 1.6337,
+      "step": 602
+    },
+    {
+      "epoch": 0.08229834857376825,
+      "grad_norm": 0.17174462974071503,
+      "learning_rate": 8e-05,
+      "loss": 1.5765,
+      "step": 603
+    },
+    {
+      "epoch": 0.08243483008052409,
+      "grad_norm": 0.17975455522537231,
+      "learning_rate": 8e-05,
+      "loss": 1.6085,
+      "step": 604
+    },
+    {
+      "epoch": 0.08257131158727993,
+      "grad_norm": 0.18230432271957397,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 605
+    },
+    {
+      "epoch": 0.08270779309403575,
+      "grad_norm": 0.18215805292129517,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 606
+    },
+    {
+      "epoch": 0.08284427460079159,
+      "grad_norm": 0.17772619426250458,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 607
+    },
+    {
+      "epoch": 0.08298075610754743,
+      "grad_norm": 0.17819859087467194,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 608
+    },
+    {
+      "epoch": 0.08311723761430326,
+      "grad_norm": 0.18093952536582947,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 609
+    },
+    {
+      "epoch": 0.0832537191210591,
+      "grad_norm": 0.17327451705932617,
+      "learning_rate": 8e-05,
+      "loss": 1.5269,
+      "step": 610
+    },
+    {
+      "epoch": 0.08339020062781494,
+      "grad_norm": 0.18055465817451477,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 611
+    },
+    {
+      "epoch": 0.08352668213457076,
+      "grad_norm": 0.17736971378326416,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 612
+    },
+    {
+      "epoch": 0.0836631636413266,
+      "grad_norm": 0.17380128800868988,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 613
+    },
+    {
+      "epoch": 0.08379964514808244,
+      "grad_norm": 0.18247854709625244,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 614
+    },
+    {
+      "epoch": 0.08393612665483827,
+      "grad_norm": 0.17850947380065918,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 615
+    },
+    {
+      "epoch": 0.0840726081615941,
+      "grad_norm": 0.1836286187171936,
+      "learning_rate": 8e-05,
+      "loss": 1.6556,
+      "step": 616
+    },
+    {
+      "epoch": 0.08420908966834993,
+      "grad_norm": 0.18170252442359924,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 617
+    },
+    {
+      "epoch": 0.08434557117510577,
+      "grad_norm": 0.17840580642223358,
+      "learning_rate": 8e-05,
+      "loss": 1.6947,
+      "step": 618
+    },
+    {
+      "epoch": 0.08448205268186161,
+      "grad_norm": 0.1744961142539978,
+      "learning_rate": 8e-05,
+      "loss": 1.5932,
+      "step": 619
+    },
+    {
+      "epoch": 0.08461853418861744,
+      "grad_norm": 0.17901203036308289,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 620
+    },
+    {
+      "epoch": 0.08475501569537328,
+      "grad_norm": 0.17852306365966797,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 621
+    },
+    {
+      "epoch": 0.08489149720212912,
+      "grad_norm": 0.18519847095012665,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 622
+    },
+    {
+      "epoch": 0.08502797870888494,
+      "grad_norm": 0.19501279294490814,
+      "learning_rate": 8e-05,
+      "loss": 1.6459,
+      "step": 623
+    },
+    {
+      "epoch": 0.08516446021564078,
+      "grad_norm": 0.18123270571231842,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 624
+    },
+    {
+      "epoch": 0.08530094172239662,
+      "grad_norm": 0.1957874447107315,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 625
+    },
+    {
+      "epoch": 0.08543742322915245,
+      "grad_norm": 0.18587222695350647,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 626
+    },
+    {
+      "epoch": 0.08557390473590828,
+      "grad_norm": 0.18250805139541626,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 627
+    },
+    {
+      "epoch": 0.08571038624266412,
+      "grad_norm": 0.184391051530838,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 628
+    },
+    {
+      "epoch": 0.08584686774941995,
+      "grad_norm": 0.17343154549598694,
+      "learning_rate": 8e-05,
+      "loss": 1.6113,
+      "step": 629
+    },
+    {
+      "epoch": 0.08598334925617579,
+      "grad_norm": 0.21289904415607452,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 630
+    },
+    {
+      "epoch": 0.08611983076293163,
+      "grad_norm": 0.17811021208763123,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 631
+    },
+    {
+      "epoch": 0.08625631226968745,
+      "grad_norm": 0.21290110051631927,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 632
+    },
+    {
+      "epoch": 0.0863927937764433,
+      "grad_norm": 0.17709466814994812,
+      "learning_rate": 8e-05,
+      "loss": 1.629,
+      "step": 633
+    },
+    {
+      "epoch": 0.08652927528319912,
+      "grad_norm": 0.20067048072814941,
+      "learning_rate": 8e-05,
+      "loss": 1.588,
+      "step": 634
+    },
+    {
+      "epoch": 0.08666575678995496,
+      "grad_norm": 0.18714983761310577,
+      "learning_rate": 8e-05,
+      "loss": 1.6911,
+      "step": 635
+    },
+    {
+      "epoch": 0.0868022382967108,
+      "grad_norm": 0.19722506403923035,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 636
+    },
+    {
+      "epoch": 0.08693871980346662,
+      "grad_norm": 0.18553335964679718,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 637
+    },
+    {
+      "epoch": 0.08707520131022246,
+      "grad_norm": 0.18402236700057983,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 638
+    },
+    {
+      "epoch": 0.0872116828169783,
+      "grad_norm": 0.20606359839439392,
+      "learning_rate": 8e-05,
+      "loss": 1.7076,
+      "step": 639
+    },
+    {
+      "epoch": 0.08734816432373413,
+      "grad_norm": 0.1979096233844757,
+      "learning_rate": 8e-05,
+      "loss": 1.6213,
+      "step": 640
+    },
+    {
+      "epoch": 0.08748464583048997,
+      "grad_norm": 0.19418217241764069,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 641
+    },
+    {
+      "epoch": 0.08762112733724581,
+      "grad_norm": 0.17649537324905396,
+      "learning_rate": 8e-05,
+      "loss": 1.6426,
+      "step": 642
+    },
+    {
+      "epoch": 0.08775760884400163,
+      "grad_norm": 0.18922159075737,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 643
+    },
+    {
+      "epoch": 0.08789409035075747,
+      "grad_norm": 0.18971461057662964,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 644
+    },
+    {
+      "epoch": 0.08803057185751331,
+      "grad_norm": 0.1711868941783905,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 645
+    },
+    {
+      "epoch": 0.08816705336426914,
+      "grad_norm": 0.1809423565864563,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 646
+    },
+    {
+      "epoch": 0.08830353487102498,
+      "grad_norm": 0.18164674937725067,
+      "learning_rate": 8e-05,
+      "loss": 1.6545,
+      "step": 647
+    },
+    {
+      "epoch": 0.08844001637778082,
+      "grad_norm": 0.20286716520786285,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 648
+    },
+    {
+      "epoch": 0.08857649788453664,
+      "grad_norm": 0.20841902494430542,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 649
+    },
+    {
+      "epoch": 0.08871297939129248,
+      "grad_norm": 0.1873120218515396,
+      "learning_rate": 8e-05,
+      "loss": 1.6197,
+      "step": 650
+    },
+    {
+      "epoch": 0.08884946089804832,
+      "grad_norm": 0.17084191739559174,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 651
+    },
+    {
+      "epoch": 0.08898594240480415,
+      "grad_norm": 0.19740137457847595,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 652
+    },
+    {
+      "epoch": 0.08912242391155999,
+      "grad_norm": 0.19413062930107117,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 653
+    },
+    {
+      "epoch": 0.08925890541831581,
+      "grad_norm": 0.18452408909797668,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 654
+    },
+    {
+      "epoch": 0.08939538692507165,
+      "grad_norm": 0.22501632571220398,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 655
+    },
+    {
+      "epoch": 0.08953186843182749,
+      "grad_norm": 0.1789039820432663,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 656
+    },
+    {
+      "epoch": 0.08966834993858332,
+      "grad_norm": 0.19610245525836945,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 657
+    },
+    {
+      "epoch": 0.08980483144533916,
+      "grad_norm": 0.17572997510433197,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 658
+    },
+    {
+      "epoch": 0.089941312952095,
+      "grad_norm": 0.19441260397434235,
+      "learning_rate": 8e-05,
+      "loss": 1.6019,
+      "step": 659
+    },
+    {
+      "epoch": 0.09007779445885082,
+      "grad_norm": 0.2057199776172638,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 660
+    },
+    {
+      "epoch": 0.09021427596560666,
+      "grad_norm": 0.1797274947166443,
+      "learning_rate": 8e-05,
+      "loss": 1.5548,
+      "step": 661
+    },
+    {
+      "epoch": 0.0903507574723625,
+      "grad_norm": 0.19070973992347717,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 662
+    },
+    {
+      "epoch": 0.09048723897911833,
+      "grad_norm": 0.17818237841129303,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 663
+    },
+    {
+      "epoch": 0.09062372048587417,
+      "grad_norm": 0.17819111049175262,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 664
+    },
+    {
+      "epoch": 0.09076020199263,
+      "grad_norm": 0.20394210517406464,
+      "learning_rate": 8e-05,
+      "loss": 1.653,
+      "step": 665
+    },
+    {
+      "epoch": 0.09089668349938583,
+      "grad_norm": 0.18936486542224884,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 666
+    },
+    {
+      "epoch": 0.09103316500614167,
+      "grad_norm": 0.1895633190870285,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 667
+    },
+    {
+      "epoch": 0.09116964651289751,
+      "grad_norm": 0.1924016922712326,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 668
+    },
+    {
+      "epoch": 0.09130612801965333,
+      "grad_norm": 0.18355004489421844,
+      "learning_rate": 8e-05,
+      "loss": 1.5788,
+      "step": 669
+    },
+    {
+      "epoch": 0.09144260952640917,
+      "grad_norm": 0.20524947345256805,
+      "learning_rate": 8e-05,
+      "loss": 1.6324,
+      "step": 670
+    },
+    {
+      "epoch": 0.091579091033165,
+      "grad_norm": 0.17362678050994873,
+      "learning_rate": 8e-05,
+      "loss": 1.5638,
+      "step": 671
+    },
+    {
+      "epoch": 0.09171557253992084,
+      "grad_norm": 0.1757184863090515,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 672
+    },
+    {
+      "epoch": 0.09185205404667668,
+      "grad_norm": 0.20039305090904236,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 673
+    },
+    {
+      "epoch": 0.0919885355534325,
+      "grad_norm": 0.17184749245643616,
+      "learning_rate": 8e-05,
+      "loss": 1.6395,
+      "step": 674
+    },
+    {
+      "epoch": 0.09212501706018834,
+      "grad_norm": 0.18741856515407562,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 675
+    },
+    {
+      "epoch": 0.09226149856694418,
+      "grad_norm": 0.19409061968326569,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 676
+    },
+    {
+      "epoch": 0.09239798007370001,
+      "grad_norm": 0.17097792029380798,
+      "learning_rate": 8e-05,
+      "loss": 1.5881,
+      "step": 677
+    },
+    {
+      "epoch": 0.09253446158045585,
+      "grad_norm": 0.1736527979373932,
+      "learning_rate": 8e-05,
+      "loss": 1.6135,
+      "step": 678
+    },
+    {
+      "epoch": 0.09267094308721169,
+      "grad_norm": 0.18509618937969208,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 679
+    },
+    {
+      "epoch": 0.09280742459396751,
+      "grad_norm": 0.17875801026821136,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 680
+    },
+    {
+      "epoch": 0.09294390610072335,
+      "grad_norm": 0.16205383837223053,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 681
+    },
+    {
+      "epoch": 0.09308038760747919,
+      "grad_norm": 0.17233863472938538,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 682
+    },
+    {
+      "epoch": 0.09321686911423502,
+      "grad_norm": 0.17813430726528168,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 683
+    },
+    {
+      "epoch": 0.09335335062099086,
+      "grad_norm": 0.1721436083316803,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 684
+    },
+    {
+      "epoch": 0.0934898321277467,
+      "grad_norm": 0.16886721551418304,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 685
+    },
+    {
+      "epoch": 0.09362631363450252,
+      "grad_norm": 0.17673741281032562,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 686
+    },
+    {
+      "epoch": 0.09376279514125836,
+      "grad_norm": 0.18254360556602478,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 687
+    },
+    {
+      "epoch": 0.09389927664801419,
+      "grad_norm": 0.1704188734292984,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 688
+    },
+    {
+      "epoch": 0.09403575815477003,
+      "grad_norm": 0.1748236119747162,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 689
+    },
+    {
+      "epoch": 0.09417223966152587,
+      "grad_norm": 0.18408794701099396,
+      "learning_rate": 8e-05,
+      "loss": 1.7105,
+      "step": 690
+    },
+    {
+      "epoch": 0.09430872116828169,
+      "grad_norm": 0.17438945174217224,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 691
+    },
+    {
+      "epoch": 0.09444520267503753,
+      "grad_norm": 0.17807906866073608,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 692
+    },
+    {
+      "epoch": 0.09458168418179337,
+      "grad_norm": 0.17539091408252716,
+      "learning_rate": 8e-05,
+      "loss": 1.613,
+      "step": 693
+    },
+    {
+      "epoch": 0.0947181656885492,
+      "grad_norm": 0.17115943133831024,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 694
+    },
+    {
+      "epoch": 0.09485464719530504,
+      "grad_norm": 0.18094389140605927,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 695
+    },
+    {
+      "epoch": 0.09499112870206088,
+      "grad_norm": 0.16728800535202026,
+      "learning_rate": 8e-05,
+      "loss": 1.5564,
+      "step": 696
+    },
+    {
+      "epoch": 0.0951276102088167,
+      "grad_norm": 0.17686955630779266,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 697
+    },
+    {
+      "epoch": 0.09526409171557254,
+      "grad_norm": 0.1698547601699829,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 698
+    },
+    {
+      "epoch": 0.09540057322232838,
+      "grad_norm": 0.17156901955604553,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 699
+    },
+    {
+      "epoch": 0.0955370547290842,
+      "grad_norm": 0.16910618543624878,
+      "learning_rate": 8e-05,
+      "loss": 1.6377,
+      "step": 700
+    },
+    {
+      "epoch": 0.09567353623584005,
+      "grad_norm": 0.1788112223148346,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 701
+    },
+    {
+      "epoch": 0.09581001774259588,
+      "grad_norm": 0.1680026799440384,
+      "learning_rate": 8e-05,
+      "loss": 1.5867,
+      "step": 702
+    },
+    {
+      "epoch": 0.09594649924935171,
+      "grad_norm": 0.17168425023555756,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 703
+    },
+    {
+      "epoch": 0.09608298075610755,
+      "grad_norm": 0.1807289719581604,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 704
+    },
+    {
+      "epoch": 0.09621946226286338,
+      "grad_norm": 0.1735907942056656,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 705
+    },
+    {
+      "epoch": 0.09635594376961922,
+      "grad_norm": 0.1873815804719925,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 706
+    },
+    {
+      "epoch": 0.09649242527637505,
+      "grad_norm": 0.17822548747062683,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 707
+    },
+    {
+      "epoch": 0.09662890678313088,
+      "grad_norm": 0.17290425300598145,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 708
+    },
+    {
+      "epoch": 0.09676538828988672,
+      "grad_norm": 0.19037611782550812,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 709
+    },
+    {
+      "epoch": 0.09690186979664256,
+      "grad_norm": 0.18396949768066406,
+      "learning_rate": 8e-05,
+      "loss": 1.6543,
+      "step": 710
+    },
+    {
+      "epoch": 0.09703835130339838,
+      "grad_norm": 0.19072912633419037,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 711
+    },
+    {
+      "epoch": 0.09717483281015422,
+      "grad_norm": 0.1756344735622406,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 712
+    },
+    {
+      "epoch": 0.09731131431691006,
+      "grad_norm": 0.17327451705932617,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 713
+    },
+    {
+      "epoch": 0.09744779582366589,
+      "grad_norm": 0.17609016597270966,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 714
+    },
+    {
+      "epoch": 0.09758427733042173,
+      "grad_norm": 0.17689797282218933,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 715
+    },
+    {
+      "epoch": 0.09772075883717757,
+      "grad_norm": 0.18054063618183136,
+      "learning_rate": 8e-05,
+      "loss": 1.6543,
+      "step": 716
+    },
+    {
+      "epoch": 0.0978572403439334,
+      "grad_norm": 0.17576654255390167,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 717
+    },
+    {
+      "epoch": 0.09799372185068923,
+      "grad_norm": 0.18059571087360382,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 718
+    },
+    {
+      "epoch": 0.09813020335744507,
+      "grad_norm": 0.18142476677894592,
+      "learning_rate": 8e-05,
+      "loss": 1.6408,
+      "step": 719
+    },
+    {
+      "epoch": 0.0982666848642009,
+      "grad_norm": 0.18520772457122803,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 720
+    },
+    {
+      "epoch": 0.09840316637095674,
+      "grad_norm": 0.18081459403038025,
+      "learning_rate": 8e-05,
+      "loss": 1.6467,
+      "step": 721
+    },
+    {
+      "epoch": 0.09853964787771256,
+      "grad_norm": 0.18610171973705292,
+      "learning_rate": 8e-05,
+      "loss": 1.7697,
+      "step": 722
+    },
+    {
+      "epoch": 0.0986761293844684,
+      "grad_norm": 0.17639221251010895,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 723
+    },
+    {
+      "epoch": 0.09881261089122424,
+      "grad_norm": 0.17549027502536774,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 724
+    },
+    {
+      "epoch": 0.09894909239798007,
+      "grad_norm": 0.18425285816192627,
+      "learning_rate": 8e-05,
+      "loss": 1.6699,
+      "step": 725
+    },
+    {
+      "epoch": 0.09908557390473591,
+      "grad_norm": 0.17593851685523987,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 726
+    },
+    {
+      "epoch": 0.09922205541149175,
+      "grad_norm": 0.18171188235282898,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 727
+    },
+    {
+      "epoch": 0.09935853691824757,
+      "grad_norm": 0.18206368386745453,
+      "learning_rate": 8e-05,
+      "loss": 1.593,
+      "step": 728
+    },
+    {
+      "epoch": 0.09949501842500341,
+      "grad_norm": 0.18444791436195374,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 729
+    },
+    {
+      "epoch": 0.09963149993175925,
+      "grad_norm": 0.19138328731060028,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 730
+    },
+    {
+      "epoch": 0.09976798143851508,
+      "grad_norm": 0.1718161404132843,
+      "learning_rate": 8e-05,
+      "loss": 1.5442,
+      "step": 731
+    },
+    {
+      "epoch": 0.09990446294527092,
+      "grad_norm": 0.18484805524349213,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 732
+    },
+    {
+      "epoch": 0.10004094445202676,
+      "grad_norm": 0.18484321236610413,
+      "learning_rate": 8e-05,
+      "loss": 1.5446,
+      "step": 733
+    },
+    {
+      "epoch": 0.10017742595878258,
+      "grad_norm": 0.1844317615032196,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 734
+    },
+    {
+      "epoch": 0.10031390746553842,
+      "grad_norm": 0.17879654467105865,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 735
+    },
+    {
+      "epoch": 0.10045038897229426,
+      "grad_norm": 0.1806444376707077,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 736
+    },
+    {
+      "epoch": 0.10058687047905009,
+      "grad_norm": 0.17929135262966156,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 737
+    },
+    {
+      "epoch": 0.10072335198580593,
+      "grad_norm": 0.1743006408214569,
+      "learning_rate": 8e-05,
+      "loss": 1.5847,
+      "step": 738
+    },
+    {
+      "epoch": 0.10085983349256175,
+      "grad_norm": 0.18439467251300812,
+      "learning_rate": 8e-05,
+      "loss": 1.6701,
+      "step": 739
+    },
+    {
+      "epoch": 0.10099631499931759,
+      "grad_norm": 0.18599484860897064,
+      "learning_rate": 8e-05,
+      "loss": 1.6073,
+      "step": 740
+    },
+    {
+      "epoch": 0.10113279650607343,
+      "grad_norm": 0.1775076687335968,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 741
+    },
+    {
+      "epoch": 0.10126927801282926,
+      "grad_norm": 0.18221992254257202,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 742
+    },
+    {
+      "epoch": 0.1014057595195851,
+      "grad_norm": 0.19071851670742035,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 743
+    },
+    {
+      "epoch": 0.10154224102634093,
+      "grad_norm": 0.1738080382347107,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 744
+    },
+    {
+      "epoch": 0.10167872253309676,
+      "grad_norm": 0.18030865490436554,
+      "learning_rate": 8e-05,
+      "loss": 1.6202,
+      "step": 745
+    },
+    {
+      "epoch": 0.1018152040398526,
+      "grad_norm": 0.18551528453826904,
+      "learning_rate": 8e-05,
+      "loss": 1.5975,
+      "step": 746
+    },
+    {
+      "epoch": 0.10195168554660844,
+      "grad_norm": 0.1839742213487625,
+      "learning_rate": 8e-05,
+      "loss": 1.7085,
+      "step": 747
+    },
+    {
+      "epoch": 0.10208816705336426,
+      "grad_norm": 0.16842563450336456,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 748
+    },
+    {
+      "epoch": 0.1022246485601201,
+      "grad_norm": 0.18717491626739502,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 749
+    },
+    {
+      "epoch": 0.10236113006687594,
+      "grad_norm": 0.1846681535243988,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 750
+    },
+    {
+      "epoch": 0.10249761157363177,
+      "grad_norm": 0.19289425015449524,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 751
+    },
+    {
+      "epoch": 0.10263409308038761,
+      "grad_norm": 0.1740955263376236,
+      "learning_rate": 8e-05,
+      "loss": 1.5315,
+      "step": 752
+    },
+    {
+      "epoch": 0.10277057458714345,
+      "grad_norm": 0.18069280683994293,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 753
+    },
+    {
+      "epoch": 0.10290705609389927,
+      "grad_norm": 0.17482346296310425,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 754
+    },
+    {
+      "epoch": 0.10304353760065511,
+      "grad_norm": 0.1850505769252777,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 755
+    },
+    {
+      "epoch": 0.10318001910741094,
+      "grad_norm": 0.17505916953086853,
+      "learning_rate": 8e-05,
+      "loss": 1.6234,
+      "step": 756
+    },
+    {
+      "epoch": 0.10331650061416678,
+      "grad_norm": 0.17290765047073364,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 757
+    },
+    {
+      "epoch": 0.10345298212092262,
+      "grad_norm": 0.17706461250782013,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 758
+    },
+    {
+      "epoch": 0.10358946362767844,
+      "grad_norm": 0.18021456897258759,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 759
+    },
+    {
+      "epoch": 0.10372594513443428,
+      "grad_norm": 0.176732137799263,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 760
+    },
+    {
+      "epoch": 0.10386242664119012,
+      "grad_norm": 0.17645902931690216,
+      "learning_rate": 8e-05,
+      "loss": 1.6628,
+      "step": 761
+    },
+    {
+      "epoch": 0.10399890814794595,
+      "grad_norm": 0.17248065769672394,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 762
+    },
+    {
+      "epoch": 0.10413538965470179,
+      "grad_norm": 0.18161921203136444,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 763
+    },
+    {
+      "epoch": 0.10427187116145763,
+      "grad_norm": 0.1725803017616272,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 764
+    },
+    {
+      "epoch": 0.10440835266821345,
+      "grad_norm": 0.18166513741016388,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 765
+    },
+    {
+      "epoch": 0.10454483417496929,
+      "grad_norm": 0.17848937213420868,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 766
+    },
+    {
+      "epoch": 0.10468131568172513,
+      "grad_norm": 0.1779319941997528,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 767
+    },
+    {
+      "epoch": 0.10481779718848096,
+      "grad_norm": 0.17400763928890228,
+      "learning_rate": 8e-05,
+      "loss": 1.5908,
+      "step": 768
+    },
+    {
+      "epoch": 0.1049542786952368,
+      "grad_norm": 0.1692686676979065,
+      "learning_rate": 8e-05,
+      "loss": 1.5829,
+      "step": 769
+    },
+    {
+      "epoch": 0.10509076020199264,
+      "grad_norm": 0.18350636959075928,
+      "learning_rate": 8e-05,
+      "loss": 1.6285,
+      "step": 770
+    },
+    {
+      "epoch": 0.10522724170874846,
+      "grad_norm": 0.1727180778980255,
+      "learning_rate": 8e-05,
+      "loss": 1.6248,
+      "step": 771
+    },
+    {
+      "epoch": 0.1053637232155043,
+      "grad_norm": 0.19921256601810455,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 772
+    },
+    {
+      "epoch": 0.10550020472226013,
+      "grad_norm": 0.1811063140630722,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 773
+    },
+    {
+      "epoch": 0.10563668622901597,
+      "grad_norm": 0.171713188290596,
+      "learning_rate": 8e-05,
+      "loss": 1.6085,
+      "step": 774
+    },
+    {
+      "epoch": 0.1057731677357718,
+      "grad_norm": 0.17383509874343872,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 775
+    },
+    {
+      "epoch": 0.10590964924252763,
+      "grad_norm": 0.1733373999595642,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 776
+    },
+    {
+      "epoch": 0.10604613074928347,
+      "grad_norm": 0.1787233203649521,
+      "learning_rate": 8e-05,
+      "loss": 1.6467,
+      "step": 777
+    },
+    {
+      "epoch": 0.10618261225603931,
+      "grad_norm": 0.17274242639541626,
+      "learning_rate": 8e-05,
+      "loss": 1.5419,
+      "step": 778
+    },
+    {
+      "epoch": 0.10631909376279514,
+      "grad_norm": 0.17590732872486115,
+      "learning_rate": 8e-05,
+      "loss": 1.6402,
+      "step": 779
+    },
+    {
+      "epoch": 0.10645557526955098,
+      "grad_norm": 0.1764359325170517,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 780
+    },
+    {
+      "epoch": 0.10659205677630681,
+      "grad_norm": 0.1733657717704773,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 781
+    },
+    {
+      "epoch": 0.10672853828306264,
+      "grad_norm": 0.183291494846344,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 782
+    },
+    {
+      "epoch": 0.10686501978981848,
+      "grad_norm": 0.1736065298318863,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 783
+    },
+    {
+      "epoch": 0.10700150129657432,
+      "grad_norm": 0.17475703358650208,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 784
+    },
+    {
+      "epoch": 0.10713798280333015,
+      "grad_norm": 0.17683954536914825,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 785
+    },
+    {
+      "epoch": 0.10727446431008598,
+      "grad_norm": 0.17981061339378357,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 786
+    },
+    {
+      "epoch": 0.10741094581684182,
+      "grad_norm": 0.17566846311092377,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 787
+    },
+    {
+      "epoch": 0.10754742732359765,
+      "grad_norm": 0.18363140523433685,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 788
+    },
+    {
+      "epoch": 0.10768390883035349,
+      "grad_norm": 0.18192534148693085,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 789
+    },
+    {
+      "epoch": 0.10782039033710931,
+      "grad_norm": 0.17498360574245453,
+      "learning_rate": 8e-05,
+      "loss": 1.579,
+      "step": 790
+    },
+    {
+      "epoch": 0.10795687184386515,
+      "grad_norm": 0.17999783158302307,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 791
+    },
+    {
+      "epoch": 0.108093353350621,
+      "grad_norm": 0.1754375696182251,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 792
+    },
+    {
+      "epoch": 0.10822983485737682,
+      "grad_norm": 0.17585621774196625,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 793
+    },
+    {
+      "epoch": 0.10836631636413266,
+      "grad_norm": 0.16878916323184967,
+      "learning_rate": 8e-05,
+      "loss": 1.5616,
+      "step": 794
+    },
+    {
+      "epoch": 0.1085027978708885,
+      "grad_norm": 0.18098954856395721,
+      "learning_rate": 8e-05,
+      "loss": 1.6234,
+      "step": 795
+    },
+    {
+      "epoch": 0.10863927937764432,
+      "grad_norm": 0.17510724067687988,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 796
+    },
+    {
+      "epoch": 0.10877576088440016,
+      "grad_norm": 0.17434807121753693,
+      "learning_rate": 8e-05,
+      "loss": 1.6585,
+      "step": 797
+    },
+    {
+      "epoch": 0.108912242391156,
+      "grad_norm": 0.1847684234380722,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 798
+    },
+    {
+      "epoch": 0.10904872389791183,
+      "grad_norm": 0.17530272901058197,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 799
+    },
+    {
+      "epoch": 0.10918520540466767,
+      "grad_norm": 0.18760628998279572,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 800
+    },
+    {
+      "epoch": 0.10932168691142351,
+      "grad_norm": 0.18097880482673645,
+      "learning_rate": 8e-05,
+      "loss": 1.6786,
+      "step": 801
+    },
+    {
+      "epoch": 0.10945816841817933,
+      "grad_norm": 0.18548883497714996,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 802
+    },
+    {
+      "epoch": 0.10959464992493517,
+      "grad_norm": 0.18543604016304016,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 803
+    },
+    {
+      "epoch": 0.10973113143169101,
+      "grad_norm": 0.17286135256290436,
+      "learning_rate": 8e-05,
+      "loss": 1.5581,
+      "step": 804
+    },
+    {
+      "epoch": 0.10986761293844684,
+      "grad_norm": 0.1733119934797287,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 805
+    },
+    {
+      "epoch": 0.11000409444520268,
+      "grad_norm": 0.17915892601013184,
+      "learning_rate": 8e-05,
+      "loss": 1.5881,
+      "step": 806
+    },
+    {
+      "epoch": 0.11014057595195852,
+      "grad_norm": 0.1865999549627304,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 807
+    },
+    {
+      "epoch": 0.11027705745871434,
+      "grad_norm": 0.18712003529071808,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 808
+    },
+    {
+      "epoch": 0.11041353896547018,
+      "grad_norm": 0.18837514519691467,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 809
+    },
+    {
+      "epoch": 0.11055002047222601,
+      "grad_norm": 0.19650830328464508,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 810
+    },
+    {
+      "epoch": 0.11068650197898185,
+      "grad_norm": 0.1765844076871872,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 811
+    },
+    {
+      "epoch": 0.11082298348573769,
+      "grad_norm": 0.19629192352294922,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 812
+    },
+    {
+      "epoch": 0.11095946499249351,
+      "grad_norm": 0.17397604882717133,
+      "learning_rate": 8e-05,
+      "loss": 1.6121,
+      "step": 813
+    },
+    {
+      "epoch": 0.11109594649924935,
+      "grad_norm": 0.18724505603313446,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 814
+    },
+    {
+      "epoch": 0.11123242800600519,
+      "grad_norm": 0.1819848269224167,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 815
+    },
+    {
+      "epoch": 0.11136890951276102,
+      "grad_norm": 0.17821615934371948,
+      "learning_rate": 8e-05,
+      "loss": 1.562,
+      "step": 816
+    },
+    {
+      "epoch": 0.11150539101951686,
+      "grad_norm": 0.1906854659318924,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 817
+    },
+    {
+      "epoch": 0.1116418725262727,
+      "grad_norm": 0.17341342568397522,
+      "learning_rate": 8e-05,
+      "loss": 1.613,
+      "step": 818
+    },
+    {
+      "epoch": 0.11177835403302852,
+      "grad_norm": 0.1838037222623825,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 819
+    },
+    {
+      "epoch": 0.11191483553978436,
+      "grad_norm": 0.17821507155895233,
+      "learning_rate": 8e-05,
+      "loss": 1.608,
+      "step": 820
+    },
+    {
+      "epoch": 0.1120513170465402,
+      "grad_norm": 0.17655165493488312,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 821
+    },
+    {
+      "epoch": 0.11218779855329603,
+      "grad_norm": 0.19318488240242004,
+      "learning_rate": 8e-05,
+      "loss": 1.7248,
+      "step": 822
+    },
+    {
+      "epoch": 0.11232428006005186,
+      "grad_norm": 0.19512103497982025,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 823
+    },
+    {
+      "epoch": 0.1124607615668077,
+      "grad_norm": 0.17627279460430145,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 824
+    },
+    {
+      "epoch": 0.11259724307356353,
+      "grad_norm": 0.17704002559185028,
+      "learning_rate": 8e-05,
+      "loss": 1.6296,
+      "step": 825
+    },
+    {
+      "epoch": 0.11273372458031937,
+      "grad_norm": 0.1744222790002823,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 826
+    },
+    {
+      "epoch": 0.1128702060870752,
+      "grad_norm": 0.18518079817295074,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 827
+    },
+    {
+      "epoch": 0.11300668759383103,
+      "grad_norm": 0.1840139776468277,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 828
+    },
+    {
+      "epoch": 0.11314316910058687,
+      "grad_norm": 0.18673016130924225,
+      "learning_rate": 8e-05,
+      "loss": 1.6941,
+      "step": 829
+    },
+    {
+      "epoch": 0.1132796506073427,
+      "grad_norm": 0.1809215247631073,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 830
+    },
+    {
+      "epoch": 0.11341613211409854,
+      "grad_norm": 0.17370539903640747,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 831
+    },
+    {
+      "epoch": 0.11355261362085438,
+      "grad_norm": 0.1796826869249344,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 832
+    },
+    {
+      "epoch": 0.1136890951276102,
+      "grad_norm": 0.18027593195438385,
+      "learning_rate": 8e-05,
+      "loss": 1.5867,
+      "step": 833
+    },
+    {
+      "epoch": 0.11382557663436604,
+      "grad_norm": 0.17990680038928986,
+      "learning_rate": 8e-05,
+      "loss": 1.6446,
+      "step": 834
+    },
+    {
+      "epoch": 0.11396205814112188,
+      "grad_norm": 0.1757950484752655,
+      "learning_rate": 8e-05,
+      "loss": 1.6375,
+      "step": 835
+    },
+    {
+      "epoch": 0.11409853964787771,
+      "grad_norm": 0.17497842013835907,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 836
+    },
+    {
+      "epoch": 0.11423502115463355,
+      "grad_norm": 0.17865315079689026,
+      "learning_rate": 8e-05,
+      "loss": 1.6132,
+      "step": 837
+    },
+    {
+      "epoch": 0.11437150266138939,
+      "grad_norm": 0.171797513961792,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 838
+    },
+    {
+      "epoch": 0.11450798416814521,
+      "grad_norm": 0.18378469347953796,
+      "learning_rate": 8e-05,
+      "loss": 1.6889,
+      "step": 839
+    },
+    {
+      "epoch": 0.11464446567490105,
+      "grad_norm": 0.17329350113868713,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 840
+    },
+    {
+      "epoch": 0.11478094718165689,
+      "grad_norm": 0.17543116211891174,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 841
+    },
+    {
+      "epoch": 0.11491742868841272,
+      "grad_norm": 0.17282266914844513,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 842
+    },
+    {
+      "epoch": 0.11505391019516856,
+      "grad_norm": 0.1784832775592804,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 843
+    },
+    {
+      "epoch": 0.11519039170192438,
+      "grad_norm": 0.17148315906524658,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 844
+    },
+    {
+      "epoch": 0.11532687320868022,
+      "grad_norm": 0.17613743245601654,
+      "learning_rate": 8e-05,
+      "loss": 1.627,
+      "step": 845
+    },
+    {
+      "epoch": 0.11546335471543606,
+      "grad_norm": 0.18533572554588318,
+      "learning_rate": 8e-05,
+      "loss": 1.7499,
+      "step": 846
+    },
+    {
+      "epoch": 0.11559983622219189,
+      "grad_norm": 0.1787489503622055,
+      "learning_rate": 8e-05,
+      "loss": 1.6186,
+      "step": 847
+    },
+    {
+      "epoch": 0.11573631772894773,
+      "grad_norm": 0.1767359972000122,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 848
+    },
+    {
+      "epoch": 0.11587279923570357,
+      "grad_norm": 0.1792907565832138,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 849
+    },
+    {
+      "epoch": 0.11600928074245939,
+      "grad_norm": 0.18009668588638306,
+      "learning_rate": 8e-05,
+      "loss": 1.5623,
+      "step": 850
+    },
+    {
+      "epoch": 0.11614576224921523,
+      "grad_norm": 0.18194961547851562,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 851
+    },
+    {
+      "epoch": 0.11628224375597107,
+      "grad_norm": 0.18576505780220032,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 852
+    },
+    {
+      "epoch": 0.1164187252627269,
+      "grad_norm": 0.18106262385845184,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 853
+    },
+    {
+      "epoch": 0.11655520676948274,
+      "grad_norm": 0.17341932654380798,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 854
+    },
+    {
+      "epoch": 0.11669168827623858,
+      "grad_norm": 0.183492511510849,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 855
+    },
+    {
+      "epoch": 0.1168281697829944,
+      "grad_norm": 0.17791801691055298,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 856
+    },
+    {
+      "epoch": 0.11696465128975024,
+      "grad_norm": 0.184437096118927,
+      "learning_rate": 8e-05,
+      "loss": 1.6706,
+      "step": 857
+    },
+    {
+      "epoch": 0.11710113279650608,
+      "grad_norm": 0.17684046924114227,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 858
+    },
+    {
+      "epoch": 0.1172376143032619,
+      "grad_norm": 0.1801089197397232,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 859
+    },
+    {
+      "epoch": 0.11737409581001775,
+      "grad_norm": 0.1937098205089569,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 860
+    },
+    {
+      "epoch": 0.11751057731677357,
+      "grad_norm": 0.18046562373638153,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 861
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 0.17740978300571442,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 862
+    },
+    {
+      "epoch": 0.11778354033028525,
+      "grad_norm": 0.1771164983510971,
+      "learning_rate": 8e-05,
+      "loss": 1.5904,
+      "step": 863
+    },
+    {
+      "epoch": 0.11792002183704108,
+      "grad_norm": 0.17637817561626434,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 864
+    },
+    {
+      "epoch": 0.11805650334379691,
+      "grad_norm": 0.17993707954883575,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 865
+    },
+    {
+      "epoch": 0.11819298485055275,
+      "grad_norm": 0.18234601616859436,
+      "learning_rate": 8e-05,
+      "loss": 1.6226,
+      "step": 866
+    },
+    {
+      "epoch": 0.11832946635730858,
+      "grad_norm": 0.18139328062534332,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 867
+    },
+    {
+      "epoch": 0.11846594786406442,
+      "grad_norm": 0.17290136218070984,
+      "learning_rate": 8e-05,
+      "loss": 1.5314,
+      "step": 868
+    },
+    {
+      "epoch": 0.11860242937082026,
+      "grad_norm": 0.1816997528076172,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 869
+    },
+    {
+      "epoch": 0.11873891087757608,
+      "grad_norm": 0.18639543652534485,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.11887539238433192,
+      "grad_norm": 0.17828713357448578,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 871
+    },
+    {
+      "epoch": 0.11901187389108776,
+      "grad_norm": 0.1788337528705597,
+      "learning_rate": 8e-05,
+      "loss": 1.6158,
+      "step": 872
+    },
+    {
+      "epoch": 0.11914835539784359,
+      "grad_norm": 0.17844010889530182,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 873
+    },
+    {
+      "epoch": 0.11928483690459943,
+      "grad_norm": 0.1788933128118515,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 874
+    },
+    {
+      "epoch": 0.11942131841135527,
+      "grad_norm": 0.18061116337776184,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 875
+    },
+    {
+      "epoch": 0.1195577999181111,
+      "grad_norm": 0.1775013953447342,
+      "learning_rate": 8e-05,
+      "loss": 1.5957,
+      "step": 876
+    },
+    {
+      "epoch": 0.11969428142486693,
+      "grad_norm": 0.17435961961746216,
+      "learning_rate": 8e-05,
+      "loss": 1.5344,
+      "step": 877
+    },
+    {
+      "epoch": 0.11983076293162276,
+      "grad_norm": 0.17335893213748932,
+      "learning_rate": 8e-05,
+      "loss": 1.471,
+      "step": 878
+    },
+    {
+      "epoch": 0.1199672444383786,
+      "grad_norm": 0.20797888934612274,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 879
+    },
+    {
+      "epoch": 0.12010372594513444,
+      "grad_norm": 0.18705587089061737,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 880
+    },
+    {
+      "epoch": 0.12024020745189026,
+      "grad_norm": 0.20102541148662567,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 881
+    },
+    {
+      "epoch": 0.1203766889586461,
+      "grad_norm": 0.18595711886882782,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 882
+    },
+    {
+      "epoch": 0.12051317046540194,
+      "grad_norm": 0.18247805535793304,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 883
+    },
+    {
+      "epoch": 0.12064965197215777,
+      "grad_norm": 0.18416698276996613,
+      "learning_rate": 8e-05,
+      "loss": 1.6469,
+      "step": 884
+    },
+    {
+      "epoch": 0.12078613347891361,
+      "grad_norm": 0.17978128790855408,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 885
+    },
+    {
+      "epoch": 0.12092261498566945,
+      "grad_norm": 0.18981236219406128,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 886
+    },
+    {
+      "epoch": 0.12105909649242527,
+      "grad_norm": 0.17937511205673218,
+      "learning_rate": 8e-05,
+      "loss": 1.5445,
+      "step": 887
+    },
+    {
+      "epoch": 0.12119557799918111,
+      "grad_norm": 0.18684163689613342,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 888
+    },
+    {
+      "epoch": 0.12133205950593695,
+      "grad_norm": 0.17635121941566467,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 889
+    },
+    {
+      "epoch": 0.12146854101269278,
+      "grad_norm": 0.1840089112520218,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 890
+    },
+    {
+      "epoch": 0.12160502251944862,
+      "grad_norm": 0.1795346438884735,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 891
+    },
+    {
+      "epoch": 0.12174150402620446,
+      "grad_norm": 0.176628977060318,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 892
+    },
+    {
+      "epoch": 0.12187798553296028,
+      "grad_norm": 0.1766405552625656,
+      "learning_rate": 8e-05,
+      "loss": 1.522,
+      "step": 893
+    },
+    {
+      "epoch": 0.12201446703971612,
+      "grad_norm": 0.1855059564113617,
+      "learning_rate": 8e-05,
+      "loss": 1.5793,
+      "step": 894
+    },
+    {
+      "epoch": 0.12215094854647195,
+      "grad_norm": 0.16692188382148743,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 895
+    },
+    {
+      "epoch": 0.12228743005322779,
+      "grad_norm": 0.18360888957977295,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 896
+    },
+    {
+      "epoch": 0.12242391155998363,
+      "grad_norm": 0.19752137362957,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 897
+    },
+    {
+      "epoch": 0.12256039306673945,
+      "grad_norm": 0.17349641025066376,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 898
+    },
+    {
+      "epoch": 0.12269687457349529,
+      "grad_norm": 0.1835305094718933,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 899
+    },
+    {
+      "epoch": 0.12283335608025113,
+      "grad_norm": 0.17954355478286743,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 900
+    },
+    {
+      "epoch": 0.12296983758700696,
+      "grad_norm": 0.17774495482444763,
+      "learning_rate": 8e-05,
+      "loss": 1.569,
+      "step": 901
+    },
+    {
+      "epoch": 0.1231063190937628,
+      "grad_norm": 0.18133053183555603,
+      "learning_rate": 8e-05,
+      "loss": 1.6259,
+      "step": 902
+    },
+    {
+      "epoch": 0.12324280060051863,
+      "grad_norm": 0.1818680465221405,
+      "learning_rate": 8e-05,
+      "loss": 1.6475,
+      "step": 903
+    },
+    {
+      "epoch": 0.12337928210727446,
+      "grad_norm": 0.17560385167598724,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 904
+    },
+    {
+      "epoch": 0.1235157636140303,
+      "grad_norm": 0.17599783837795258,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 905
+    },
+    {
+      "epoch": 0.12365224512078614,
+      "grad_norm": 0.18171408772468567,
+      "learning_rate": 8e-05,
+      "loss": 1.6056,
+      "step": 906
+    },
+    {
+      "epoch": 0.12378872662754196,
+      "grad_norm": 0.190260112285614,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 907
+    },
+    {
+      "epoch": 0.1239252081342978,
+      "grad_norm": 0.18249057233333588,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 908
+    },
+    {
+      "epoch": 0.12406168964105364,
+      "grad_norm": 0.1811327487230301,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 909
+    },
+    {
+      "epoch": 0.12419817114780947,
+      "grad_norm": 0.18575896322727203,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 910
+    },
+    {
+      "epoch": 0.12433465265456531,
+      "grad_norm": 0.20230241119861603,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 911
+    },
+    {
+      "epoch": 0.12447113416132113,
+      "grad_norm": 0.1710241734981537,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 912
+    },
+    {
+      "epoch": 0.12460761566807697,
+      "grad_norm": 0.21191109716892242,
+      "learning_rate": 8e-05,
+      "loss": 1.6232,
+      "step": 913
+    },
+    {
+      "epoch": 0.12474409717483281,
+      "grad_norm": 0.1764707863330841,
+      "learning_rate": 8e-05,
+      "loss": 1.5538,
+      "step": 914
+    },
+    {
+      "epoch": 0.12488057868158864,
+      "grad_norm": 0.1901218146085739,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 915
+    },
+    {
+      "epoch": 0.1250170601883445,
+      "grad_norm": 0.17722094058990479,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 916
+    },
+    {
+      "epoch": 0.12515354169510032,
+      "grad_norm": 0.1798778921365738,
+      "learning_rate": 8e-05,
+      "loss": 1.6525,
+      "step": 917
+    },
+    {
+      "epoch": 0.12529002320185614,
+      "grad_norm": 0.17167997360229492,
+      "learning_rate": 8e-05,
+      "loss": 1.5566,
+      "step": 918
+    },
+    {
+      "epoch": 0.125426504708612,
+      "grad_norm": 0.17248013615608215,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 919
+    },
+    {
+      "epoch": 0.12556298621536782,
+      "grad_norm": 0.194889098405838,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 920
+    },
+    {
+      "epoch": 0.12569946772212365,
+      "grad_norm": 0.17680509388446808,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 921
+    },
+    {
+      "epoch": 0.12583594922887947,
+      "grad_norm": 0.18244650959968567,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 922
+    },
+    {
+      "epoch": 0.12597243073563533,
+      "grad_norm": 0.17713268101215363,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 923
+    },
+    {
+      "epoch": 0.12610891224239115,
+      "grad_norm": 0.17911219596862793,
+      "learning_rate": 8e-05,
+      "loss": 1.6386,
+      "step": 924
+    },
+    {
+      "epoch": 0.12624539374914698,
+      "grad_norm": 0.17873425781726837,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 925
+    },
+    {
+      "epoch": 0.12638187525590283,
+      "grad_norm": 0.1766316145658493,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 926
+    },
+    {
+      "epoch": 0.12651835676265866,
+      "grad_norm": 0.17669962346553802,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 927
+    },
+    {
+      "epoch": 0.12665483826941448,
+      "grad_norm": 0.18439628183841705,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 928
+    },
+    {
+      "epoch": 0.12679131977617034,
+      "grad_norm": 0.1775929480791092,
+      "learning_rate": 8e-05,
+      "loss": 1.5436,
+      "step": 929
+    },
+    {
+      "epoch": 0.12692780128292616,
+      "grad_norm": 0.18298010528087616,
+      "learning_rate": 8e-05,
+      "loss": 1.5582,
+      "step": 930
+    },
+    {
+      "epoch": 0.127064282789682,
+      "grad_norm": 0.1873892992734909,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 931
+    },
+    {
+      "epoch": 0.12720076429643784,
+      "grad_norm": 0.17163720726966858,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 932
+    },
+    {
+      "epoch": 0.12733724580319367,
+      "grad_norm": 0.18349236249923706,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 933
+    },
+    {
+      "epoch": 0.1274737273099495,
+      "grad_norm": 0.1902233362197876,
+      "learning_rate": 8e-05,
+      "loss": 1.6826,
+      "step": 934
+    },
+    {
+      "epoch": 0.12761020881670534,
+      "grad_norm": 0.1762106865644455,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 935
+    },
+    {
+      "epoch": 0.12774669032346117,
+      "grad_norm": 0.18263155221939087,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 936
+    },
+    {
+      "epoch": 0.127883171830217,
+      "grad_norm": 0.1742410808801651,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 937
+    },
+    {
+      "epoch": 0.12801965333697285,
+      "grad_norm": 0.1760226935148239,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 938
+    },
+    {
+      "epoch": 0.12815613484372868,
+      "grad_norm": 0.17952662706375122,
+      "learning_rate": 8e-05,
+      "loss": 1.5442,
+      "step": 939
+    },
+    {
+      "epoch": 0.1282926163504845,
+      "grad_norm": 0.1722107231616974,
+      "learning_rate": 8e-05,
+      "loss": 1.6329,
+      "step": 940
+    },
+    {
+      "epoch": 0.12842909785724035,
+      "grad_norm": 0.19019272923469543,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 941
+    },
+    {
+      "epoch": 0.12856557936399618,
+      "grad_norm": 0.1751847267150879,
+      "learning_rate": 8e-05,
+      "loss": 1.6177,
+      "step": 942
+    },
+    {
+      "epoch": 0.128702060870752,
+      "grad_norm": 0.17907164990901947,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 943
+    },
+    {
+      "epoch": 0.12883854237750786,
+      "grad_norm": 0.18604299426078796,
+      "learning_rate": 8e-05,
+      "loss": 1.5701,
+      "step": 944
+    },
+    {
+      "epoch": 0.12897502388426368,
+      "grad_norm": 0.18538393080234528,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 945
+    },
+    {
+      "epoch": 0.1291115053910195,
+      "grad_norm": 0.19442118704319,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 946
+    },
+    {
+      "epoch": 0.12924798689777536,
+      "grad_norm": 0.17790429294109344,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 947
+    },
+    {
+      "epoch": 0.1293844684045312,
+      "grad_norm": 0.2003907859325409,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 948
+    },
+    {
+      "epoch": 0.12952094991128701,
+      "grad_norm": 0.17197178304195404,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 949
+    },
+    {
+      "epoch": 0.12965743141804287,
+      "grad_norm": 0.17345573008060455,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 950
+    },
+    {
+      "epoch": 0.1297939129247987,
+      "grad_norm": 0.18658950924873352,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 951
+    },
+    {
+      "epoch": 0.12993039443155452,
+      "grad_norm": 0.18130122125148773,
+      "learning_rate": 8e-05,
+      "loss": 1.6065,
+      "step": 952
+    },
+    {
+      "epoch": 0.13006687593831037,
+      "grad_norm": 0.19098636507987976,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 953
+    },
+    {
+      "epoch": 0.1302033574450662,
+      "grad_norm": 0.187925323843956,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 954
+    },
+    {
+      "epoch": 0.13033983895182202,
+      "grad_norm": 0.1866169571876526,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 955
+    },
+    {
+      "epoch": 0.13047632045857785,
+      "grad_norm": 0.18686704337596893,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 956
+    },
+    {
+      "epoch": 0.1306128019653337,
+      "grad_norm": 0.17064644396305084,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 957
+    },
+    {
+      "epoch": 0.13074928347208953,
+      "grad_norm": 0.199337437748909,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 958
+    },
+    {
+      "epoch": 0.13088576497884535,
+      "grad_norm": 0.17565563321113586,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 959
+    },
+    {
+      "epoch": 0.1310222464856012,
+      "grad_norm": 0.18454402685165405,
+      "learning_rate": 8e-05,
+      "loss": 1.5688,
+      "step": 960
+    },
+    {
+      "epoch": 0.13115872799235703,
+      "grad_norm": 0.17383253574371338,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 961
+    },
+    {
+      "epoch": 0.13129520949911286,
+      "grad_norm": 0.17539046704769135,
+      "learning_rate": 8e-05,
+      "loss": 1.533,
+      "step": 962
+    },
+    {
+      "epoch": 0.1314316910058687,
+      "grad_norm": 0.18956957757472992,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 963
+    },
+    {
+      "epoch": 0.13156817251262454,
+      "grad_norm": 0.1752173900604248,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 964
+    },
+    {
+      "epoch": 0.13170465401938036,
+      "grad_norm": 0.17612940073013306,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 965
+    },
+    {
+      "epoch": 0.13184113552613622,
+      "grad_norm": 0.1937987357378006,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 966
+    },
+    {
+      "epoch": 0.13197761703289204,
+      "grad_norm": 0.17771616578102112,
+      "learning_rate": 8e-05,
+      "loss": 1.606,
+      "step": 967
+    },
+    {
+      "epoch": 0.13211409853964787,
+      "grad_norm": 0.17480716109275818,
+      "learning_rate": 8e-05,
+      "loss": 1.4751,
+      "step": 968
+    },
+    {
+      "epoch": 0.13225058004640372,
+      "grad_norm": 0.17813922464847565,
+      "learning_rate": 8e-05,
+      "loss": 1.5524,
+      "step": 969
+    },
+    {
+      "epoch": 0.13238706155315955,
+      "grad_norm": 0.18162702023983002,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 970
+    },
+    {
+      "epoch": 0.13252354305991537,
+      "grad_norm": 0.18407806754112244,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 971
+    },
+    {
+      "epoch": 0.13266002456667123,
+      "grad_norm": 0.18449129164218903,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 972
+    },
+    {
+      "epoch": 0.13279650607342705,
+      "grad_norm": 0.20189370214939117,
+      "learning_rate": 8e-05,
+      "loss": 1.6208,
+      "step": 973
+    },
+    {
+      "epoch": 0.13293298758018288,
+      "grad_norm": 0.17681576311588287,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 974
+    },
+    {
+      "epoch": 0.13306946908693873,
+      "grad_norm": 0.1731061041355133,
+      "learning_rate": 8e-05,
+      "loss": 1.5359,
+      "step": 975
+    },
+    {
+      "epoch": 0.13320595059369456,
+      "grad_norm": 0.18673290312290192,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 976
+    },
+    {
+      "epoch": 0.13334243210045038,
+      "grad_norm": 0.18171891570091248,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 977
+    },
+    {
+      "epoch": 0.13347891360720623,
+      "grad_norm": 0.1831137239933014,
+      "learning_rate": 8e-05,
+      "loss": 1.608,
+      "step": 978
+    },
+    {
+      "epoch": 0.13361539511396206,
+      "grad_norm": 0.17454639077186584,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 979
+    },
+    {
+      "epoch": 0.13375187662071789,
+      "grad_norm": 0.18279410898685455,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 980
+    },
+    {
+      "epoch": 0.13388835812747374,
+      "grad_norm": 0.17709580063819885,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 981
+    },
+    {
+      "epoch": 0.13402483963422956,
+      "grad_norm": 0.1750289499759674,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 982
+    },
+    {
+      "epoch": 0.1341613211409854,
+      "grad_norm": 0.17871804535388947,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 983
+    },
+    {
+      "epoch": 0.13429780264774124,
+      "grad_norm": 0.17914311587810516,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 984
+    },
+    {
+      "epoch": 0.13443428415449707,
+      "grad_norm": 0.18388959765434265,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 985
+    },
+    {
+      "epoch": 0.1345707656612529,
+      "grad_norm": 0.18308115005493164,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 986
+    },
+    {
+      "epoch": 0.13470724716800875,
+      "grad_norm": 0.1827654242515564,
+      "learning_rate": 8e-05,
+      "loss": 1.661,
+      "step": 987
+    },
+    {
+      "epoch": 0.13484372867476457,
+      "grad_norm": 0.17717143893241882,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 988
+    },
+    {
+      "epoch": 0.1349802101815204,
+      "grad_norm": 0.1805904507637024,
+      "learning_rate": 8e-05,
+      "loss": 1.6351,
+      "step": 989
+    },
+    {
+      "epoch": 0.13511669168827622,
+      "grad_norm": 0.17695428431034088,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 990
+    },
+    {
+      "epoch": 0.13525317319503208,
+      "grad_norm": 0.17206792533397675,
+      "learning_rate": 8e-05,
+      "loss": 1.5984,
+      "step": 991
+    },
+    {
+      "epoch": 0.1353896547017879,
+      "grad_norm": 0.1792447865009308,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 992
+    },
+    {
+      "epoch": 0.13552613620854373,
+      "grad_norm": 0.18265284597873688,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 993
+    },
+    {
+      "epoch": 0.13566261771529958,
+      "grad_norm": 0.17521649599075317,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 994
+    },
+    {
+      "epoch": 0.1357990992220554,
+      "grad_norm": 0.1735079139471054,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 995
+    },
+    {
+      "epoch": 0.13593558072881123,
+      "grad_norm": 0.18370352685451508,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 996
+    },
+    {
+      "epoch": 0.1360720622355671,
+      "grad_norm": 0.18222841620445251,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 997
+    },
+    {
+      "epoch": 0.1362085437423229,
+      "grad_norm": 0.18838071823120117,
+      "learning_rate": 8e-05,
+      "loss": 1.6012,
+      "step": 998
+    },
+    {
+      "epoch": 0.13634502524907874,
+      "grad_norm": 0.17374905943870544,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 999
+    },
+    {
+      "epoch": 0.1364815067558346,
+      "grad_norm": 0.18086311221122742,
+      "learning_rate": 8e-05,
+      "loss": 1.5805,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13661798826259042,
+      "grad_norm": 0.19914259016513824,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13675446976934624,
+      "grad_norm": 0.18456502258777618,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1368909512761021,
+      "grad_norm": 0.1976194530725479,
+      "learning_rate": 8e-05,
+      "loss": 1.6784,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13702743278285792,
+      "grad_norm": 0.1925603747367859,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13716391428961375,
+      "grad_norm": 0.17244145274162292,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1373003957963696,
+      "grad_norm": 0.19100700318813324,
+      "learning_rate": 8e-05,
+      "loss": 1.5489,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13743687730312543,
+      "grad_norm": 0.17784613370895386,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13757335880988125,
+      "grad_norm": 0.17631614208221436,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1377098403166371,
+      "grad_norm": 0.18062086403369904,
+      "learning_rate": 8e-05,
+      "loss": 1.622,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13784632182339293,
+      "grad_norm": 0.16928671300411224,
+      "learning_rate": 8e-05,
+      "loss": 1.5581,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13798280333014876,
+      "grad_norm": 0.17381703853607178,
+      "learning_rate": 8e-05,
+      "loss": 1.5035,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1381192848369046,
+      "grad_norm": 0.17410334944725037,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13825576634366044,
+      "grad_norm": 0.1883874237537384,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13839224785041626,
+      "grad_norm": 0.1813386082649231,
+      "learning_rate": 8e-05,
+      "loss": 1.6014,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13852872935717211,
+      "grad_norm": 0.1805773824453354,
+      "learning_rate": 8e-05,
+      "loss": 1.6754,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13866521086392794,
+      "grad_norm": 0.17270179092884064,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13880169237068377,
+      "grad_norm": 0.1726042479276657,
+      "learning_rate": 8e-05,
+      "loss": 1.5009,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13893817387743962,
+      "grad_norm": 0.17762930691242218,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13907465538419544,
+      "grad_norm": 0.1860552728176117,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13921113689095127,
+      "grad_norm": 0.1832042634487152,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13934761839770712,
+      "grad_norm": 0.1861710250377655,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13948409990446295,
+      "grad_norm": 0.1767587512731552,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13962058141121877,
+      "grad_norm": 0.18185380101203918,
+      "learning_rate": 8e-05,
+      "loss": 1.6492,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13975706291797463,
+      "grad_norm": 0.1820162683725357,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13989354442473045,
+      "grad_norm": 0.17909790575504303,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 1025
+    },
+    {
+      "epoch": 0.14003002593148628,
+      "grad_norm": 0.1793469786643982,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1401665074382421,
+      "grad_norm": 0.18177662789821625,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 1027
+    },
+    {
+      "epoch": 0.14030298894499796,
+      "grad_norm": 0.17323514819145203,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 1028
+    },
+    {
+      "epoch": 0.14043947045175378,
+      "grad_norm": 0.18185874819755554,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1405759519585096,
+      "grad_norm": 0.1708773374557495,
+      "learning_rate": 8e-05,
+      "loss": 1.5021,
+      "step": 1030
+    },
+    {
+      "epoch": 0.14071243346526546,
+      "grad_norm": 0.18672265112400055,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1408489149720213,
+      "grad_norm": 0.18373003602027893,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 1032
+    },
+    {
+      "epoch": 0.14098539647877711,
+      "grad_norm": 0.18454372882843018,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 1033
+    },
+    {
+      "epoch": 0.14112187798553297,
+      "grad_norm": 0.1784476488828659,
+      "learning_rate": 8e-05,
+      "loss": 1.5652,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1412583594922888,
+      "grad_norm": 0.17644748091697693,
+      "learning_rate": 8e-05,
+      "loss": 1.5939,
+      "step": 1035
+    },
+    {
+      "epoch": 0.14139484099904462,
+      "grad_norm": 0.1777205467224121,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 1036
+    },
+    {
+      "epoch": 0.14153132250580047,
+      "grad_norm": 0.18302205204963684,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1416678040125563,
+      "grad_norm": 0.1734844446182251,
+      "learning_rate": 8e-05,
+      "loss": 1.584,
+      "step": 1038
+    },
+    {
+      "epoch": 0.14180428551931212,
+      "grad_norm": 0.18382154405117035,
+      "learning_rate": 8e-05,
+      "loss": 1.6426,
+      "step": 1039
+    },
+    {
+      "epoch": 0.14194076702606798,
+      "grad_norm": 0.18865056335926056,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1420772485328238,
+      "grad_norm": 0.17658798396587372,
+      "learning_rate": 8e-05,
+      "loss": 1.5401,
+      "step": 1041
+    },
+    {
+      "epoch": 0.14221373003957963,
+      "grad_norm": 0.19382302463054657,
+      "learning_rate": 8e-05,
+      "loss": 1.5658,
+      "step": 1042
+    },
+    {
+      "epoch": 0.14235021154633548,
+      "grad_norm": 0.18951554596424103,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1424866930530913,
+      "grad_norm": 0.17096170783042908,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 1044
+    },
+    {
+      "epoch": 0.14262317455984713,
+      "grad_norm": 0.18710759282112122,
+      "learning_rate": 8e-05,
+      "loss": 1.6514,
+      "step": 1045
+    },
+    {
+      "epoch": 0.14275965606660299,
+      "grad_norm": 0.17742778360843658,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 1046
+    },
+    {
+      "epoch": 0.1428961375733588,
+      "grad_norm": 0.17701713740825653,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 1047
+    },
+    {
+      "epoch": 0.14303261908011464,
+      "grad_norm": 0.17899398505687714,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1431691005868705,
+      "grad_norm": 0.17460735142230988,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1049
+    },
+    {
+      "epoch": 0.14330558209362632,
+      "grad_norm": 0.18164092302322388,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14344206360038214,
+      "grad_norm": 0.1788603514432907,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 1051
+    },
+    {
+      "epoch": 0.143578545107138,
+      "grad_norm": 0.18057799339294434,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 1052
+    },
+    {
+      "epoch": 0.14371502661389382,
+      "grad_norm": 0.17332379519939423,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 1053
+    },
+    {
+      "epoch": 0.14385150812064965,
+      "grad_norm": 0.18277396261692047,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1439879896274055,
+      "grad_norm": 0.1774376928806305,
+      "learning_rate": 8e-05,
+      "loss": 1.5576,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14412447113416132,
+      "grad_norm": 0.17968496680259705,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14426095264091715,
+      "grad_norm": 0.18069961667060852,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 1057
+    },
+    {
+      "epoch": 0.144397434147673,
+      "grad_norm": 0.17182733118534088,
+      "learning_rate": 8e-05,
+      "loss": 1.5373,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14453391565442883,
+      "grad_norm": 0.17266228795051575,
+      "learning_rate": 8e-05,
+      "loss": 1.6098,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14467039716118466,
+      "grad_norm": 0.17059507966041565,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14480687866794048,
+      "grad_norm": 0.17278394103050232,
+      "learning_rate": 8e-05,
+      "loss": 1.5623,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14494336017469633,
+      "grad_norm": 0.17384158074855804,
+      "learning_rate": 8e-05,
+      "loss": 1.5201,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14507984168145216,
+      "grad_norm": 0.1728159636259079,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14521632318820799,
+      "grad_norm": 0.1760879009962082,
+      "learning_rate": 8e-05,
+      "loss": 1.563,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14535280469496384,
+      "grad_norm": 0.17579421401023865,
+      "learning_rate": 8e-05,
+      "loss": 1.593,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14548928620171966,
+      "grad_norm": 0.17799431085586548,
+      "learning_rate": 8e-05,
+      "loss": 1.6455,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1456257677084755,
+      "grad_norm": 0.17833682894706726,
+      "learning_rate": 8e-05,
+      "loss": 1.5442,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14576224921523134,
+      "grad_norm": 0.1798112988471985,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14589873072198717,
+      "grad_norm": 0.1862846314907074,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 1069
+    },
+    {
+      "epoch": 0.146035212228743,
+      "grad_norm": 0.1751265525817871,
+      "learning_rate": 8e-05,
+      "loss": 1.5103,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14617169373549885,
+      "grad_norm": 0.17471760511398315,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14630817524225467,
+      "grad_norm": 0.18728645145893097,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1464446567490105,
+      "grad_norm": 0.18203704059123993,
+      "learning_rate": 8e-05,
+      "loss": 1.6135,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14658113825576635,
+      "grad_norm": 0.1761341243982315,
+      "learning_rate": 8e-05,
+      "loss": 1.6291,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14671761976252218,
+      "grad_norm": 0.17968878149986267,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 1075
+    },
+    {
+      "epoch": 0.146854101269278,
+      "grad_norm": 0.18649674952030182,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14699058277603386,
+      "grad_norm": 0.18363752961158752,
+      "learning_rate": 8e-05,
+      "loss": 1.5694,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14712706428278968,
+      "grad_norm": 0.17442214488983154,
+      "learning_rate": 8e-05,
+      "loss": 1.5561,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1472635457895455,
+      "grad_norm": 0.17706051468849182,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14740002729630136,
+      "grad_norm": 0.17640429735183716,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1475365088030572,
+      "grad_norm": 0.18479065597057343,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 1081
+    },
+    {
+      "epoch": 0.147672990309813,
+      "grad_norm": 0.17930950224399567,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14780947181656887,
+      "grad_norm": 0.18708758056163788,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1479459533233247,
+      "grad_norm": 0.17782355844974518,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14808243483008052,
+      "grad_norm": 0.19263522326946259,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14821891633683637,
+      "grad_norm": 0.18563781678676605,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1483553978435922,
+      "grad_norm": 0.18000678718090057,
+      "learning_rate": 8e-05,
+      "loss": 1.4983,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14849187935034802,
+      "grad_norm": 0.19443407654762268,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14862836085710388,
+      "grad_norm": 0.1867537498474121,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1487648423638597,
+      "grad_norm": 0.21880941092967987,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14890132387061553,
+      "grad_norm": 0.1861274689435959,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14903780537737138,
+      "grad_norm": 0.17828530073165894,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1491742868841272,
+      "grad_norm": 0.1942828744649887,
+      "learning_rate": 8e-05,
+      "loss": 1.6211,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14931076839088303,
+      "grad_norm": 0.1772737354040146,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14944724989763886,
+      "grad_norm": 0.1927632838487625,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1495837314043947,
+      "grad_norm": 0.1839390993118286,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14972021291115054,
+      "grad_norm": 0.1866489201784134,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14985669441790636,
+      "grad_norm": 0.19737054407596588,
+      "learning_rate": 8e-05,
+      "loss": 1.6586,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14999317592466221,
+      "grad_norm": 0.20163214206695557,
+      "learning_rate": 8e-05,
+      "loss": 1.6707,
+      "step": 1099
+    },
+    {
+      "epoch": 0.15012965743141804,
+      "grad_norm": 0.1980806142091751,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15026613893817387,
+      "grad_norm": 0.19414465129375458,
+      "learning_rate": 8e-05,
+      "loss": 1.5875,
+      "step": 1101
+    },
+    {
+      "epoch": 0.15040262044492972,
+      "grad_norm": 0.1710263043642044,
+      "learning_rate": 8e-05,
+      "loss": 1.5526,
+      "step": 1102
+    },
+    {
+      "epoch": 0.15053910195168554,
+      "grad_norm": 0.19318680465221405,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 1103
+    },
+    {
+      "epoch": 0.15067558345844137,
+      "grad_norm": 0.1879231184720993,
+      "learning_rate": 8e-05,
+      "loss": 1.5401,
+      "step": 1104
+    },
+    {
+      "epoch": 0.15081206496519722,
+      "grad_norm": 0.1749289333820343,
+      "learning_rate": 8e-05,
+      "loss": 1.4879,
+      "step": 1105
+    },
+    {
+      "epoch": 0.15094854647195305,
+      "grad_norm": 0.18455862998962402,
+      "learning_rate": 8e-05,
+      "loss": 1.5904,
+      "step": 1106
+    },
+    {
+      "epoch": 0.15108502797870887,
+      "grad_norm": 0.18272286653518677,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 1107
+    },
+    {
+      "epoch": 0.15122150948546473,
+      "grad_norm": 0.18324877321720123,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 1108
+    },
+    {
+      "epoch": 0.15135799099222055,
+      "grad_norm": 0.19138990342617035,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 1109
+    },
+    {
+      "epoch": 0.15149447249897638,
+      "grad_norm": 0.18146942555904388,
+      "learning_rate": 8e-05,
+      "loss": 1.6572,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15163095400573223,
+      "grad_norm": 0.18672722578048706,
+      "learning_rate": 8e-05,
+      "loss": 1.6006,
+      "step": 1111
+    },
+    {
+      "epoch": 0.15176743551248806,
+      "grad_norm": 0.17490947246551514,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 1112
+    },
+    {
+      "epoch": 0.15190391701924388,
+      "grad_norm": 0.18168552219867706,
+      "learning_rate": 8e-05,
+      "loss": 1.5323,
+      "step": 1113
+    },
+    {
+      "epoch": 0.15204039852599974,
+      "grad_norm": 0.18998941779136658,
+      "learning_rate": 8e-05,
+      "loss": 1.6606,
+      "step": 1114
+    },
+    {
+      "epoch": 0.15217688003275556,
+      "grad_norm": 0.17711807787418365,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1523133615395114,
+      "grad_norm": 0.19640128314495087,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 1116
+    },
+    {
+      "epoch": 0.15244984304626724,
+      "grad_norm": 0.18015168607234955,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 1117
+    },
+    {
+      "epoch": 0.15258632455302307,
+      "grad_norm": 0.19263866543769836,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 1118
+    },
+    {
+      "epoch": 0.1527228060597789,
+      "grad_norm": 0.183291494846344,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 1119
+    },
+    {
+      "epoch": 0.15285928756653475,
+      "grad_norm": 0.18031246960163116,
+      "learning_rate": 8e-05,
+      "loss": 1.5428,
+      "step": 1120
+    },
+    {
+      "epoch": 0.15299576907329057,
+      "grad_norm": 0.20082776248455048,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 1121
+    },
+    {
+      "epoch": 0.1531322505800464,
+      "grad_norm": 0.1699168086051941,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 1122
+    },
+    {
+      "epoch": 0.15326873208680225,
+      "grad_norm": 0.19546018540859222,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 1123
+    },
+    {
+      "epoch": 0.15340521359355808,
+      "grad_norm": 0.18226487934589386,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1535416951003139,
+      "grad_norm": 0.1791285276412964,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 1125
+    },
+    {
+      "epoch": 0.15367817660706976,
+      "grad_norm": 0.18853293359279633,
+      "learning_rate": 8e-05,
+      "loss": 1.5212,
+      "step": 1126
+    },
+    {
+      "epoch": 0.15381465811382558,
+      "grad_norm": 0.1931140422821045,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1539511396205814,
+      "grad_norm": 0.19364455342292786,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 1128
+    },
+    {
+      "epoch": 0.15408762112733723,
+      "grad_norm": 0.19626347720623016,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15422410263409309,
+      "grad_norm": 0.1855393946170807,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1543605841408489,
+      "grad_norm": 0.1744105964899063,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15449706564760474,
+      "grad_norm": 0.18494953215122223,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1546335471543606,
+      "grad_norm": 0.1748018115758896,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15477002866111642,
+      "grad_norm": 0.1800168752670288,
+      "learning_rate": 8e-05,
+      "loss": 1.5526,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15490651016787224,
+      "grad_norm": 0.18198733031749725,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1550429916746281,
+      "grad_norm": 0.1802837997674942,
+      "learning_rate": 8e-05,
+      "loss": 1.602,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15517947318138392,
+      "grad_norm": 0.1845933347940445,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15531595468813975,
+      "grad_norm": 0.19058282673358917,
+      "learning_rate": 8e-05,
+      "loss": 1.5956,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1554524361948956,
+      "grad_norm": 0.18110552430152893,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15558891770165142,
+      "grad_norm": 0.18208782374858856,
+      "learning_rate": 8e-05,
+      "loss": 1.5625,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15572539920840725,
+      "grad_norm": 0.19271939992904663,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1558618807151631,
+      "grad_norm": 0.18664176762104034,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15599836222191893,
+      "grad_norm": 0.195692241191864,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15613484372867475,
+      "grad_norm": 0.1937275528907776,
+      "learning_rate": 8e-05,
+      "loss": 1.5908,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1562713252354306,
+      "grad_norm": 0.16840621829032898,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15640780674218643,
+      "grad_norm": 0.1681431531906128,
+      "learning_rate": 8e-05,
+      "loss": 1.5064,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15654428824894226,
+      "grad_norm": 0.18363936245441437,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1566807697556981,
+      "grad_norm": 0.17105446755886078,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15681725126245394,
+      "grad_norm": 0.177464559674263,
+      "learning_rate": 8e-05,
+      "loss": 1.5689,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15695373276920976,
+      "grad_norm": 0.17929357290267944,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15709021427596562,
+      "grad_norm": 0.17444851994514465,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15722669578272144,
+      "grad_norm": 0.18968383967876434,
+      "learning_rate": 8e-05,
+      "loss": 1.653,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15736317728947727,
+      "grad_norm": 0.17355993390083313,
+      "learning_rate": 8e-05,
+      "loss": 1.4527,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15749965879623312,
+      "grad_norm": 0.17717935144901276,
+      "learning_rate": 8e-05,
+      "loss": 1.5335,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15763614030298895,
+      "grad_norm": 0.18725042045116425,
+      "learning_rate": 8e-05,
+      "loss": 1.5996,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15777262180974477,
+      "grad_norm": 0.17206521332263947,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15790910331650063,
+      "grad_norm": 0.17939825356006622,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15804558482325645,
+      "grad_norm": 0.1848825365304947,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15818206633001228,
+      "grad_norm": 0.17308993637561798,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15831854783676813,
+      "grad_norm": 0.17614364624023438,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15845502934352396,
+      "grad_norm": 0.18215034902095795,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15859151085027978,
+      "grad_norm": 0.19466570019721985,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15872799235703564,
+      "grad_norm": 0.17851008474826813,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15886447386379146,
+      "grad_norm": 0.19332507252693176,
+      "learning_rate": 8e-05,
+      "loss": 1.5907,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1590009553705473,
+      "grad_norm": 0.18777388334274292,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1591374368773031,
+      "grad_norm": 0.17552988231182098,
+      "learning_rate": 8e-05,
+      "loss": 1.5109,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15927391838405897,
+      "grad_norm": 0.2035273313522339,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1594103998908148,
+      "grad_norm": 0.18196377158164978,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15954688139757062,
+      "grad_norm": 0.19131863117218018,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15968336290432647,
+      "grad_norm": 0.20446893572807312,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1598198444110823,
+      "grad_norm": 0.17946723103523254,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15995632591783812,
+      "grad_norm": 0.1865733414888382,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 1172
+    },
+    {
+      "epoch": 0.16009280742459397,
+      "grad_norm": 0.21724280714988708,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1602292889313498,
+      "grad_norm": 0.1725117415189743,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 1174
+    },
+    {
+      "epoch": 0.16036577043810563,
+      "grad_norm": 0.20278382301330566,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 1175
+    },
+    {
+      "epoch": 0.16050225194486148,
+      "grad_norm": 0.2073919028043747,
+      "learning_rate": 8e-05,
+      "loss": 1.5558,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1606387334516173,
+      "grad_norm": 0.17578734457492828,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 1177
+    },
+    {
+      "epoch": 0.16077521495837313,
+      "grad_norm": 0.20177391171455383,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 1178
+    },
+    {
+      "epoch": 0.16091169646512898,
+      "grad_norm": 0.191424161195755,
+      "learning_rate": 8e-05,
+      "loss": 1.5903,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1610481779718848,
+      "grad_norm": 0.17382898926734924,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 1180
+    },
+    {
+      "epoch": 0.16118465947864064,
+      "grad_norm": 0.1912289559841156,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1613211409853965,
+      "grad_norm": 0.1886594593524933,
+      "learning_rate": 8e-05,
+      "loss": 1.5413,
+      "step": 1182
+    },
+    {
+      "epoch": 0.16145762249215231,
+      "grad_norm": 0.18339189887046814,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 1183
+    },
+    {
+      "epoch": 0.16159410399890814,
+      "grad_norm": 0.19257821142673492,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 1184
+    },
+    {
+      "epoch": 0.161730585505664,
+      "grad_norm": 0.1859150528907776,
+      "learning_rate": 8e-05,
+      "loss": 1.62,
+      "step": 1185
+    },
+    {
+      "epoch": 0.16186706701241982,
+      "grad_norm": 0.19414544105529785,
+      "learning_rate": 8e-05,
+      "loss": 1.5491,
+      "step": 1186
+    },
+    {
+      "epoch": 0.16200354851917564,
+      "grad_norm": 0.18056927621364594,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1621400300259315,
+      "grad_norm": 0.18426348268985748,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 1188
+    },
+    {
+      "epoch": 0.16227651153268732,
+      "grad_norm": 0.18608851730823517,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 1189
+    },
+    {
+      "epoch": 0.16241299303944315,
+      "grad_norm": 0.1818545013666153,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 1190
+    },
+    {
+      "epoch": 0.162549474546199,
+      "grad_norm": 0.18049830198287964,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 1191
+    },
+    {
+      "epoch": 0.16268595605295483,
+      "grad_norm": 0.17946931719779968,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 1192
+    },
+    {
+      "epoch": 0.16282243755971065,
+      "grad_norm": 0.1907290369272232,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1629589190664665,
+      "grad_norm": 0.18133649230003357,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 1194
+    },
+    {
+      "epoch": 0.16309540057322233,
+      "grad_norm": 0.19502057135105133,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 1195
+    },
+    {
+      "epoch": 0.16323188207997816,
+      "grad_norm": 0.18882198631763458,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 1196
+    },
+    {
+      "epoch": 0.163368363586734,
+      "grad_norm": 0.1806669384241104,
+      "learning_rate": 8e-05,
+      "loss": 1.6159,
+      "step": 1197
+    },
+    {
+      "epoch": 0.16350484509348984,
+      "grad_norm": 0.18993724882602692,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 1198
+    },
+    {
+      "epoch": 0.16364132660024566,
+      "grad_norm": 0.18915489315986633,
+      "learning_rate": 8e-05,
+      "loss": 1.5938,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1637778081070015,
+      "grad_norm": 0.17887987196445465,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16391428961375734,
+      "grad_norm": 0.22455234825611115,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 1201
+    },
+    {
+      "epoch": 0.16405077112051317,
+      "grad_norm": 0.18593254685401917,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 1202
+    },
+    {
+      "epoch": 0.164187252627269,
+      "grad_norm": 0.18921193480491638,
+      "learning_rate": 8e-05,
+      "loss": 1.5405,
+      "step": 1203
+    },
+    {
+      "epoch": 0.16432373413402485,
+      "grad_norm": 0.21956679224967957,
+      "learning_rate": 8e-05,
+      "loss": 1.6019,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16446021564078067,
+      "grad_norm": 0.18349327147006989,
+      "learning_rate": 8e-05,
+      "loss": 1.6218,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1645966971475365,
+      "grad_norm": 0.19205377995967865,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 1206
+    },
+    {
+      "epoch": 0.16473317865429235,
+      "grad_norm": 0.1899394392967224,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16486966016104818,
+      "grad_norm": 0.18048936128616333,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 1208
+    },
+    {
+      "epoch": 0.165006141667804,
+      "grad_norm": 0.1869986206293106,
+      "learning_rate": 8e-05,
+      "loss": 1.6498,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16514262317455985,
+      "grad_norm": 0.18800325691699982,
+      "learning_rate": 8e-05,
+      "loss": 1.5452,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16527910468131568,
+      "grad_norm": 0.1765373796224594,
+      "learning_rate": 8e-05,
+      "loss": 1.6174,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1654155861880715,
+      "grad_norm": 0.17699959874153137,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16555206769482736,
+      "grad_norm": 0.18276195228099823,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16568854920158319,
+      "grad_norm": 0.18691441416740417,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 1214
+    },
+    {
+      "epoch": 0.165825030708339,
+      "grad_norm": 0.17904722690582275,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16596151221509486,
+      "grad_norm": 0.17879913747310638,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1660979937218507,
+      "grad_norm": 0.17948831617832184,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16623447522860652,
+      "grad_norm": 0.18168850243091583,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16637095673536237,
+      "grad_norm": 0.17248106002807617,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1665074382421182,
+      "grad_norm": 0.18305379152297974,
+      "learning_rate": 8e-05,
+      "loss": 1.525,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16664391974887402,
+      "grad_norm": 0.1820470541715622,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16678040125562987,
+      "grad_norm": 0.18057158589363098,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1669168827623857,
+      "grad_norm": 0.190862238407135,
+      "learning_rate": 8e-05,
+      "loss": 1.5501,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16705336426914152,
+      "grad_norm": 0.18769824504852295,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16718984577589738,
+      "grad_norm": 0.18200239539146423,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1673263272826532,
+      "grad_norm": 0.18543142080307007,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16746280878940903,
+      "grad_norm": 0.1815391331911087,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16759929029616488,
+      "grad_norm": 0.17922349274158478,
+      "learning_rate": 8e-05,
+      "loss": 1.5171,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1677357718029207,
+      "grad_norm": 0.17526231706142426,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16787225330967653,
+      "grad_norm": 0.17742040753364563,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1680087348164324,
+      "grad_norm": 0.1920635998249054,
+      "learning_rate": 8e-05,
+      "loss": 1.602,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1681452163231882,
+      "grad_norm": 0.17591023445129395,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16828169782994404,
+      "grad_norm": 0.18112918734550476,
+      "learning_rate": 8e-05,
+      "loss": 1.5914,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16841817933669986,
+      "grad_norm": 0.17961393296718597,
+      "learning_rate": 8e-05,
+      "loss": 1.5291,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16855466084345572,
+      "grad_norm": 0.17470215260982513,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16869114235021154,
+      "grad_norm": 0.1790974885225296,
+      "learning_rate": 8e-05,
+      "loss": 1.5562,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16882762385696737,
+      "grad_norm": 0.180716410279274,
+      "learning_rate": 8e-05,
+      "loss": 1.5557,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16896410536372322,
+      "grad_norm": 0.1740095019340515,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16910058687047905,
+      "grad_norm": 0.17068733274936676,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16923706837723487,
+      "grad_norm": 0.19151034951210022,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16937354988399073,
+      "grad_norm": 0.17609147727489471,
+      "learning_rate": 8e-05,
+      "loss": 1.529,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16951003139074655,
+      "grad_norm": 0.17521561682224274,
+      "learning_rate": 8e-05,
+      "loss": 1.5431,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16964651289750238,
+      "grad_norm": 0.18549153208732605,
+      "learning_rate": 8e-05,
+      "loss": 1.5748,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16978299440425823,
+      "grad_norm": 0.1770140379667282,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16991947591101406,
+      "grad_norm": 0.1812058389186859,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 1245
+    },
+    {
+      "epoch": 0.17005595741776988,
+      "grad_norm": 0.18413046002388,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 1246
+    },
+    {
+      "epoch": 0.17019243892452574,
+      "grad_norm": 0.18512527644634247,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 1247
+    },
+    {
+      "epoch": 0.17032892043128156,
+      "grad_norm": 0.18638159334659576,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1704654019380374,
+      "grad_norm": 0.1843906044960022,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 1249
+    },
+    {
+      "epoch": 0.17060188344479324,
+      "grad_norm": 0.1919764131307602,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17073836495154907,
+      "grad_norm": 0.18232044577598572,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1708748464583049,
+      "grad_norm": 0.1758478581905365,
+      "learning_rate": 8e-05,
+      "loss": 1.6234,
+      "step": 1252
+    },
+    {
+      "epoch": 0.17101132796506074,
+      "grad_norm": 0.18299835920333862,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 1253
+    },
+    {
+      "epoch": 0.17114780947181657,
+      "grad_norm": 0.17757123708724976,
+      "learning_rate": 8e-05,
+      "loss": 1.5759,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1712842909785724,
+      "grad_norm": 0.18284830451011658,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 1255
+    },
+    {
+      "epoch": 0.17142077248532825,
+      "grad_norm": 0.17563997209072113,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 1256
+    },
+    {
+      "epoch": 0.17155725399208407,
+      "grad_norm": 0.1784861832857132,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1716937354988399,
+      "grad_norm": 0.17500747740268707,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 1258
+    },
+    {
+      "epoch": 0.17183021700559575,
+      "grad_norm": 0.17861445248126984,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17196669851235158,
+      "grad_norm": 0.18527667224407196,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1721031800191074,
+      "grad_norm": 0.1847296804189682,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17223966152586326,
+      "grad_norm": 0.18138012290000916,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 1262
+    },
+    {
+      "epoch": 0.17237614303261908,
+      "grad_norm": 0.1790272295475006,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1725126245393749,
+      "grad_norm": 0.18722257018089294,
+      "learning_rate": 8e-05,
+      "loss": 1.599,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17264910604613076,
+      "grad_norm": 0.17761902511119843,
+      "learning_rate": 8e-05,
+      "loss": 1.5586,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1727855875528866,
+      "grad_norm": 0.1800829917192459,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1729220690596424,
+      "grad_norm": 0.18426303565502167,
+      "learning_rate": 8e-05,
+      "loss": 1.6557,
+      "step": 1267
+    },
+    {
+      "epoch": 0.17305855056639824,
+      "grad_norm": 0.18017345666885376,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1731950320731541,
+      "grad_norm": 0.1756056249141693,
+      "learning_rate": 8e-05,
+      "loss": 1.6008,
+      "step": 1269
+    },
+    {
+      "epoch": 0.17333151357990992,
+      "grad_norm": 0.1743742674589157,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 1270
+    },
+    {
+      "epoch": 0.17346799508666574,
+      "grad_norm": 0.17890498042106628,
+      "learning_rate": 8e-05,
+      "loss": 1.5899,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1736044765934216,
+      "grad_norm": 0.18561498820781708,
+      "learning_rate": 8e-05,
+      "loss": 1.5761,
+      "step": 1272
+    },
+    {
+      "epoch": 0.17374095810017742,
+      "grad_norm": 0.1821655035018921,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17387743960693325,
+      "grad_norm": 0.1756887286901474,
+      "learning_rate": 8e-05,
+      "loss": 1.5573,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1740139211136891,
+      "grad_norm": 0.17907758057117462,
+      "learning_rate": 8e-05,
+      "loss": 1.609,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17415040262044493,
+      "grad_norm": 0.17647773027420044,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 1276
+    },
+    {
+      "epoch": 0.17428688412720075,
+      "grad_norm": 0.18217086791992188,
+      "learning_rate": 8e-05,
+      "loss": 1.6304,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1744233656339566,
+      "grad_norm": 0.1770544797182083,
+      "learning_rate": 8e-05,
+      "loss": 1.5869,
+      "step": 1278
+    },
+    {
+      "epoch": 0.17455984714071243,
+      "grad_norm": 0.1738310307264328,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17469632864746826,
+      "grad_norm": 0.1832553595304489,
+      "learning_rate": 8e-05,
+      "loss": 1.628,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1748328101542241,
+      "grad_norm": 0.18110144138336182,
+      "learning_rate": 8e-05,
+      "loss": 1.5282,
+      "step": 1281
+    },
+    {
+      "epoch": 0.17496929166097994,
+      "grad_norm": 0.191071555018425,
+      "learning_rate": 8e-05,
+      "loss": 1.5891,
+      "step": 1282
+    },
+    {
+      "epoch": 0.17510577316773576,
+      "grad_norm": 0.18352948129177094,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17524225467449162,
+      "grad_norm": 0.1845688819885254,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17537873618124744,
+      "grad_norm": 0.17512784898281097,
+      "learning_rate": 8e-05,
+      "loss": 1.5424,
+      "step": 1285
+    },
+    {
+      "epoch": 0.17551521768800327,
+      "grad_norm": 0.18274016678333282,
+      "learning_rate": 8e-05,
+      "loss": 1.5847,
+      "step": 1286
+    },
+    {
+      "epoch": 0.17565169919475912,
+      "grad_norm": 0.18950589001178741,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 1287
+    },
+    {
+      "epoch": 0.17578818070151495,
+      "grad_norm": 0.1836417019367218,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17592466220827077,
+      "grad_norm": 0.18523946404457092,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17606114371502662,
+      "grad_norm": 0.17753663659095764,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17619762522178245,
+      "grad_norm": 0.18328975141048431,
+      "learning_rate": 8e-05,
+      "loss": 1.563,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17633410672853828,
+      "grad_norm": 0.1872308999300003,
+      "learning_rate": 8e-05,
+      "loss": 1.5452,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 0.1936081498861313,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17660706974204995,
+      "grad_norm": 0.18803086876869202,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17674355124880578,
+      "grad_norm": 0.1898953914642334,
+      "learning_rate": 8e-05,
+      "loss": 1.5856,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17688003275556163,
+      "grad_norm": 0.19058725237846375,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17701651426231746,
+      "grad_norm": 0.17649631202220917,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17715299576907328,
+      "grad_norm": 0.18933144211769104,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17728947727582914,
+      "grad_norm": 0.19089755415916443,
+      "learning_rate": 8e-05,
+      "loss": 1.5853,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17742595878258496,
+      "grad_norm": 0.17789152264595032,
+      "learning_rate": 8e-05,
+      "loss": 1.5872,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1775624402893408,
+      "grad_norm": 0.18564368784427643,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17769892179609664,
+      "grad_norm": 0.17861820757389069,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17783540330285247,
+      "grad_norm": 0.17816072702407837,
+      "learning_rate": 8e-05,
+      "loss": 1.5364,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1779718848096083,
+      "grad_norm": 0.181040957570076,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17810836631636412,
+      "grad_norm": 0.19521668553352356,
+      "learning_rate": 8e-05,
+      "loss": 1.5515,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17824484782311997,
+      "grad_norm": 0.17776930332183838,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1783813293298758,
+      "grad_norm": 0.18215477466583252,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17851781083663162,
+      "grad_norm": 0.19597412645816803,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17865429234338748,
+      "grad_norm": 0.17734640836715698,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1787907738501433,
+      "grad_norm": 0.18767887353897095,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17892725535689913,
+      "grad_norm": 0.18095117807388306,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17906373686365498,
+      "grad_norm": 0.1826968640089035,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1792002183704108,
+      "grad_norm": 0.1729998141527176,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17933669987716663,
+      "grad_norm": 0.1798374056816101,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1794731813839225,
+      "grad_norm": 0.18414777517318726,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1796096628906783,
+      "grad_norm": 0.17512325942516327,
+      "learning_rate": 8e-05,
+      "loss": 1.5287,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17974614439743414,
+      "grad_norm": 0.17899727821350098,
+      "learning_rate": 8e-05,
+      "loss": 1.5239,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17988262590419,
+      "grad_norm": 0.17925399541854858,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 1318
+    },
+    {
+      "epoch": 0.18001910741094582,
+      "grad_norm": 0.17679765820503235,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 1319
+    },
+    {
+      "epoch": 0.18015558891770164,
+      "grad_norm": 0.17553366720676422,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1802920704244575,
+      "grad_norm": 0.18355882167816162,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 1321
+    },
+    {
+      "epoch": 0.18042855193121332,
+      "grad_norm": 0.18068662285804749,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 1322
+    },
+    {
+      "epoch": 0.18056503343796915,
+      "grad_norm": 0.18069930374622345,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 1323
+    },
+    {
+      "epoch": 0.180701514944725,
+      "grad_norm": 0.17888805270195007,
+      "learning_rate": 8e-05,
+      "loss": 1.5438,
+      "step": 1324
+    },
+    {
+      "epoch": 0.18083799645148083,
+      "grad_norm": 0.17886215448379517,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 1325
+    },
+    {
+      "epoch": 0.18097447795823665,
+      "grad_norm": 0.18087361752986908,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1811109594649925,
+      "grad_norm": 0.17912723124027252,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 1327
+    },
+    {
+      "epoch": 0.18124744097174833,
+      "grad_norm": 0.1830485761165619,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 1328
+    },
+    {
+      "epoch": 0.18138392247850416,
+      "grad_norm": 0.1840202957391739,
+      "learning_rate": 8e-05,
+      "loss": 1.5615,
+      "step": 1329
+    },
+    {
+      "epoch": 0.18152040398526,
+      "grad_norm": 0.17745667695999146,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 1330
+    },
+    {
+      "epoch": 0.18165688549201583,
+      "grad_norm": 0.17821282148361206,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 1331
+    },
+    {
+      "epoch": 0.18179336699877166,
+      "grad_norm": 0.1944548338651657,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1819298485055275,
+      "grad_norm": 0.18436874449253082,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 1333
+    },
+    {
+      "epoch": 0.18206633001228334,
+      "grad_norm": 0.17645595967769623,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 1334
+    },
+    {
+      "epoch": 0.18220281151903917,
+      "grad_norm": 0.18822963535785675,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 1335
+    },
+    {
+      "epoch": 0.18233929302579502,
+      "grad_norm": 0.1763017475605011,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1336
+    },
+    {
+      "epoch": 0.18247577453255084,
+      "grad_norm": 0.1805952787399292,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 1337
+    },
+    {
+      "epoch": 0.18261225603930667,
+      "grad_norm": 0.17240872979164124,
+      "learning_rate": 8e-05,
+      "loss": 1.5551,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1827487375460625,
+      "grad_norm": 0.17618606984615326,
+      "learning_rate": 8e-05,
+      "loss": 1.5382,
+      "step": 1339
+    },
+    {
+      "epoch": 0.18288521905281835,
+      "grad_norm": 0.17841020226478577,
+      "learning_rate": 8e-05,
+      "loss": 1.6514,
+      "step": 1340
+    },
+    {
+      "epoch": 0.18302170055957417,
+      "grad_norm": 0.17210696637630463,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1341
+    },
+    {
+      "epoch": 0.18315818206633,
+      "grad_norm": 0.1769982874393463,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 1342
+    },
+    {
+      "epoch": 0.18329466357308585,
+      "grad_norm": 0.18421314656734467,
+      "learning_rate": 8e-05,
+      "loss": 1.6176,
+      "step": 1343
+    },
+    {
+      "epoch": 0.18343114507984168,
+      "grad_norm": 0.1886877715587616,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1835676265865975,
+      "grad_norm": 0.17790721356868744,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 1345
+    },
+    {
+      "epoch": 0.18370410809335336,
+      "grad_norm": 0.18201424181461334,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 1346
+    },
+    {
+      "epoch": 0.18384058960010918,
+      "grad_norm": 0.1772538274526596,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1347
+    },
+    {
+      "epoch": 0.183977071106865,
+      "grad_norm": 0.18448364734649658,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 1348
+    },
+    {
+      "epoch": 0.18411355261362086,
+      "grad_norm": 0.17663274705410004,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1842500341203767,
+      "grad_norm": 0.18552815914154053,
+      "learning_rate": 8e-05,
+      "loss": 1.6226,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1843865156271325,
+      "grad_norm": 0.17485670745372772,
+      "learning_rate": 8e-05,
+      "loss": 1.5477,
+      "step": 1351
+    },
+    {
+      "epoch": 0.18452299713388837,
+      "grad_norm": 0.17874416708946228,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1846594786406442,
+      "grad_norm": 0.19542373716831207,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 1353
+    },
+    {
+      "epoch": 0.18479596014740002,
+      "grad_norm": 0.18012094497680664,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18493244165415587,
+      "grad_norm": 0.18003003299236298,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1850689231609117,
+      "grad_norm": 0.19407962262630463,
+      "learning_rate": 8e-05,
+      "loss": 1.6061,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18520540466766752,
+      "grad_norm": 0.1789020597934723,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 1357
+    },
+    {
+      "epoch": 0.18534188617442338,
+      "grad_norm": 0.20204922556877136,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1854783676811792,
+      "grad_norm": 0.1791933923959732,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 1359
+    },
+    {
+      "epoch": 0.18561484918793503,
+      "grad_norm": 0.19147218763828278,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 1360
+    },
+    {
+      "epoch": 0.18575133069469088,
+      "grad_norm": 0.18052275478839874,
+      "learning_rate": 8e-05,
+      "loss": 1.5237,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1858878122014467,
+      "grad_norm": 0.17625939846038818,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 1362
+    },
+    {
+      "epoch": 0.18602429370820253,
+      "grad_norm": 0.17764948308467865,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 1363
+    },
+    {
+      "epoch": 0.18616077521495838,
+      "grad_norm": 0.18250398337841034,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1862972567217142,
+      "grad_norm": 0.18556956946849823,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18643373822847004,
+      "grad_norm": 0.18151864409446716,
+      "learning_rate": 8e-05,
+      "loss": 1.5757,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1865702197352259,
+      "grad_norm": 0.17581619322299957,
+      "learning_rate": 8e-05,
+      "loss": 1.5534,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18670670124198172,
+      "grad_norm": 0.17195409536361694,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18684318274873754,
+      "grad_norm": 0.18987657129764557,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1869796642554934,
+      "grad_norm": 0.18373243510723114,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18711614576224922,
+      "grad_norm": 0.18615441024303436,
+      "learning_rate": 8e-05,
+      "loss": 1.594,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18725262726900505,
+      "grad_norm": 0.1862824410200119,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18738910877576087,
+      "grad_norm": 0.18179596960544586,
+      "learning_rate": 8e-05,
+      "loss": 1.5886,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18752559028251672,
+      "grad_norm": 0.17452482879161835,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18766207178927255,
+      "grad_norm": 0.18409357964992523,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18779855329602838,
+      "grad_norm": 0.17680881917476654,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18793503480278423,
+      "grad_norm": 0.1769135594367981,
+      "learning_rate": 8e-05,
+      "loss": 1.5392,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18807151630954005,
+      "grad_norm": 0.18122313916683197,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18820799781629588,
+      "grad_norm": 0.18079566955566406,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18834447932305173,
+      "grad_norm": 0.1818685680627823,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18848096082980756,
+      "grad_norm": 0.19014553725719452,
+      "learning_rate": 8e-05,
+      "loss": 1.5558,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18861744233656338,
+      "grad_norm": 0.17578458786010742,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18875392384331924,
+      "grad_norm": 0.18871986865997314,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18889040535007506,
+      "grad_norm": 0.18032406270503998,
+      "learning_rate": 8e-05,
+      "loss": 1.5184,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1890268868568309,
+      "grad_norm": 0.18657973408699036,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18916336836358674,
+      "grad_norm": 0.18356724083423615,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18929984987034257,
+      "grad_norm": 0.17990297079086304,
+      "learning_rate": 8e-05,
+      "loss": 1.5812,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1894363313770984,
+      "grad_norm": 0.17911188304424286,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18957281288385425,
+      "grad_norm": 0.1806691288948059,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18970929439061007,
+      "grad_norm": 0.18714362382888794,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1898457758973659,
+      "grad_norm": 0.19810731709003448,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18998225740412175,
+      "grad_norm": 0.1779014766216278,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 1392
+    },
+    {
+      "epoch": 0.19011873891087758,
+      "grad_norm": 0.19377335906028748,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1902552204176334,
+      "grad_norm": 0.17974470555782318,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 1394
+    },
+    {
+      "epoch": 0.19039170192438926,
+      "grad_norm": 0.19470670819282532,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 1395
+    },
+    {
+      "epoch": 0.19052818343114508,
+      "grad_norm": 0.1894291341304779,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1906646649379009,
+      "grad_norm": 0.17871151864528656,
+      "learning_rate": 8e-05,
+      "loss": 1.5975,
+      "step": 1397
+    },
+    {
+      "epoch": 0.19080114644465676,
+      "grad_norm": 0.19233465194702148,
+      "learning_rate": 8e-05,
+      "loss": 1.5793,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1909376279514126,
+      "grad_norm": 0.18489322066307068,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1910741094581684,
+      "grad_norm": 0.17781518399715424,
+      "learning_rate": 8e-05,
+      "loss": 1.4896,
+      "step": 1400
+    },
+    {
+      "epoch": 0.19121059096492427,
+      "grad_norm": 0.18936896324157715,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1913470724716801,
+      "grad_norm": 0.19243223965168,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 1402
+    },
+    {
+      "epoch": 0.19148355397843592,
+      "grad_norm": 0.18140558898448944,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 1403
+    },
+    {
+      "epoch": 0.19162003548519177,
+      "grad_norm": 0.1888154149055481,
+      "learning_rate": 8e-05,
+      "loss": 1.5025,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1917565169919476,
+      "grad_norm": 0.17819145321846008,
+      "learning_rate": 8e-05,
+      "loss": 1.546,
+      "step": 1405
+    },
+    {
+      "epoch": 0.19189299849870342,
+      "grad_norm": 0.1893412172794342,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 1406
+    },
+    {
+      "epoch": 0.19202948000545925,
+      "grad_norm": 0.18471944332122803,
+      "learning_rate": 8e-05,
+      "loss": 1.5434,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1921659615122151,
+      "grad_norm": 0.185714453458786,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 1408
+    },
+    {
+      "epoch": 0.19230244301897093,
+      "grad_norm": 0.18834368884563446,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 1409
+    },
+    {
+      "epoch": 0.19243892452572675,
+      "grad_norm": 0.18776118755340576,
+      "learning_rate": 8e-05,
+      "loss": 1.5496,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1925754060324826,
+      "grad_norm": 0.18822896480560303,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 1411
+    },
+    {
+      "epoch": 0.19271188753923843,
+      "grad_norm": 0.1927414983510971,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 1412
+    },
+    {
+      "epoch": 0.19284836904599426,
+      "grad_norm": 0.17877045273780823,
+      "learning_rate": 8e-05,
+      "loss": 1.4872,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1929848505527501,
+      "grad_norm": 0.17581462860107422,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 1414
+    },
+    {
+      "epoch": 0.19312133205950593,
+      "grad_norm": 0.18094876408576965,
+      "learning_rate": 8e-05,
+      "loss": 1.58,
+      "step": 1415
+    },
+    {
+      "epoch": 0.19325781356626176,
+      "grad_norm": 0.1943579465150833,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1933942950730176,
+      "grad_norm": 0.1700141727924347,
+      "learning_rate": 8e-05,
+      "loss": 1.452,
+      "step": 1417
+    },
+    {
+      "epoch": 0.19353077657977344,
+      "grad_norm": 0.18419000506401062,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 1418
+    },
+    {
+      "epoch": 0.19366725808652926,
+      "grad_norm": 0.17995068430900574,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 1419
+    },
+    {
+      "epoch": 0.19380373959328512,
+      "grad_norm": 0.18322820961475372,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 1420
+    },
+    {
+      "epoch": 0.19394022110004094,
+      "grad_norm": 0.18575672805309296,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1421
+    },
+    {
+      "epoch": 0.19407670260679677,
+      "grad_norm": 0.17758823931217194,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1422
+    },
+    {
+      "epoch": 0.19421318411355262,
+      "grad_norm": 0.17651227116584778,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 1423
+    },
+    {
+      "epoch": 0.19434966562030845,
+      "grad_norm": 0.18368026614189148,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 1424
+    },
+    {
+      "epoch": 0.19448614712706427,
+      "grad_norm": 0.18148313462734222,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 1425
+    },
+    {
+      "epoch": 0.19462262863382013,
+      "grad_norm": 0.18397633731365204,
+      "learning_rate": 8e-05,
+      "loss": 1.5528,
+      "step": 1426
+    },
+    {
+      "epoch": 0.19475911014057595,
+      "grad_norm": 0.18574011325836182,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19489559164733178,
+      "grad_norm": 0.18528436124324799,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 1428
+    },
+    {
+      "epoch": 0.19503207315408763,
+      "grad_norm": 0.18001489341259003,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 1429
+    },
+    {
+      "epoch": 0.19516855466084346,
+      "grad_norm": 0.1796899288892746,
+      "learning_rate": 8e-05,
+      "loss": 1.5733,
+      "step": 1430
+    },
+    {
+      "epoch": 0.19530503616759928,
+      "grad_norm": 0.18439553678035736,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 1431
+    },
+    {
+      "epoch": 0.19544151767435514,
+      "grad_norm": 0.18113921582698822,
+      "learning_rate": 8e-05,
+      "loss": 1.588,
+      "step": 1432
+    },
+    {
+      "epoch": 0.19557799918111096,
+      "grad_norm": 0.19196051359176636,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 1433
+    },
+    {
+      "epoch": 0.1957144806878668,
+      "grad_norm": 0.17708393931388855,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 1434
+    },
+    {
+      "epoch": 0.19585096219462264,
+      "grad_norm": 0.18627843260765076,
+      "learning_rate": 8e-05,
+      "loss": 1.6045,
+      "step": 1435
+    },
+    {
+      "epoch": 0.19598744370137847,
+      "grad_norm": 0.18533670902252197,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1961239252081343,
+      "grad_norm": 0.18477529287338257,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 1437
+    },
+    {
+      "epoch": 0.19626040671489015,
+      "grad_norm": 0.1898396909236908,
+      "learning_rate": 8e-05,
+      "loss": 1.5633,
+      "step": 1438
+    },
+    {
+      "epoch": 0.19639688822164597,
+      "grad_norm": 0.18457242846488953,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1965333697284018,
+      "grad_norm": 0.19674451649188995,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19666985123515765,
+      "grad_norm": 0.1788880079984665,
+      "learning_rate": 8e-05,
+      "loss": 1.5564,
+      "step": 1441
+    },
+    {
+      "epoch": 0.19680633274191348,
+      "grad_norm": 0.19101595878601074,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1969428142486693,
+      "grad_norm": 0.1784212589263916,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 1443
+    },
+    {
+      "epoch": 0.19707929575542513,
+      "grad_norm": 0.18406438827514648,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 1444
+    },
+    {
+      "epoch": 0.19721577726218098,
+      "grad_norm": 0.19459623098373413,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1973522587689368,
+      "grad_norm": 0.17270679771900177,
+      "learning_rate": 8e-05,
+      "loss": 1.5593,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19748874027569263,
+      "grad_norm": 0.1803906261920929,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19762522178244848,
+      "grad_norm": 0.1875525414943695,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1977617032892043,
+      "grad_norm": 0.19411605596542358,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19789818479596014,
+      "grad_norm": 0.18178801238536835,
+      "learning_rate": 8e-05,
+      "loss": 1.5548,
+      "step": 1450
+    },
+    {
+      "epoch": 0.198034666302716,
+      "grad_norm": 0.18365299701690674,
+      "learning_rate": 8e-05,
+      "loss": 1.5959,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19817114780947181,
+      "grad_norm": 0.18477144837379456,
+      "learning_rate": 8e-05,
+      "loss": 1.5584,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19830762931622764,
+      "grad_norm": 0.18778949975967407,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1984441108229835,
+      "grad_norm": 0.1929863840341568,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19858059232973932,
+      "grad_norm": 0.1916200965642929,
+      "learning_rate": 8e-05,
+      "loss": 1.5693,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19871707383649515,
+      "grad_norm": 0.18464729189872742,
+      "learning_rate": 8e-05,
+      "loss": 1.6588,
+      "step": 1456
+    },
+    {
+      "epoch": 0.198853555343251,
+      "grad_norm": 0.1821947544813156,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19899003685000682,
+      "grad_norm": 0.19039857387542725,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19912651835676265,
+      "grad_norm": 0.1781628131866455,
+      "learning_rate": 8e-05,
+      "loss": 1.503,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1992629998635185,
+      "grad_norm": 0.18593347072601318,
+      "learning_rate": 8e-05,
+      "loss": 1.5386,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19939948137027433,
+      "grad_norm": 0.19762296974658966,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19953596287703015,
+      "grad_norm": 0.1734614372253418,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 1462
+    },
+    {
+      "epoch": 0.199672444383786,
+      "grad_norm": 0.18353652954101562,
+      "learning_rate": 8e-05,
+      "loss": 1.4828,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19980892589054183,
+      "grad_norm": 0.18647031486034393,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19994540739729766,
+      "grad_norm": 0.17637072503566742,
+      "learning_rate": 8e-05,
+      "loss": 1.5307,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2000818889040535,
+      "grad_norm": 0.18160459399223328,
+      "learning_rate": 8e-05,
+      "loss": 1.6175,
+      "step": 1466
+    },
+    {
+      "epoch": 0.20021837041080934,
+      "grad_norm": 0.185297429561615,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 1467
+    },
+    {
+      "epoch": 0.20035485191756516,
+      "grad_norm": 0.18620960414409637,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 1468
+    },
+    {
+      "epoch": 0.20049133342432102,
+      "grad_norm": 0.17840619385242462,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 1469
+    },
+    {
+      "epoch": 0.20062781493107684,
+      "grad_norm": 0.1843876987695694,
+      "learning_rate": 8e-05,
+      "loss": 1.5255,
+      "step": 1470
+    },
+    {
+      "epoch": 0.20076429643783267,
+      "grad_norm": 0.18618100881576538,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 1471
+    },
+    {
+      "epoch": 0.20090077794458852,
+      "grad_norm": 0.19297049939632416,
+      "learning_rate": 8e-05,
+      "loss": 1.5528,
+      "step": 1472
+    },
+    {
+      "epoch": 0.20103725945134435,
+      "grad_norm": 0.1964980512857437,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 1473
+    },
+    {
+      "epoch": 0.20117374095810017,
+      "grad_norm": 0.17532451450824738,
+      "learning_rate": 8e-05,
+      "loss": 1.6164,
+      "step": 1474
+    },
+    {
+      "epoch": 0.20131022246485603,
+      "grad_norm": 0.19287341833114624,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 1475
+    },
+    {
+      "epoch": 0.20144670397161185,
+      "grad_norm": 0.18532688915729523,
+      "learning_rate": 8e-05,
+      "loss": 1.5219,
+      "step": 1476
+    },
+    {
+      "epoch": 0.20158318547836768,
+      "grad_norm": 0.18044430017471313,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2017196669851235,
+      "grad_norm": 0.19123688340187073,
+      "learning_rate": 8e-05,
+      "loss": 1.5733,
+      "step": 1478
+    },
+    {
+      "epoch": 0.20185614849187936,
+      "grad_norm": 0.18336538970470428,
+      "learning_rate": 8e-05,
+      "loss": 1.5159,
+      "step": 1479
+    },
+    {
+      "epoch": 0.20199262999863518,
+      "grad_norm": 0.1817123144865036,
+      "learning_rate": 8e-05,
+      "loss": 1.5417,
+      "step": 1480
+    },
+    {
+      "epoch": 0.202129111505391,
+      "grad_norm": 0.19311976432800293,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 1481
+    },
+    {
+      "epoch": 0.20226559301214686,
+      "grad_norm": 0.1777079701423645,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 1482
+    },
+    {
+      "epoch": 0.20240207451890269,
+      "grad_norm": 0.17433413863182068,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2025385560256585,
+      "grad_norm": 0.17603716254234314,
+      "learning_rate": 8e-05,
+      "loss": 1.5548,
+      "step": 1484
+    },
+    {
+      "epoch": 0.20267503753241436,
+      "grad_norm": 0.17388878762722015,
+      "learning_rate": 8e-05,
+      "loss": 1.5054,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2028115190391702,
+      "grad_norm": 0.17641131579875946,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1486
+    },
+    {
+      "epoch": 0.20294800054592602,
+      "grad_norm": 0.1832215040922165,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 1487
+    },
+    {
+      "epoch": 0.20308448205268187,
+      "grad_norm": 0.183164581656456,
+      "learning_rate": 8e-05,
+      "loss": 1.5961,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2032209635594377,
+      "grad_norm": 0.18779879808425903,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 1489
+    },
+    {
+      "epoch": 0.20335744506619352,
+      "grad_norm": 0.17913123965263367,
+      "learning_rate": 8e-05,
+      "loss": 1.5698,
+      "step": 1490
+    },
+    {
+      "epoch": 0.20349392657294937,
+      "grad_norm": 0.17858372628688812,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2036304080797052,
+      "grad_norm": 0.18160991370677948,
+      "learning_rate": 8e-05,
+      "loss": 1.6053,
+      "step": 1492
+    },
+    {
+      "epoch": 0.20376688958646103,
+      "grad_norm": 0.18251852691173553,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 1493
+    },
+    {
+      "epoch": 0.20390337109321688,
+      "grad_norm": 0.17210716009140015,
+      "learning_rate": 8e-05,
+      "loss": 1.5618,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2040398525999727,
+      "grad_norm": 0.1758878231048584,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 1495
+    },
+    {
+      "epoch": 0.20417633410672853,
+      "grad_norm": 0.18337933719158173,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 1496
+    },
+    {
+      "epoch": 0.20431281561348438,
+      "grad_norm": 0.17764587700366974,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2044492971202402,
+      "grad_norm": 0.18335145711898804,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 1498
+    },
+    {
+      "epoch": 0.20458577862699603,
+      "grad_norm": 0.1805732250213623,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2047222601337519,
+      "grad_norm": 0.1846403032541275,
+      "learning_rate": 8e-05,
+      "loss": 1.5964,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2048587416405077,
+      "grad_norm": 0.18115833401679993,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 1501
+    },
+    {
+      "epoch": 0.20499522314726354,
+      "grad_norm": 0.18207481503486633,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2051317046540194,
+      "grad_norm": 0.19865632057189941,
+      "learning_rate": 8e-05,
+      "loss": 1.579,
+      "step": 1503
+    },
+    {
+      "epoch": 0.20526818616077522,
+      "grad_norm": 0.18002501130104065,
+      "learning_rate": 8e-05,
+      "loss": 1.6192,
+      "step": 1504
+    },
+    {
+      "epoch": 0.20540466766753104,
+      "grad_norm": 0.19746966660022736,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2055411491742869,
+      "grad_norm": 0.1856650859117508,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 1506
+    },
+    {
+      "epoch": 0.20567763068104272,
+      "grad_norm": 0.18077273666858673,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 1507
+    },
+    {
+      "epoch": 0.20581411218779855,
+      "grad_norm": 0.1812048852443695,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2059505936945544,
+      "grad_norm": 0.1826622486114502,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 1509
+    },
+    {
+      "epoch": 0.20608707520131023,
+      "grad_norm": 0.18820056319236755,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 1510
+    },
+    {
+      "epoch": 0.20622355670806605,
+      "grad_norm": 0.17998521029949188,
+      "learning_rate": 8e-05,
+      "loss": 1.5749,
+      "step": 1511
+    },
+    {
+      "epoch": 0.20636003821482188,
+      "grad_norm": 0.17936627566814423,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 1512
+    },
+    {
+      "epoch": 0.20649651972157773,
+      "grad_norm": 0.17615634202957153,
+      "learning_rate": 8e-05,
+      "loss": 1.5345,
+      "step": 1513
+    },
+    {
+      "epoch": 0.20663300122833356,
+      "grad_norm": 0.18781231343746185,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 1514
+    },
+    {
+      "epoch": 0.20676948273508938,
+      "grad_norm": 0.174053356051445,
+      "learning_rate": 8e-05,
+      "loss": 1.5677,
+      "step": 1515
+    },
+    {
+      "epoch": 0.20690596424184524,
+      "grad_norm": 0.1739943027496338,
+      "learning_rate": 8e-05,
+      "loss": 1.5526,
+      "step": 1516
+    },
+    {
+      "epoch": 0.20704244574860106,
+      "grad_norm": 0.19110843539237976,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2071789272553569,
+      "grad_norm": 0.18332581222057343,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 1518
+    },
+    {
+      "epoch": 0.20731540876211274,
+      "grad_norm": 0.17619259655475616,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 1519
+    },
+    {
+      "epoch": 0.20745189026886857,
+      "grad_norm": 0.18023639917373657,
+      "learning_rate": 8e-05,
+      "loss": 1.6248,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2075883717756244,
+      "grad_norm": 0.18323586881160736,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 1521
+    },
+    {
+      "epoch": 0.20772485328238025,
+      "grad_norm": 0.17915664613246918,
+      "learning_rate": 8e-05,
+      "loss": 1.5775,
+      "step": 1522
+    },
+    {
+      "epoch": 0.20786133478913607,
+      "grad_norm": 0.1844443380832672,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2079978162958919,
+      "grad_norm": 0.1829414814710617,
+      "learning_rate": 8e-05,
+      "loss": 1.5234,
+      "step": 1524
+    },
+    {
+      "epoch": 0.20813429780264775,
+      "grad_norm": 0.18842652440071106,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 1525
+    },
+    {
+      "epoch": 0.20827077930940358,
+      "grad_norm": 0.19175831973552704,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2084072608161594,
+      "grad_norm": 0.18748562037944794,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 1527
+    },
+    {
+      "epoch": 0.20854374232291525,
+      "grad_norm": 0.18532401323318481,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 1528
+    },
+    {
+      "epoch": 0.20868022382967108,
+      "grad_norm": 0.18803122639656067,
+      "learning_rate": 8e-05,
+      "loss": 1.5104,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2088167053364269,
+      "grad_norm": 0.1791721135377884,
+      "learning_rate": 8e-05,
+      "loss": 1.5574,
+      "step": 1530
+    },
+    {
+      "epoch": 0.20895318684318276,
+      "grad_norm": 0.1932784765958786,
+      "learning_rate": 8e-05,
+      "loss": 1.6243,
+      "step": 1531
+    },
+    {
+      "epoch": 0.20908966834993858,
+      "grad_norm": 0.18730276823043823,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2092261498566944,
+      "grad_norm": 0.17879685759544373,
+      "learning_rate": 8e-05,
+      "loss": 1.4858,
+      "step": 1533
+    },
+    {
+      "epoch": 0.20936263136345026,
+      "grad_norm": 0.1810392588376999,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2094991128702061,
+      "grad_norm": 0.1829460859298706,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 1535
+    },
+    {
+      "epoch": 0.20963559437696191,
+      "grad_norm": 0.1807374358177185,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 1536
+    },
+    {
+      "epoch": 0.20977207588371777,
+      "grad_norm": 0.18087249994277954,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2099085573904736,
+      "grad_norm": 0.17787086963653564,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 1538
+    },
+    {
+      "epoch": 0.21004503889722942,
+      "grad_norm": 0.18385079503059387,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1539
+    },
+    {
+      "epoch": 0.21018152040398527,
+      "grad_norm": 0.1800709366798401,
+      "learning_rate": 8e-05,
+      "loss": 1.6858,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2103180019107411,
+      "grad_norm": 0.17953120172023773,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 1541
+    },
+    {
+      "epoch": 0.21045448341749692,
+      "grad_norm": 0.18363916873931885,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 1542
+    },
+    {
+      "epoch": 0.21059096492425278,
+      "grad_norm": 0.18143697082996368,
+      "learning_rate": 8e-05,
+      "loss": 1.5613,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2107274464310086,
+      "grad_norm": 0.18854306638240814,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 1544
+    },
+    {
+      "epoch": 0.21086392793776443,
+      "grad_norm": 0.18977026641368866,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 1545
+    },
+    {
+      "epoch": 0.21100040944452025,
+      "grad_norm": 0.18302160501480103,
+      "learning_rate": 8e-05,
+      "loss": 1.62,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2111368909512761,
+      "grad_norm": 0.18566632270812988,
+      "learning_rate": 8e-05,
+      "loss": 1.6041,
+      "step": 1547
+    },
+    {
+      "epoch": 0.21127337245803193,
+      "grad_norm": 0.18153445422649384,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 1548
+    },
+    {
+      "epoch": 0.21140985396478776,
+      "grad_norm": 0.18662768602371216,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2115463354715436,
+      "grad_norm": 0.18222737312316895,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 1550
+    },
+    {
+      "epoch": 0.21168281697829944,
+      "grad_norm": 0.18449488282203674,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 1551
+    },
+    {
+      "epoch": 0.21181929848505526,
+      "grad_norm": 0.17821605503559113,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 1552
+    },
+    {
+      "epoch": 0.21195577999181112,
+      "grad_norm": 0.18590347468852997,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 1553
+    },
+    {
+      "epoch": 0.21209226149856694,
+      "grad_norm": 0.18014498054981232,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 1554
+    },
+    {
+      "epoch": 0.21222874300532277,
+      "grad_norm": 0.19307273626327515,
+      "learning_rate": 8e-05,
+      "loss": 1.5766,
+      "step": 1555
+    },
+    {
+      "epoch": 0.21236522451207862,
+      "grad_norm": 0.18647314608097076,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 1556
+    },
+    {
+      "epoch": 0.21250170601883445,
+      "grad_norm": 0.18725787103176117,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 1557
+    },
+    {
+      "epoch": 0.21263818752559027,
+      "grad_norm": 0.1864556521177292,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1558
+    },
+    {
+      "epoch": 0.21277466903234613,
+      "grad_norm": 0.17538176476955414,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 1559
+    },
+    {
+      "epoch": 0.21291115053910195,
+      "grad_norm": 0.1799948811531067,
+      "learning_rate": 8e-05,
+      "loss": 1.4613,
+      "step": 1560
+    },
+    {
+      "epoch": 0.21304763204585778,
+      "grad_norm": 0.18626180291175842,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 1561
+    },
+    {
+      "epoch": 0.21318411355261363,
+      "grad_norm": 0.1895618587732315,
+      "learning_rate": 8e-05,
+      "loss": 1.4661,
+      "step": 1562
+    },
+    {
+      "epoch": 0.21332059505936946,
+      "grad_norm": 0.20060861110687256,
+      "learning_rate": 8e-05,
+      "loss": 1.6723,
+      "step": 1563
+    },
+    {
+      "epoch": 0.21345707656612528,
+      "grad_norm": 0.1764329969882965,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 1564
+    },
+    {
+      "epoch": 0.21359355807288113,
+      "grad_norm": 0.185940220952034,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 1565
+    },
+    {
+      "epoch": 0.21373003957963696,
+      "grad_norm": 0.18389998376369476,
+      "learning_rate": 8e-05,
+      "loss": 1.5572,
+      "step": 1566
+    },
+    {
+      "epoch": 0.21386652108639279,
+      "grad_norm": 0.1898891031742096,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 1567
+    },
+    {
+      "epoch": 0.21400300259314864,
+      "grad_norm": 0.18268591165542603,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 1568
+    },
+    {
+      "epoch": 0.21413948409990446,
+      "grad_norm": 0.18448932468891144,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2142759656066603,
+      "grad_norm": 0.1885816603899002,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 1570
+    },
+    {
+      "epoch": 0.21441244711341614,
+      "grad_norm": 0.19512155652046204,
+      "learning_rate": 8e-05,
+      "loss": 1.6871,
+      "step": 1571
+    },
+    {
+      "epoch": 0.21454892862017197,
+      "grad_norm": 0.18985328078269958,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2146854101269278,
+      "grad_norm": 0.19342605769634247,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 1573
+    },
+    {
+      "epoch": 0.21482189163368365,
+      "grad_norm": 0.184814453125,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 1574
+    },
+    {
+      "epoch": 0.21495837314043947,
+      "grad_norm": 0.1896548569202423,
+      "learning_rate": 8e-05,
+      "loss": 1.5218,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2150948546471953,
+      "grad_norm": 0.18027977645397186,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 1576
+    },
+    {
+      "epoch": 0.21523133615395115,
+      "grad_norm": 0.17841024696826935,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 1577
+    },
+    {
+      "epoch": 0.21536781766070698,
+      "grad_norm": 0.18876133859157562,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2155042991674628,
+      "grad_norm": 0.19057638943195343,
+      "learning_rate": 8e-05,
+      "loss": 1.6539,
+      "step": 1579
+    },
+    {
+      "epoch": 0.21564078067421863,
+      "grad_norm": 0.18067972362041473,
+      "learning_rate": 8e-05,
+      "loss": 1.5582,
+      "step": 1580
+    },
+    {
+      "epoch": 0.21577726218097448,
+      "grad_norm": 0.19346143305301666,
+      "learning_rate": 8e-05,
+      "loss": 1.6287,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2159137436877303,
+      "grad_norm": 0.1792355626821518,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 1582
+    },
+    {
+      "epoch": 0.21605022519448613,
+      "grad_norm": 0.19466540217399597,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 1583
+    },
+    {
+      "epoch": 0.216186706701242,
+      "grad_norm": 0.1889699399471283,
+      "learning_rate": 8e-05,
+      "loss": 1.4937,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2163231882079978,
+      "grad_norm": 0.1760951727628708,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 1585
+    },
+    {
+      "epoch": 0.21645966971475364,
+      "grad_norm": 0.20095914602279663,
+      "learning_rate": 8e-05,
+      "loss": 1.5902,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2165961512215095,
+      "grad_norm": 0.18775932490825653,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 1587
+    },
+    {
+      "epoch": 0.21673263272826532,
+      "grad_norm": 0.18992333114147186,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 1588
+    },
+    {
+      "epoch": 0.21686911423502114,
+      "grad_norm": 0.18613582849502563,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 1589
+    },
+    {
+      "epoch": 0.217005595741777,
+      "grad_norm": 0.19453835487365723,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 1590
+    },
+    {
+      "epoch": 0.21714207724853282,
+      "grad_norm": 0.17925287783145905,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 1591
+    },
+    {
+      "epoch": 0.21727855875528865,
+      "grad_norm": 0.19244016706943512,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2174150402620445,
+      "grad_norm": 0.18168294429779053,
+      "learning_rate": 8e-05,
+      "loss": 1.5767,
+      "step": 1593
+    },
+    {
+      "epoch": 0.21755152176880033,
+      "grad_norm": 0.18998073041439056,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 1594
+    },
+    {
+      "epoch": 0.21768800327555615,
+      "grad_norm": 0.18540899455547333,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 1595
+    },
+    {
+      "epoch": 0.217824484782312,
+      "grad_norm": 0.179458886384964,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 1596
+    },
+    {
+      "epoch": 0.21796096628906783,
+      "grad_norm": 0.185092031955719,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 1597
+    },
+    {
+      "epoch": 0.21809744779582366,
+      "grad_norm": 0.18470264971256256,
+      "learning_rate": 8e-05,
+      "loss": 1.5739,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2182339293025795,
+      "grad_norm": 0.1853020042181015,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 1599
+    },
+    {
+      "epoch": 0.21837041080933534,
+      "grad_norm": 0.1860482096672058,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21850689231609116,
+      "grad_norm": 0.18579822778701782,
+      "learning_rate": 8e-05,
+      "loss": 1.597,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21864337382284701,
+      "grad_norm": 0.17685216665267944,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 1602
+    },
+    {
+      "epoch": 0.21877985532960284,
+      "grad_norm": 0.1827811300754547,
+      "learning_rate": 8e-05,
+      "loss": 1.5486,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21891633683635867,
+      "grad_norm": 0.18281586468219757,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21905281834311452,
+      "grad_norm": 0.18391123414039612,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21918929984987034,
+      "grad_norm": 0.17738930881023407,
+      "learning_rate": 8e-05,
+      "loss": 1.5435,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21932578135662617,
+      "grad_norm": 0.18947827816009521,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 1607
+    },
+    {
+      "epoch": 0.21946226286338202,
+      "grad_norm": 0.1755830943584442,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21959874437013785,
+      "grad_norm": 0.18073181807994843,
+      "learning_rate": 8e-05,
+      "loss": 1.5401,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21973522587689368,
+      "grad_norm": 0.1794690489768982,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21987170738364953,
+      "grad_norm": 0.1837550550699234,
+      "learning_rate": 8e-05,
+      "loss": 1.6088,
+      "step": 1611
+    },
+    {
+      "epoch": 0.22000818889040535,
+      "grad_norm": 0.18821464478969574,
+      "learning_rate": 8e-05,
+      "loss": 1.5675,
+      "step": 1612
+    },
+    {
+      "epoch": 0.22014467039716118,
+      "grad_norm": 0.1863572895526886,
+      "learning_rate": 8e-05,
+      "loss": 1.5586,
+      "step": 1613
+    },
+    {
+      "epoch": 0.22028115190391703,
+      "grad_norm": 0.17708463966846466,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 1614
+    },
+    {
+      "epoch": 0.22041763341067286,
+      "grad_norm": 0.17648367583751678,
+      "learning_rate": 8e-05,
+      "loss": 1.5344,
+      "step": 1615
+    },
+    {
+      "epoch": 0.22055411491742868,
+      "grad_norm": 0.18008606135845184,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2206905964241845,
+      "grad_norm": 0.17940223217010498,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 1617
+    },
+    {
+      "epoch": 0.22082707793094036,
+      "grad_norm": 0.18638667464256287,
+      "learning_rate": 8e-05,
+      "loss": 1.579,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2209635594376962,
+      "grad_norm": 0.18122965097427368,
+      "learning_rate": 8e-05,
+      "loss": 1.5054,
+      "step": 1619
+    },
+    {
+      "epoch": 0.22110004094445201,
+      "grad_norm": 0.17930862307548523,
+      "learning_rate": 8e-05,
+      "loss": 1.4994,
+      "step": 1620
+    },
+    {
+      "epoch": 0.22123652245120787,
+      "grad_norm": 0.18551065027713776,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 1621
+    },
+    {
+      "epoch": 0.2213730039579637,
+      "grad_norm": 0.18968766927719116,
+      "learning_rate": 8e-05,
+      "loss": 1.6192,
+      "step": 1622
+    },
+    {
+      "epoch": 0.22150948546471952,
+      "grad_norm": 0.18144211173057556,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 1623
+    },
+    {
+      "epoch": 0.22164596697147537,
+      "grad_norm": 0.19915767014026642,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2217824484782312,
+      "grad_norm": 0.18128690123558044,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 1625
+    },
+    {
+      "epoch": 0.22191892998498702,
+      "grad_norm": 0.18224012851715088,
+      "learning_rate": 8e-05,
+      "loss": 1.5499,
+      "step": 1626
+    },
+    {
+      "epoch": 0.22205541149174288,
+      "grad_norm": 0.18859682977199554,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2221918929984987,
+      "grad_norm": 0.1913030445575714,
+      "learning_rate": 8e-05,
+      "loss": 1.6293,
+      "step": 1628
+    },
+    {
+      "epoch": 0.22232837450525453,
+      "grad_norm": 0.1801142692565918,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1629
+    },
+    {
+      "epoch": 0.22246485601201038,
+      "grad_norm": 0.18184086680412292,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2226013375187662,
+      "grad_norm": 0.18206322193145752,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 1631
+    },
+    {
+      "epoch": 0.22273781902552203,
+      "grad_norm": 0.18237844109535217,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 1632
+    },
+    {
+      "epoch": 0.22287430053227789,
+      "grad_norm": 0.18598276376724243,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2230107820390337,
+      "grad_norm": 0.18124960362911224,
+      "learning_rate": 8e-05,
+      "loss": 1.6666,
+      "step": 1634
+    },
+    {
+      "epoch": 0.22314726354578954,
+      "grad_norm": 0.18319760262966156,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2232837450525454,
+      "grad_norm": 0.1866164356470108,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 1636
+    },
+    {
+      "epoch": 0.22342022655930122,
+      "grad_norm": 0.1916627287864685,
+      "learning_rate": 8e-05,
+      "loss": 1.6515,
+      "step": 1637
+    },
+    {
+      "epoch": 0.22355670806605704,
+      "grad_norm": 0.19214844703674316,
+      "learning_rate": 8e-05,
+      "loss": 1.6045,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2236931895728129,
+      "grad_norm": 0.189527690410614,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 1639
+    },
+    {
+      "epoch": 0.22382967107956872,
+      "grad_norm": 0.19260188937187195,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 1640
+    },
+    {
+      "epoch": 0.22396615258632455,
+      "grad_norm": 0.18151207268238068,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2241026340930804,
+      "grad_norm": 0.18015263974666595,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 1642
+    },
+    {
+      "epoch": 0.22423911559983623,
+      "grad_norm": 0.17515723407268524,
+      "learning_rate": 8e-05,
+      "loss": 1.4826,
+      "step": 1643
+    },
+    {
+      "epoch": 0.22437559710659205,
+      "grad_norm": 0.2210821658372879,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2245120786133479,
+      "grad_norm": 0.17541536688804626,
+      "learning_rate": 8e-05,
+      "loss": 1.4722,
+      "step": 1645
+    },
+    {
+      "epoch": 0.22464856012010373,
+      "grad_norm": 0.18410120904445648,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 1646
+    },
+    {
+      "epoch": 0.22478504162685956,
+      "grad_norm": 0.18484385311603546,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2249215231336154,
+      "grad_norm": 0.18690799176692963,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1648
+    },
+    {
+      "epoch": 0.22505800464037123,
+      "grad_norm": 0.18821720778942108,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 1649
+    },
+    {
+      "epoch": 0.22519448614712706,
+      "grad_norm": 0.1808300018310547,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22533096765388289,
+      "grad_norm": 0.18520498275756836,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 1651
+    },
+    {
+      "epoch": 0.22546744916063874,
+      "grad_norm": 0.1838509738445282,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 1652
+    },
+    {
+      "epoch": 0.22560393066739456,
+      "grad_norm": 0.18807122111320496,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 1653
+    },
+    {
+      "epoch": 0.2257404121741504,
+      "grad_norm": 0.18304729461669922,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 1654
+    },
+    {
+      "epoch": 0.22587689368090624,
+      "grad_norm": 0.17820486426353455,
+      "learning_rate": 8e-05,
+      "loss": 1.5461,
+      "step": 1655
+    },
+    {
+      "epoch": 0.22601337518766207,
+      "grad_norm": 0.19268783926963806,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2261498566944179,
+      "grad_norm": 0.17507818341255188,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 1657
+    },
+    {
+      "epoch": 0.22628633820117375,
+      "grad_norm": 0.18350698053836823,
+      "learning_rate": 8e-05,
+      "loss": 1.5497,
+      "step": 1658
+    },
+    {
+      "epoch": 0.22642281970792957,
+      "grad_norm": 0.18189282715320587,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2265593012146854,
+      "grad_norm": 0.18441444635391235,
+      "learning_rate": 8e-05,
+      "loss": 1.562,
+      "step": 1660
+    },
+    {
+      "epoch": 0.22669578272144125,
+      "grad_norm": 0.18575315177440643,
+      "learning_rate": 8e-05,
+      "loss": 1.5492,
+      "step": 1661
+    },
+    {
+      "epoch": 0.22683226422819708,
+      "grad_norm": 0.18621037900447845,
+      "learning_rate": 8e-05,
+      "loss": 1.5294,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2269687457349529,
+      "grad_norm": 0.18469302356243134,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1663
+    },
+    {
+      "epoch": 0.22710522724170876,
+      "grad_norm": 0.1926560252904892,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 1664
+    },
+    {
+      "epoch": 0.22724170874846458,
+      "grad_norm": 0.1905696988105774,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2273781902552204,
+      "grad_norm": 0.1913861632347107,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 1666
+    },
+    {
+      "epoch": 0.22751467176197626,
+      "grad_norm": 0.19802449643611908,
+      "learning_rate": 8e-05,
+      "loss": 1.5716,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2276511532687321,
+      "grad_norm": 0.18814772367477417,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2277876347754879,
+      "grad_norm": 0.19677011668682098,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 1669
+    },
+    {
+      "epoch": 0.22792411628224377,
+      "grad_norm": 0.19284853339195251,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2280605977889996,
+      "grad_norm": 0.17543110251426697,
+      "learning_rate": 8e-05,
+      "loss": 1.5528,
+      "step": 1671
+    },
+    {
+      "epoch": 0.22819707929575542,
+      "grad_norm": 0.1940460354089737,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 1672
+    },
+    {
+      "epoch": 0.22833356080251127,
+      "grad_norm": 0.19870352745056152,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2284700423092671,
+      "grad_norm": 0.19045959413051605,
+      "learning_rate": 8e-05,
+      "loss": 1.5823,
+      "step": 1674
+    },
+    {
+      "epoch": 0.22860652381602292,
+      "grad_norm": 0.19584600627422333,
+      "learning_rate": 8e-05,
+      "loss": 1.6097,
+      "step": 1675
+    },
+    {
+      "epoch": 0.22874300532277878,
+      "grad_norm": 0.1913304328918457,
+      "learning_rate": 8e-05,
+      "loss": 1.6193,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2288794868295346,
+      "grad_norm": 0.17441055178642273,
+      "learning_rate": 8e-05,
+      "loss": 1.5261,
+      "step": 1677
+    },
+    {
+      "epoch": 0.22901596833629043,
+      "grad_norm": 0.19493237137794495,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 1678
+    },
+    {
+      "epoch": 0.22915244984304628,
+      "grad_norm": 0.1975114345550537,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2292889313498021,
+      "grad_norm": 0.18074315786361694,
+      "learning_rate": 8e-05,
+      "loss": 1.5748,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22942541285655793,
+      "grad_norm": 0.19960017502307892,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22956189436331378,
+      "grad_norm": 0.17777031660079956,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 1682
+    },
+    {
+      "epoch": 0.2296983758700696,
+      "grad_norm": 0.18749453127384186,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22983485737682544,
+      "grad_norm": 0.18549850583076477,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1684
+    },
+    {
+      "epoch": 0.22997133888358126,
+      "grad_norm": 0.18856392800807953,
+      "learning_rate": 8e-05,
+      "loss": 1.5913,
+      "step": 1685
+    },
+    {
+      "epoch": 0.23010782039033711,
+      "grad_norm": 0.18631874024868011,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 1686
+    },
+    {
+      "epoch": 0.23024430189709294,
+      "grad_norm": 0.18141521513462067,
+      "learning_rate": 8e-05,
+      "loss": 1.5771,
+      "step": 1687
+    },
+    {
+      "epoch": 0.23038078340384877,
+      "grad_norm": 0.18833263218402863,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 1688
+    },
+    {
+      "epoch": 0.23051726491060462,
+      "grad_norm": 0.17912141978740692,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 1689
+    },
+    {
+      "epoch": 0.23065374641736044,
+      "grad_norm": 0.17798754572868347,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 1690
+    },
+    {
+      "epoch": 0.23079022792411627,
+      "grad_norm": 0.192842498421669,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 1691
+    },
+    {
+      "epoch": 0.23092670943087212,
+      "grad_norm": 0.18777163326740265,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 1692
+    },
+    {
+      "epoch": 0.23106319093762795,
+      "grad_norm": 0.2039753496646881,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1693
+    },
+    {
+      "epoch": 0.23119967244438377,
+      "grad_norm": 0.1855362504720688,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 1694
+    },
+    {
+      "epoch": 0.23133615395113963,
+      "grad_norm": 0.18319478631019592,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 1695
+    },
+    {
+      "epoch": 0.23147263545789545,
+      "grad_norm": 0.19037805497646332,
+      "learning_rate": 8e-05,
+      "loss": 1.5054,
+      "step": 1696
+    },
+    {
+      "epoch": 0.23160911696465128,
+      "grad_norm": 0.19100148975849152,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 1697
+    },
+    {
+      "epoch": 0.23174559847140713,
+      "grad_norm": 0.18612314760684967,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 1698
+    },
+    {
+      "epoch": 0.23188207997816296,
+      "grad_norm": 0.1895783245563507,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 1699
+    },
+    {
+      "epoch": 0.23201856148491878,
+      "grad_norm": 0.18185438215732574,
+      "learning_rate": 8e-05,
+      "loss": 1.4859,
+      "step": 1700
+    },
+    {
+      "epoch": 0.23215504299167464,
+      "grad_norm": 0.18597343564033508,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 1701
+    },
+    {
+      "epoch": 0.23229152449843046,
+      "grad_norm": 0.18556249141693115,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2324280060051863,
+      "grad_norm": 0.17800599336624146,
+      "learning_rate": 8e-05,
+      "loss": 1.485,
+      "step": 1703
+    },
+    {
+      "epoch": 0.23256448751194214,
+      "grad_norm": 0.1902007907629013,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 1704
+    },
+    {
+      "epoch": 0.23270096901869797,
+      "grad_norm": 0.18084144592285156,
+      "learning_rate": 8e-05,
+      "loss": 1.4766,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2328374505254538,
+      "grad_norm": 0.18228298425674438,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 1706
+    },
+    {
+      "epoch": 0.23297393203220965,
+      "grad_norm": 0.18154548108577728,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 1707
+    },
+    {
+      "epoch": 0.23311041353896547,
+      "grad_norm": 0.19297222793102264,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2332468950457213,
+      "grad_norm": 0.18998752534389496,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 1709
+    },
+    {
+      "epoch": 0.23338337655247715,
+      "grad_norm": 0.18729344010353088,
+      "learning_rate": 8e-05,
+      "loss": 1.5876,
+      "step": 1710
+    },
+    {
+      "epoch": 0.23351985805923298,
+      "grad_norm": 0.1814780831336975,
+      "learning_rate": 8e-05,
+      "loss": 1.5308,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2336563395659888,
+      "grad_norm": 0.1783227175474167,
+      "learning_rate": 8e-05,
+      "loss": 1.4944,
+      "step": 1712
+    },
+    {
+      "epoch": 0.23379282107274466,
+      "grad_norm": 0.17750617861747742,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 1713
+    },
+    {
+      "epoch": 0.23392930257950048,
+      "grad_norm": 0.18692000210285187,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2340657840862563,
+      "grad_norm": 0.19738642871379852,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 1715
+    },
+    {
+      "epoch": 0.23420226559301216,
+      "grad_norm": 0.18085099756717682,
+      "learning_rate": 8e-05,
+      "loss": 1.5419,
+      "step": 1716
+    },
+    {
+      "epoch": 0.23433874709976799,
+      "grad_norm": 0.18681450188159943,
+      "learning_rate": 8e-05,
+      "loss": 1.5564,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2344752286065238,
+      "grad_norm": 0.18711727857589722,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 1718
+    },
+    {
+      "epoch": 0.23461171011327964,
+      "grad_norm": 0.17836855351924896,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2347481916200355,
+      "grad_norm": 0.18290987610816956,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 1720
+    },
+    {
+      "epoch": 0.23488467312679132,
+      "grad_norm": 0.1832231879234314,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 1721
+    },
+    {
+      "epoch": 0.23502115463354714,
+      "grad_norm": 0.18527549505233765,
+      "learning_rate": 8e-05,
+      "loss": 1.5402,
+      "step": 1722
+    },
+    {
+      "epoch": 0.235157636140303,
+      "grad_norm": 0.18735694885253906,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 1723
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.1856762021780014,
+      "learning_rate": 8e-05,
+      "loss": 1.5542,
+      "step": 1724
+    },
+    {
+      "epoch": 0.23543059915381465,
+      "grad_norm": 0.18207597732543945,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2355670806605705,
+      "grad_norm": 0.18174079060554504,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 1726
+    },
+    {
+      "epoch": 0.23570356216732632,
+      "grad_norm": 0.17819462716579437,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1727
+    },
+    {
+      "epoch": 0.23584004367408215,
+      "grad_norm": 0.1815337985754013,
+      "learning_rate": 8e-05,
+      "loss": 1.5385,
+      "step": 1728
+    },
+    {
+      "epoch": 0.235976525180838,
+      "grad_norm": 0.18723762035369873,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 1729
+    },
+    {
+      "epoch": 0.23611300668759383,
+      "grad_norm": 0.18837349116802216,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 1730
+    },
+    {
+      "epoch": 0.23624948819434965,
+      "grad_norm": 0.1874261498451233,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2363859697011055,
+      "grad_norm": 0.1893012374639511,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 1732
+    },
+    {
+      "epoch": 0.23652245120786133,
+      "grad_norm": 0.18179252743721008,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 1733
+    },
+    {
+      "epoch": 0.23665893271461716,
+      "grad_norm": 0.18713226914405823,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 1734
+    },
+    {
+      "epoch": 0.236795414221373,
+      "grad_norm": 0.1828542947769165,
+      "learning_rate": 8e-05,
+      "loss": 1.5939,
+      "step": 1735
+    },
+    {
+      "epoch": 0.23693189572812884,
+      "grad_norm": 0.1856721192598343,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 1736
+    },
+    {
+      "epoch": 0.23706837723488466,
+      "grad_norm": 0.17282211780548096,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 1737
+    },
+    {
+      "epoch": 0.23720485874164052,
+      "grad_norm": 0.18401271104812622,
+      "learning_rate": 8e-05,
+      "loss": 1.5894,
+      "step": 1738
+    },
+    {
+      "epoch": 0.23734134024839634,
+      "grad_norm": 0.18857260048389435,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 1739
+    },
+    {
+      "epoch": 0.23747782175515217,
+      "grad_norm": 0.1815832406282425,
+      "learning_rate": 8e-05,
+      "loss": 1.5199,
+      "step": 1740
+    },
+    {
+      "epoch": 0.23761430326190802,
+      "grad_norm": 0.18803003430366516,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 1741
+    },
+    {
+      "epoch": 0.23775078476866385,
+      "grad_norm": 0.18170347809791565,
+      "learning_rate": 8e-05,
+      "loss": 1.5681,
+      "step": 1742
+    },
+    {
+      "epoch": 0.23788726627541967,
+      "grad_norm": 0.18565987050533295,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 1743
+    },
+    {
+      "epoch": 0.23802374778217553,
+      "grad_norm": 0.1861896961927414,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 1744
+    },
+    {
+      "epoch": 0.23816022928893135,
+      "grad_norm": 0.18452812731266022,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 1745
+    },
+    {
+      "epoch": 0.23829671079568718,
+      "grad_norm": 0.18803074955940247,
+      "learning_rate": 8e-05,
+      "loss": 1.5837,
+      "step": 1746
+    },
+    {
+      "epoch": 0.23843319230244303,
+      "grad_norm": 0.18187682330608368,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 1747
+    },
+    {
+      "epoch": 0.23856967380919886,
+      "grad_norm": 0.1811991035938263,
+      "learning_rate": 8e-05,
+      "loss": 1.5395,
+      "step": 1748
+    },
+    {
+      "epoch": 0.23870615531595468,
+      "grad_norm": 0.18209730088710785,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1749
+    },
+    {
+      "epoch": 0.23884263682271054,
+      "grad_norm": 0.17959095537662506,
+      "learning_rate": 8e-05,
+      "loss": 1.4725,
+      "step": 1750
+    },
+    {
+      "epoch": 0.23897911832946636,
+      "grad_norm": 0.18598006665706635,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2391155998362222,
+      "grad_norm": 0.1867654174566269,
+      "learning_rate": 8e-05,
+      "loss": 1.5633,
+      "step": 1752
+    },
+    {
+      "epoch": 0.23925208134297804,
+      "grad_norm": 0.18545357882976532,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 1753
+    },
+    {
+      "epoch": 0.23938856284973387,
+      "grad_norm": 0.1776314377784729,
+      "learning_rate": 8e-05,
+      "loss": 1.5031,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2395250443564897,
+      "grad_norm": 0.18356305360794067,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 1755
+    },
+    {
+      "epoch": 0.23966152586324552,
+      "grad_norm": 0.19067412614822388,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 1756
+    },
+    {
+      "epoch": 0.23979800737000137,
+      "grad_norm": 0.18596555292606354,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2399344888767572,
+      "grad_norm": 0.1766899973154068,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 1758
+    },
+    {
+      "epoch": 0.24007097038351302,
+      "grad_norm": 0.18611909449100494,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 1759
+    },
+    {
+      "epoch": 0.24020745189026887,
+      "grad_norm": 0.2000034898519516,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2403439333970247,
+      "grad_norm": 0.18233288824558258,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 1761
+    },
+    {
+      "epoch": 0.24048041490378053,
+      "grad_norm": 0.18243840336799622,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 1762
+    },
+    {
+      "epoch": 0.24061689641053638,
+      "grad_norm": 0.18466459214687347,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2407533779172922,
+      "grad_norm": 0.1838509738445282,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1764
+    },
+    {
+      "epoch": 0.24088985942404803,
+      "grad_norm": 0.18624763190746307,
+      "learning_rate": 8e-05,
+      "loss": 1.5161,
+      "step": 1765
+    },
+    {
+      "epoch": 0.24102634093080388,
+      "grad_norm": 0.18495598435401917,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2411628224375597,
+      "grad_norm": 0.177808940410614,
+      "learning_rate": 8e-05,
+      "loss": 1.4451,
+      "step": 1767
+    },
+    {
+      "epoch": 0.24129930394431554,
+      "grad_norm": 0.1832917183637619,
+      "learning_rate": 8e-05,
+      "loss": 1.5571,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2414357854510714,
+      "grad_norm": 0.18603256344795227,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 1769
+    },
+    {
+      "epoch": 0.24157226695782721,
+      "grad_norm": 0.19211837649345398,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 1770
+    },
+    {
+      "epoch": 0.24170874846458304,
+      "grad_norm": 0.18810877203941345,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 1771
+    },
+    {
+      "epoch": 0.2418452299713389,
+      "grad_norm": 0.1866840422153473,
+      "learning_rate": 8e-05,
+      "loss": 1.6311,
+      "step": 1772
+    },
+    {
+      "epoch": 0.24198171147809472,
+      "grad_norm": 0.18082091212272644,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 1773
+    },
+    {
+      "epoch": 0.24211819298485054,
+      "grad_norm": 0.18563398718833923,
+      "learning_rate": 8e-05,
+      "loss": 1.503,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2422546744916064,
+      "grad_norm": 0.1880609095096588,
+      "learning_rate": 8e-05,
+      "loss": 1.564,
+      "step": 1775
+    },
+    {
+      "epoch": 0.24239115599836222,
+      "grad_norm": 0.19007337093353271,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 1776
+    },
+    {
+      "epoch": 0.24252763750511805,
+      "grad_norm": 0.19348379969596863,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2426641190118739,
+      "grad_norm": 0.18118761479854584,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 1778
+    },
+    {
+      "epoch": 0.24280060051862973,
+      "grad_norm": 0.17686647176742554,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 1779
+    },
+    {
+      "epoch": 0.24293708202538555,
+      "grad_norm": 0.17891229689121246,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2430735635321414,
+      "grad_norm": 0.18298952281475067,
+      "learning_rate": 8e-05,
+      "loss": 1.5231,
+      "step": 1781
+    },
+    {
+      "epoch": 0.24321004503889723,
+      "grad_norm": 0.18165840208530426,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 1782
+    },
+    {
+      "epoch": 0.24334652654565306,
+      "grad_norm": 0.18146726489067078,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2434830080524089,
+      "grad_norm": 0.18730725347995758,
+      "learning_rate": 8e-05,
+      "loss": 1.5268,
+      "step": 1784
+    },
+    {
+      "epoch": 0.24361948955916474,
+      "grad_norm": 0.18791694939136505,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 1785
+    },
+    {
+      "epoch": 0.24375597106592056,
+      "grad_norm": 0.1907023787498474,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 1786
+    },
+    {
+      "epoch": 0.24389245257267642,
+      "grad_norm": 0.18179954588413239,
+      "learning_rate": 8e-05,
+      "loss": 1.558,
+      "step": 1787
+    },
+    {
+      "epoch": 0.24402893407943224,
+      "grad_norm": 0.1894650012254715,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 1788
+    },
+    {
+      "epoch": 0.24416541558618807,
+      "grad_norm": 0.18077702820301056,
+      "learning_rate": 8e-05,
+      "loss": 1.5108,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2443018970929439,
+      "grad_norm": 0.18770000338554382,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 1790
+    },
+    {
+      "epoch": 0.24443837859969975,
+      "grad_norm": 0.18682396411895752,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1791
+    },
+    {
+      "epoch": 0.24457486010645557,
+      "grad_norm": 0.18906524777412415,
+      "learning_rate": 8e-05,
+      "loss": 1.5516,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2447113416132114,
+      "grad_norm": 0.18724487721920013,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 1793
+    },
+    {
+      "epoch": 0.24484782311996725,
+      "grad_norm": 0.1839170902967453,
+      "learning_rate": 8e-05,
+      "loss": 1.5458,
+      "step": 1794
+    },
+    {
+      "epoch": 0.24498430462672308,
+      "grad_norm": 0.18446402251720428,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1795
+    },
+    {
+      "epoch": 0.2451207861334789,
+      "grad_norm": 0.185521200299263,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 1796
+    },
+    {
+      "epoch": 0.24525726764023476,
+      "grad_norm": 0.18370167911052704,
+      "learning_rate": 8e-05,
+      "loss": 1.5303,
+      "step": 1797
+    },
+    {
+      "epoch": 0.24539374914699058,
+      "grad_norm": 0.17842985689640045,
+      "learning_rate": 8e-05,
+      "loss": 1.5075,
+      "step": 1798
+    },
+    {
+      "epoch": 0.2455302306537464,
+      "grad_norm": 0.1852072924375534,
+      "learning_rate": 8e-05,
+      "loss": 1.6184,
+      "step": 1799
+    },
+    {
+      "epoch": 0.24566671216050226,
+      "grad_norm": 0.18539951741695404,
+      "learning_rate": 8e-05,
+      "loss": 1.5343,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24580319366725809,
+      "grad_norm": 0.18143805861473083,
+      "learning_rate": 8e-05,
+      "loss": 1.5167,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2459396751740139,
+      "grad_norm": 0.1851867288351059,
+      "learning_rate": 8e-05,
+      "loss": 1.6343,
+      "step": 1802
+    },
+    {
+      "epoch": 0.24607615668076976,
+      "grad_norm": 0.18274852633476257,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2462126381875256,
+      "grad_norm": 0.19242529571056366,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 1804
+    },
+    {
+      "epoch": 0.24634911969428142,
+      "grad_norm": 0.20002421736717224,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 1805
+    },
+    {
+      "epoch": 0.24648560120103727,
+      "grad_norm": 0.18668735027313232,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2466220827077931,
+      "grad_norm": 0.20423966646194458,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 1807
+    },
+    {
+      "epoch": 0.24675856421454892,
+      "grad_norm": 0.19704192876815796,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 1808
+    },
+    {
+      "epoch": 0.24689504572130477,
+      "grad_norm": 0.19407778978347778,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2470315272280606,
+      "grad_norm": 0.19340352714061737,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 1810
+    },
+    {
+      "epoch": 0.24716800873481642,
+      "grad_norm": 0.18367774784564972,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 1811
+    },
+    {
+      "epoch": 0.24730449024157228,
+      "grad_norm": 0.19014742970466614,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2474409717483281,
+      "grad_norm": 0.1854240447282791,
+      "learning_rate": 8e-05,
+      "loss": 1.6484,
+      "step": 1813
+    },
+    {
+      "epoch": 0.24757745325508393,
+      "grad_norm": 0.18356232345104218,
+      "learning_rate": 8e-05,
+      "loss": 1.5342,
+      "step": 1814
+    },
+    {
+      "epoch": 0.24771393476183978,
+      "grad_norm": 0.18884854018688202,
+      "learning_rate": 8e-05,
+      "loss": 1.5875,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2478504162685956,
+      "grad_norm": 0.19298192858695984,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 1816
+    },
+    {
+      "epoch": 0.24798689777535143,
+      "grad_norm": 0.1852366030216217,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2481233792821073,
+      "grad_norm": 0.18575924634933472,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2482598607888631,
+      "grad_norm": 0.18977604806423187,
+      "learning_rate": 8e-05,
+      "loss": 1.5989,
+      "step": 1819
+    },
+    {
+      "epoch": 0.24839634229561894,
+      "grad_norm": 0.1818329095840454,
+      "learning_rate": 8e-05,
+      "loss": 1.588,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2485328238023748,
+      "grad_norm": 0.1889045685529709,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 1821
+    },
+    {
+      "epoch": 0.24866930530913062,
+      "grad_norm": 0.18959684669971466,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 1822
+    },
+    {
+      "epoch": 0.24880578681588644,
+      "grad_norm": 0.18398132920265198,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 1823
+    },
+    {
+      "epoch": 0.24894226832264227,
+      "grad_norm": 0.1816786378622055,
+      "learning_rate": 8e-05,
+      "loss": 1.5566,
+      "step": 1824
+    },
+    {
+      "epoch": 0.24907874982939812,
+      "grad_norm": 0.18479779362678528,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 1825
+    },
+    {
+      "epoch": 0.24921523133615395,
+      "grad_norm": 0.19588877260684967,
+      "learning_rate": 8e-05,
+      "loss": 1.5449,
+      "step": 1826
+    },
+    {
+      "epoch": 0.24935171284290977,
+      "grad_norm": 0.18671122193336487,
+      "learning_rate": 8e-05,
+      "loss": 1.6456,
+      "step": 1827
+    },
+    {
+      "epoch": 0.24948819434966563,
+      "grad_norm": 0.1876567304134369,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 1828
+    },
+    {
+      "epoch": 0.24962467585642145,
+      "grad_norm": 0.1924601048231125,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1829
+    },
+    {
+      "epoch": 0.24976115736317728,
+      "grad_norm": 0.17967765033245087,
+      "learning_rate": 8e-05,
+      "loss": 1.542,
+      "step": 1830
+    },
+    {
+      "epoch": 0.24989763886993313,
+      "grad_norm": 0.1899469792842865,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 1831
+    },
+    {
+      "epoch": 0.250034120376689,
+      "grad_norm": 0.1814192235469818,
+      "learning_rate": 8e-05,
+      "loss": 1.5524,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2501706018834448,
+      "grad_norm": 0.18947488069534302,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 1833
+    },
+    {
+      "epoch": 0.25030708339020064,
+      "grad_norm": 0.1923573613166809,
+      "learning_rate": 8e-05,
+      "loss": 1.5693,
+      "step": 1834
+    },
+    {
+      "epoch": 0.25044356489695646,
+      "grad_norm": 0.19979293644428253,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2505800464037123,
+      "grad_norm": 0.1897367388010025,
+      "learning_rate": 8e-05,
+      "loss": 1.5894,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2507165279104681,
+      "grad_norm": 0.194015771150589,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 1837
+    },
+    {
+      "epoch": 0.250853009417224,
+      "grad_norm": 0.1856689304113388,
+      "learning_rate": 8e-05,
+      "loss": 1.4927,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2509894909239798,
+      "grad_norm": 0.1950702965259552,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 1839
+    },
+    {
+      "epoch": 0.25112597243073564,
+      "grad_norm": 0.18865469098091125,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 1840
+    },
+    {
+      "epoch": 0.25126245393749147,
+      "grad_norm": 0.18866439163684845,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2513989354442473,
+      "grad_norm": 0.18741683661937714,
+      "learning_rate": 8e-05,
+      "loss": 1.5315,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2515354169510031,
+      "grad_norm": 0.18092559278011322,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 1843
+    },
+    {
+      "epoch": 0.25167189845775895,
+      "grad_norm": 0.1817786991596222,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 1844
+    },
+    {
+      "epoch": 0.25180837996451483,
+      "grad_norm": 0.18489402532577515,
+      "learning_rate": 8e-05,
+      "loss": 1.594,
+      "step": 1845
+    },
+    {
+      "epoch": 0.25194486147127065,
+      "grad_norm": 0.18681347370147705,
+      "learning_rate": 8e-05,
+      "loss": 1.5142,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2520813429780265,
+      "grad_norm": 0.1880158931016922,
+      "learning_rate": 8e-05,
+      "loss": 1.6058,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2522178244847823,
+      "grad_norm": 0.18516340851783752,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 1848
+    },
+    {
+      "epoch": 0.25235430599153813,
+      "grad_norm": 0.18830637633800507,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 1849
+    },
+    {
+      "epoch": 0.25249078749829396,
+      "grad_norm": 0.17914266884326935,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 1850
+    },
+    {
+      "epoch": 0.25262726900504984,
+      "grad_norm": 0.1884596347808838,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 1851
+    },
+    {
+      "epoch": 0.25276375051180566,
+      "grad_norm": 0.18159277737140656,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2529002320185615,
+      "grad_norm": 0.18422812223434448,
+      "learning_rate": 8e-05,
+      "loss": 1.5541,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2530367135253173,
+      "grad_norm": 0.191286101937294,
+      "learning_rate": 8e-05,
+      "loss": 1.5673,
+      "step": 1854
+    },
+    {
+      "epoch": 0.25317319503207314,
+      "grad_norm": 0.18754881620407104,
+      "learning_rate": 8e-05,
+      "loss": 1.5608,
+      "step": 1855
+    },
+    {
+      "epoch": 0.25330967653882897,
+      "grad_norm": 0.18286220729351044,
+      "learning_rate": 8e-05,
+      "loss": 1.5371,
+      "step": 1856
+    },
+    {
+      "epoch": 0.25344615804558485,
+      "grad_norm": 0.17880402505397797,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 1857
+    },
+    {
+      "epoch": 0.25358263955234067,
+      "grad_norm": 0.18703843653202057,
+      "learning_rate": 8e-05,
+      "loss": 1.5652,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2537191210590965,
+      "grad_norm": 0.1764143854379654,
+      "learning_rate": 8e-05,
+      "loss": 1.512,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2538556025658523,
+      "grad_norm": 0.1798507571220398,
+      "learning_rate": 8e-05,
+      "loss": 1.5014,
+      "step": 1860
+    },
+    {
+      "epoch": 0.25399208407260815,
+      "grad_norm": 0.1905338168144226,
+      "learning_rate": 8e-05,
+      "loss": 1.5993,
+      "step": 1861
+    },
+    {
+      "epoch": 0.254128565579364,
+      "grad_norm": 0.18906225264072418,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 1862
+    },
+    {
+      "epoch": 0.25426504708611986,
+      "grad_norm": 0.18377509713172913,
+      "learning_rate": 8e-05,
+      "loss": 1.5638,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2544015285928757,
+      "grad_norm": 0.18969833850860596,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2545380100996315,
+      "grad_norm": 0.17398011684417725,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 1865
+    },
+    {
+      "epoch": 0.25467449160638733,
+      "grad_norm": 0.18954981863498688,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 1866
+    },
+    {
+      "epoch": 0.25481097311314316,
+      "grad_norm": 0.18197040259838104,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 1867
+    },
+    {
+      "epoch": 0.254947454619899,
+      "grad_norm": 0.18149541318416595,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 1868
+    },
+    {
+      "epoch": 0.25508393612665486,
+      "grad_norm": 0.1875280886888504,
+      "learning_rate": 8e-05,
+      "loss": 1.5445,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2552204176334107,
+      "grad_norm": 0.19960227608680725,
+      "learning_rate": 8e-05,
+      "loss": 1.5625,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2553568991401665,
+      "grad_norm": 0.1862519383430481,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 1871
+    },
+    {
+      "epoch": 0.25549338064692234,
+      "grad_norm": 0.20189553499221802,
+      "learning_rate": 8e-05,
+      "loss": 1.6515,
+      "step": 1872
+    },
+    {
+      "epoch": 0.25562986215367817,
+      "grad_norm": 0.18987150490283966,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 1873
+    },
+    {
+      "epoch": 0.255766343660434,
+      "grad_norm": 0.19546480476856232,
+      "learning_rate": 8e-05,
+      "loss": 1.5187,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2559028251671899,
+      "grad_norm": 0.19442078471183777,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2560393066739457,
+      "grad_norm": 0.1887025535106659,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2561757881807015,
+      "grad_norm": 0.20818020403385162,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 1877
+    },
+    {
+      "epoch": 0.25631226968745735,
+      "grad_norm": 0.18358318507671356,
+      "learning_rate": 8e-05,
+      "loss": 1.5764,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2564487511942132,
+      "grad_norm": 0.18465085327625275,
+      "learning_rate": 8e-05,
+      "loss": 1.5487,
+      "step": 1879
+    },
+    {
+      "epoch": 0.256585232700969,
+      "grad_norm": 0.18243341147899628,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2567217142077248,
+      "grad_norm": 0.18170012533664703,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2568581957144807,
+      "grad_norm": 0.18276526033878326,
+      "learning_rate": 8e-05,
+      "loss": 1.5194,
+      "step": 1882
+    },
+    {
+      "epoch": 0.25699467722123653,
+      "grad_norm": 0.1819353848695755,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 1883
+    },
+    {
+      "epoch": 0.25713115872799236,
+      "grad_norm": 0.18505947291851044,
+      "learning_rate": 8e-05,
+      "loss": 1.4915,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2572676402347482,
+      "grad_norm": 0.18670497834682465,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 1885
+    },
+    {
+      "epoch": 0.257404121741504,
+      "grad_norm": 0.1910771280527115,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 1886
+    },
+    {
+      "epoch": 0.25754060324825984,
+      "grad_norm": 0.18426668643951416,
+      "learning_rate": 8e-05,
+      "loss": 1.4876,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2576770847550157,
+      "grad_norm": 0.1860811412334442,
+      "learning_rate": 8e-05,
+      "loss": 1.5694,
+      "step": 1888
+    },
+    {
+      "epoch": 0.25781356626177154,
+      "grad_norm": 0.18157057464122772,
+      "learning_rate": 8e-05,
+      "loss": 1.5061,
+      "step": 1889
+    },
+    {
+      "epoch": 0.25795004776852737,
+      "grad_norm": 0.18245695531368256,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2580865292752832,
+      "grad_norm": 0.1781587302684784,
+      "learning_rate": 8e-05,
+      "loss": 1.4823,
+      "step": 1891
+    },
+    {
+      "epoch": 0.258223010782039,
+      "grad_norm": 0.18184123933315277,
+      "learning_rate": 8e-05,
+      "loss": 1.5435,
+      "step": 1892
+    },
+    {
+      "epoch": 0.25835949228879485,
+      "grad_norm": 0.1817796677350998,
+      "learning_rate": 8e-05,
+      "loss": 1.5381,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2584959737955507,
+      "grad_norm": 0.18690018355846405,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 1894
+    },
+    {
+      "epoch": 0.25863245530230655,
+      "grad_norm": 0.1951090395450592,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2587689368090624,
+      "grad_norm": 0.17705518007278442,
+      "learning_rate": 8e-05,
+      "loss": 1.5145,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2589054183158182,
+      "grad_norm": 0.18770405650138855,
+      "learning_rate": 8e-05,
+      "loss": 1.514,
+      "step": 1897
+    },
+    {
+      "epoch": 0.25904189982257403,
+      "grad_norm": 0.1851692795753479,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 1898
+    },
+    {
+      "epoch": 0.25917838132932985,
+      "grad_norm": 0.1832483559846878,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 1899
+    },
+    {
+      "epoch": 0.25931486283608574,
+      "grad_norm": 0.1904088258743286,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 1900
+    },
+    {
+      "epoch": 0.25945134434284156,
+      "grad_norm": 0.1859157830476761,
+      "learning_rate": 8e-05,
+      "loss": 1.5848,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2595878258495974,
+      "grad_norm": 0.18166902661323547,
+      "learning_rate": 8e-05,
+      "loss": 1.5155,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2597243073563532,
+      "grad_norm": 0.18620115518569946,
+      "learning_rate": 8e-05,
+      "loss": 1.5589,
+      "step": 1903
+    },
+    {
+      "epoch": 0.25986078886310904,
+      "grad_norm": 0.1801372766494751,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 1904
+    },
+    {
+      "epoch": 0.25999727036986486,
+      "grad_norm": 0.19733835756778717,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 1905
+    },
+    {
+      "epoch": 0.26013375187662074,
+      "grad_norm": 0.19153082370758057,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 1906
+    },
+    {
+      "epoch": 0.26027023338337657,
+      "grad_norm": 0.18736200034618378,
+      "learning_rate": 8e-05,
+      "loss": 1.5959,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2604067148901324,
+      "grad_norm": 0.18396945297718048,
+      "learning_rate": 8e-05,
+      "loss": 1.5616,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2605431963968882,
+      "grad_norm": 0.18430747091770172,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1909
+    },
+    {
+      "epoch": 0.26067967790364405,
+      "grad_norm": 0.18824610114097595,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2608161594103999,
+      "grad_norm": 0.174940288066864,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2609526409171557,
+      "grad_norm": 0.1898488998413086,
+      "learning_rate": 8e-05,
+      "loss": 1.6297,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2610891224239116,
+      "grad_norm": 0.188020721077919,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2612256039306674,
+      "grad_norm": 0.19418668746948242,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 1914
+    },
+    {
+      "epoch": 0.26136208543742323,
+      "grad_norm": 0.1825471967458725,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 1915
+    },
+    {
+      "epoch": 0.26149856694417906,
+      "grad_norm": 0.18769685924053192,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2616350484509349,
+      "grad_norm": 0.1926618069410324,
+      "learning_rate": 8e-05,
+      "loss": 1.6287,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2617715299576907,
+      "grad_norm": 0.1830102503299713,
+      "learning_rate": 8e-05,
+      "loss": 1.6395,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2619080114644466,
+      "grad_norm": 0.1960916668176651,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2620444929712024,
+      "grad_norm": 0.17569278180599213,
+      "learning_rate": 8e-05,
+      "loss": 1.511,
+      "step": 1920
+    },
+    {
+      "epoch": 0.26218097447795824,
+      "grad_norm": 0.19341592490673065,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 1921
+    },
+    {
+      "epoch": 0.26231745598471407,
+      "grad_norm": 0.19142979383468628,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2624539374914699,
+      "grad_norm": 0.20391963422298431,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2625904189982257,
+      "grad_norm": 0.18116265535354614,
+      "learning_rate": 8e-05,
+      "loss": 1.5303,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2627269005049816,
+      "grad_norm": 0.19184935092926025,
+      "learning_rate": 8e-05,
+      "loss": 1.5891,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2628633820117374,
+      "grad_norm": 0.18622086942195892,
+      "learning_rate": 8e-05,
+      "loss": 1.5305,
+      "step": 1926
+    },
+    {
+      "epoch": 0.26299986351849325,
+      "grad_norm": 0.20018920302391052,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2631363450252491,
+      "grad_norm": 0.20352910459041595,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2632728265320049,
+      "grad_norm": 0.181611105799675,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2634093080387607,
+      "grad_norm": 0.18917249143123627,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2635457895455166,
+      "grad_norm": 0.19118757545948029,
+      "learning_rate": 8e-05,
+      "loss": 1.5691,
+      "step": 1931
+    },
+    {
+      "epoch": 0.26368227105227243,
+      "grad_norm": 0.18285700678825378,
+      "learning_rate": 8e-05,
+      "loss": 1.5444,
+      "step": 1932
+    },
+    {
+      "epoch": 0.26381875255902826,
+      "grad_norm": 0.18428394198417664,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2639552340657841,
+      "grad_norm": 0.18672525882720947,
+      "learning_rate": 8e-05,
+      "loss": 1.5277,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2640917155725399,
+      "grad_norm": 0.1911453753709793,
+      "learning_rate": 8e-05,
+      "loss": 1.4905,
+      "step": 1935
+    },
+    {
+      "epoch": 0.26422819707929573,
+      "grad_norm": 0.18629948794841766,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2643646785860516,
+      "grad_norm": 0.18960405886173248,
+      "learning_rate": 8e-05,
+      "loss": 1.5701,
+      "step": 1937
+    },
+    {
+      "epoch": 0.26450116009280744,
+      "grad_norm": 0.19084347784519196,
+      "learning_rate": 8e-05,
+      "loss": 1.6068,
+      "step": 1938
+    },
+    {
+      "epoch": 0.26463764159956327,
+      "grad_norm": 0.18919622898101807,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2647741231063191,
+      "grad_norm": 0.18236729502677917,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2649106046130749,
+      "grad_norm": 0.18604658544063568,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 1941
+    },
+    {
+      "epoch": 0.26504708611983074,
+      "grad_norm": 0.1972835212945938,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2651835676265866,
+      "grad_norm": 0.18087676167488098,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 1943
+    },
+    {
+      "epoch": 0.26532004913334245,
+      "grad_norm": 0.18288066983222961,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2654565306400983,
+      "grad_norm": 0.19971388578414917,
+      "learning_rate": 8e-05,
+      "loss": 1.5837,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2655930121468541,
+      "grad_norm": 0.19132724404335022,
+      "learning_rate": 8e-05,
+      "loss": 1.5639,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2657294936536099,
+      "grad_norm": 0.18988388776779175,
+      "learning_rate": 8e-05,
+      "loss": 1.5873,
+      "step": 1947
+    },
+    {
+      "epoch": 0.26586597516036575,
+      "grad_norm": 0.18902939558029175,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2660024566671216,
+      "grad_norm": 0.18844576179981232,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 1949
+    },
+    {
+      "epoch": 0.26613893817387746,
+      "grad_norm": 0.1893915832042694,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2662754196806333,
+      "grad_norm": 0.19330920279026031,
+      "learning_rate": 8e-05,
+      "loss": 1.5569,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2664119011873891,
+      "grad_norm": 0.18004919588565826,
+      "learning_rate": 8e-05,
+      "loss": 1.593,
+      "step": 1952
+    },
+    {
+      "epoch": 0.26654838269414494,
+      "grad_norm": 0.18775908648967743,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 1953
+    },
+    {
+      "epoch": 0.26668486420090076,
+      "grad_norm": 0.19320730865001678,
+      "learning_rate": 8e-05,
+      "loss": 1.5666,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2668213457076566,
+      "grad_norm": 0.1863574981689453,
+      "learning_rate": 8e-05,
+      "loss": 1.5345,
+      "step": 1955
+    },
+    {
+      "epoch": 0.26695782721441247,
+      "grad_norm": 0.1857500523328781,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2670943087211683,
+      "grad_norm": 0.18506550788879395,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2672307902279241,
+      "grad_norm": 0.181392103433609,
+      "learning_rate": 8e-05,
+      "loss": 1.4983,
+      "step": 1958
+    },
+    {
+      "epoch": 0.26736727173467995,
+      "grad_norm": 0.19811207056045532,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 1959
+    },
+    {
+      "epoch": 0.26750375324143577,
+      "grad_norm": 0.18951451778411865,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2676402347481916,
+      "grad_norm": 0.18197423219680786,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2677767162549475,
+      "grad_norm": 0.1882587969303131,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2679131977617033,
+      "grad_norm": 0.19073234498500824,
+      "learning_rate": 8e-05,
+      "loss": 1.5673,
+      "step": 1963
+    },
+    {
+      "epoch": 0.26804967926845913,
+      "grad_norm": 0.19055069983005524,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 1964
+    },
+    {
+      "epoch": 0.26818616077521495,
+      "grad_norm": 0.1990155577659607,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2683226422819708,
+      "grad_norm": 0.18501326441764832,
+      "learning_rate": 8e-05,
+      "loss": 1.5013,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2684591237887266,
+      "grad_norm": 0.1913275271654129,
+      "learning_rate": 8e-05,
+      "loss": 1.5332,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2685956052954825,
+      "grad_norm": 0.18184155225753784,
+      "learning_rate": 8e-05,
+      "loss": 1.5156,
+      "step": 1968
+    },
+    {
+      "epoch": 0.2687320868022383,
+      "grad_norm": 0.1934329718351364,
+      "learning_rate": 8e-05,
+      "loss": 1.5777,
+      "step": 1969
+    },
+    {
+      "epoch": 0.26886856830899414,
+      "grad_norm": 0.18897759914398193,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1970
+    },
+    {
+      "epoch": 0.26900504981574996,
+      "grad_norm": 0.18884557485580444,
+      "learning_rate": 8e-05,
+      "loss": 1.5869,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2691415313225058,
+      "grad_norm": 0.1875036358833313,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2692780128292616,
+      "grad_norm": 0.18670396506786346,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2694144943360175,
+      "grad_norm": 0.18138957023620605,
+      "learning_rate": 8e-05,
+      "loss": 1.5251,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2695509758427733,
+      "grad_norm": 0.1885509043931961,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1975
+    },
+    {
+      "epoch": 0.26968745734952915,
+      "grad_norm": 0.187229722738266,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 1976
+    },
+    {
+      "epoch": 0.269823938856285,
+      "grad_norm": 0.19409427046775818,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2699604203630408,
+      "grad_norm": 0.1868712157011032,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2700969018697966,
+      "grad_norm": 0.19214344024658203,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 1979
+    },
+    {
+      "epoch": 0.27023338337655245,
+      "grad_norm": 0.19068008661270142,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1980
+    },
+    {
+      "epoch": 0.27036986488330833,
+      "grad_norm": 0.18215326964855194,
+      "learning_rate": 8e-05,
+      "loss": 1.5185,
+      "step": 1981
+    },
+    {
+      "epoch": 0.27050634639006416,
+      "grad_norm": 0.18147170543670654,
+      "learning_rate": 8e-05,
+      "loss": 1.5206,
+      "step": 1982
+    },
+    {
+      "epoch": 0.27064282789682,
+      "grad_norm": 0.18436045944690704,
+      "learning_rate": 8e-05,
+      "loss": 1.5095,
+      "step": 1983
+    },
+    {
+      "epoch": 0.2707793094035758,
+      "grad_norm": 0.18467599153518677,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 1984
+    },
+    {
+      "epoch": 0.27091579091033163,
+      "grad_norm": 0.18867486715316772,
+      "learning_rate": 8e-05,
+      "loss": 1.5409,
+      "step": 1985
+    },
+    {
+      "epoch": 0.27105227241708746,
+      "grad_norm": 0.18096183240413666,
+      "learning_rate": 8e-05,
+      "loss": 1.5082,
+      "step": 1986
+    },
+    {
+      "epoch": 0.27118875392384334,
+      "grad_norm": 0.1838560700416565,
+      "learning_rate": 8e-05,
+      "loss": 1.5188,
+      "step": 1987
+    },
+    {
+      "epoch": 0.27132523543059917,
+      "grad_norm": 0.1841355562210083,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 1988
+    },
+    {
+      "epoch": 0.271461716937355,
+      "grad_norm": 0.18679086863994598,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 1989
+    },
+    {
+      "epoch": 0.2715981984441108,
+      "grad_norm": 0.18134665489196777,
+      "learning_rate": 8e-05,
+      "loss": 1.5034,
+      "step": 1990
+    },
+    {
+      "epoch": 0.27173467995086664,
+      "grad_norm": 0.18904753029346466,
+      "learning_rate": 8e-05,
+      "loss": 1.5461,
+      "step": 1991
+    },
+    {
+      "epoch": 0.27187116145762247,
+      "grad_norm": 0.18858948349952698,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 1992
+    },
+    {
+      "epoch": 0.27200764296437835,
+      "grad_norm": 0.1885107159614563,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 1993
+    },
+    {
+      "epoch": 0.2721441244711342,
+      "grad_norm": 0.19152052700519562,
+      "learning_rate": 8e-05,
+      "loss": 1.5104,
+      "step": 1994
+    },
+    {
+      "epoch": 0.27228060597789,
+      "grad_norm": 0.18891370296478271,
+      "learning_rate": 8e-05,
+      "loss": 1.5267,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2724170874846458,
+      "grad_norm": 0.19039413332939148,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 1996
+    },
+    {
+      "epoch": 0.27255356899140165,
+      "grad_norm": 0.18724073469638824,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 1997
+    },
+    {
+      "epoch": 0.2726900504981575,
+      "grad_norm": 0.18726858496665955,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 1998
+    },
+    {
+      "epoch": 0.27282653200491336,
+      "grad_norm": 0.17973366379737854,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2729630135116692,
+      "grad_norm": 0.18358615040779114,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 2000
+    },
+    {
+      "epoch": 0.273099495018425,
+      "grad_norm": 0.18650315701961517,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 2001
+    },
+    {
+      "epoch": 0.27323597652518083,
+      "grad_norm": 0.18275026977062225,
+      "learning_rate": 8e-05,
+      "loss": 1.5492,
+      "step": 2002
+    },
+    {
+      "epoch": 0.27337245803193666,
+      "grad_norm": 0.1825461983680725,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2735089395386925,
+      "grad_norm": 0.18888992071151733,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 2004
+    },
+    {
+      "epoch": 0.27364542104544837,
+      "grad_norm": 0.1880376785993576,
+      "learning_rate": 8e-05,
+      "loss": 1.5013,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2737819025522042,
+      "grad_norm": 0.18573841452598572,
+      "learning_rate": 8e-05,
+      "loss": 1.5665,
+      "step": 2006
+    },
+    {
+      "epoch": 0.27391838405896,
+      "grad_norm": 0.18665440380573273,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 2007
+    },
+    {
+      "epoch": 0.27405486556571584,
+      "grad_norm": 0.18374697864055634,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 2008
+    },
+    {
+      "epoch": 0.27419134707247167,
+      "grad_norm": 0.1893589198589325,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2743278285792275,
+      "grad_norm": 0.18848778307437897,
+      "learning_rate": 8e-05,
+      "loss": 1.4803,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2744643100859834,
+      "grad_norm": 0.18755941092967987,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 2011
+    },
+    {
+      "epoch": 0.2746007915927392,
+      "grad_norm": 0.18306218087673187,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 2012
+    },
+    {
+      "epoch": 0.274737273099495,
+      "grad_norm": 0.18841566145420074,
+      "learning_rate": 8e-05,
+      "loss": 1.5486,
+      "step": 2013
+    },
+    {
+      "epoch": 0.27487375460625085,
+      "grad_norm": 0.18870852887630463,
+      "learning_rate": 8e-05,
+      "loss": 1.5111,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2750102361130067,
+      "grad_norm": 0.18545715510845184,
+      "learning_rate": 8e-05,
+      "loss": 1.5179,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2751467176197625,
+      "grad_norm": 0.1913764774799347,
+      "learning_rate": 8e-05,
+      "loss": 1.5866,
+      "step": 2016
+    },
+    {
+      "epoch": 0.27528319912651833,
+      "grad_norm": 0.18971224129199982,
+      "learning_rate": 8e-05,
+      "loss": 1.5129,
+      "step": 2017
+    },
+    {
+      "epoch": 0.2754196806332742,
+      "grad_norm": 0.18147705495357513,
+      "learning_rate": 8e-05,
+      "loss": 1.5084,
+      "step": 2018
+    },
+    {
+      "epoch": 0.27555616214003004,
+      "grad_norm": 0.18673628568649292,
+      "learning_rate": 8e-05,
+      "loss": 1.5013,
+      "step": 2019
+    },
+    {
+      "epoch": 0.27569264364678586,
+      "grad_norm": 0.1802338808774948,
+      "learning_rate": 8e-05,
+      "loss": 1.4617,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2758291251535417,
+      "grad_norm": 0.19209244847297668,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2759656066602975,
+      "grad_norm": 0.20253556966781616,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 2022
+    },
+    {
+      "epoch": 0.27610208816705334,
+      "grad_norm": 0.18339714407920837,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 2023
+    },
+    {
+      "epoch": 0.2762385696738092,
+      "grad_norm": 0.18465369939804077,
+      "learning_rate": 8e-05,
+      "loss": 1.4795,
+      "step": 2024
+    },
+    {
+      "epoch": 0.27637505118056505,
+      "grad_norm": 0.20187842845916748,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2025
+    },
+    {
+      "epoch": 0.27651153268732087,
+      "grad_norm": 0.18822062015533447,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 2026
+    },
+    {
+      "epoch": 0.2766480141940767,
+      "grad_norm": 0.1913299262523651,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 2027
+    },
+    {
+      "epoch": 0.2767844957008325,
+      "grad_norm": 0.19084617495536804,
+      "learning_rate": 8e-05,
+      "loss": 1.571,
+      "step": 2028
+    },
+    {
+      "epoch": 0.27692097720758835,
+      "grad_norm": 0.19963937997817993,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 2029
+    },
+    {
+      "epoch": 0.27705745871434423,
+      "grad_norm": 0.18904776871204376,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 2030
+    },
+    {
+      "epoch": 0.27719394022110005,
+      "grad_norm": 0.1909492164850235,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 2031
+    },
+    {
+      "epoch": 0.2773304217278559,
+      "grad_norm": 0.18875887989997864,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 2032
+    },
+    {
+      "epoch": 0.2774669032346117,
+      "grad_norm": 0.18893785774707794,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 2033
+    },
+    {
+      "epoch": 0.27760338474136753,
+      "grad_norm": 0.18837638199329376,
+      "learning_rate": 8e-05,
+      "loss": 1.5717,
+      "step": 2034
+    },
+    {
+      "epoch": 0.27773986624812336,
+      "grad_norm": 0.19914820790290833,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 2035
+    },
+    {
+      "epoch": 0.27787634775487924,
+      "grad_norm": 0.18605150282382965,
+      "learning_rate": 8e-05,
+      "loss": 1.5508,
+      "step": 2036
+    },
+    {
+      "epoch": 0.27801282926163506,
+      "grad_norm": 0.1822100281715393,
+      "learning_rate": 8e-05,
+      "loss": 1.5239,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2781493107683909,
+      "grad_norm": 0.1860920935869217,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2782857922751467,
+      "grad_norm": 0.18487825989723206,
+      "learning_rate": 8e-05,
+      "loss": 1.5003,
+      "step": 2039
+    },
+    {
+      "epoch": 0.27842227378190254,
+      "grad_norm": 0.21461714804172516,
+      "learning_rate": 8e-05,
+      "loss": 1.5303,
+      "step": 2040
+    },
+    {
+      "epoch": 0.27855875528865837,
+      "grad_norm": 0.1883653998374939,
+      "learning_rate": 8e-05,
+      "loss": 1.6343,
+      "step": 2041
+    },
+    {
+      "epoch": 0.27869523679541425,
+      "grad_norm": 0.18029989302158356,
+      "learning_rate": 8e-05,
+      "loss": 1.4902,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2788317183021701,
+      "grad_norm": 0.1968306452035904,
+      "learning_rate": 8e-05,
+      "loss": 1.5404,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2789681998089259,
+      "grad_norm": 0.1841675192117691,
+      "learning_rate": 8e-05,
+      "loss": 1.5112,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2791046813156817,
+      "grad_norm": 0.18856370449066162,
+      "learning_rate": 8e-05,
+      "loss": 1.4514,
+      "step": 2045
+    },
+    {
+      "epoch": 0.27924116282243755,
+      "grad_norm": 0.19120898842811584,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2793776443291934,
+      "grad_norm": 0.1992325335741043,
+      "learning_rate": 8e-05,
+      "loss": 1.617,
+      "step": 2047
+    },
+    {
+      "epoch": 0.27951412583594926,
+      "grad_norm": 0.1858394593000412,
+      "learning_rate": 8e-05,
+      "loss": 1.562,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2796506073427051,
+      "grad_norm": 0.18444325029850006,
+      "learning_rate": 8e-05,
+      "loss": 1.5448,
+      "step": 2049
+    },
+    {
+      "epoch": 0.2797870888494609,
+      "grad_norm": 0.18600699305534363,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 2050
+    },
+    {
+      "epoch": 0.27992357035621673,
+      "grad_norm": 0.18486641347408295,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 2051
+    },
+    {
+      "epoch": 0.28006005186297256,
+      "grad_norm": 0.18436729907989502,
+      "learning_rate": 8e-05,
+      "loss": 1.5727,
+      "step": 2052
+    },
+    {
+      "epoch": 0.2801965333697284,
+      "grad_norm": 0.18976762890815735,
+      "learning_rate": 8e-05,
+      "loss": 1.584,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2803330148764842,
+      "grad_norm": 0.1834273636341095,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2804694963832401,
+      "grad_norm": 0.1896577775478363,
+      "learning_rate": 8e-05,
+      "loss": 1.5514,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2806059778899959,
+      "grad_norm": 0.19044657051563263,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 2056
+    },
+    {
+      "epoch": 0.28074245939675174,
+      "grad_norm": 0.19076402485370636,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 2057
+    },
+    {
+      "epoch": 0.28087894090350757,
+      "grad_norm": 0.18252645432949066,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 2058
+    },
+    {
+      "epoch": 0.2810154224102634,
+      "grad_norm": 0.18642456829547882,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2811519039170192,
+      "grad_norm": 0.198273703455925,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2812883854237751,
+      "grad_norm": 0.188239187002182,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2814248669305309,
+      "grad_norm": 0.18614868819713593,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2062
+    },
+    {
+      "epoch": 0.28156134843728675,
+      "grad_norm": 0.19303081929683685,
+      "learning_rate": 8e-05,
+      "loss": 1.5311,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2816978299440426,
+      "grad_norm": 0.18951718509197235,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2818343114507984,
+      "grad_norm": 0.19406652450561523,
+      "learning_rate": 8e-05,
+      "loss": 1.6268,
+      "step": 2065
+    },
+    {
+      "epoch": 0.28197079295755423,
+      "grad_norm": 0.19498033821582794,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2821072744643101,
+      "grad_norm": 0.18559187650680542,
+      "learning_rate": 8e-05,
+      "loss": 1.486,
+      "step": 2067
+    },
+    {
+      "epoch": 0.28224375597106593,
+      "grad_norm": 0.18782497942447662,
+      "learning_rate": 8e-05,
+      "loss": 1.4472,
+      "step": 2068
+    },
+    {
+      "epoch": 0.28238023747782176,
+      "grad_norm": 0.1820422261953354,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2825167189845776,
+      "grad_norm": 0.18505264818668365,
+      "learning_rate": 8e-05,
+      "loss": 1.599,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2826532004913334,
+      "grad_norm": 0.18905840814113617,
+      "learning_rate": 8e-05,
+      "loss": 1.5395,
+      "step": 2071
+    },
+    {
+      "epoch": 0.28278968199808924,
+      "grad_norm": 0.19095966219902039,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2829261635048451,
+      "grad_norm": 0.19496643543243408,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 2073
+    },
+    {
+      "epoch": 0.28306264501160094,
+      "grad_norm": 0.19167523086071014,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 2074
+    },
+    {
+      "epoch": 0.28319912651835677,
+      "grad_norm": 0.1842069774866104,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2833356080251126,
+      "grad_norm": 0.19793909788131714,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2834720895318684,
+      "grad_norm": 0.19219326972961426,
+      "learning_rate": 8e-05,
+      "loss": 1.5475,
+      "step": 2077
+    },
+    {
+      "epoch": 0.28360857103862425,
+      "grad_norm": 0.18906018137931824,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2837450525453801,
+      "grad_norm": 0.196896493434906,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 2079
+    },
+    {
+      "epoch": 0.28388153405213595,
+      "grad_norm": 0.1978483647108078,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2840180155588918,
+      "grad_norm": 0.1934979259967804,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2841544970656476,
+      "grad_norm": 0.19172784686088562,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 2082
+    },
+    {
+      "epoch": 0.28429097857240343,
+      "grad_norm": 0.19835710525512695,
+      "learning_rate": 8e-05,
+      "loss": 1.4582,
+      "step": 2083
+    },
+    {
+      "epoch": 0.28442746007915926,
+      "grad_norm": 0.2018197625875473,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2845639415859151,
+      "grad_norm": 0.1912423074245453,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 2085
+    },
+    {
+      "epoch": 0.28470042309267096,
+      "grad_norm": 0.19440267980098724,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2848369045994268,
+      "grad_norm": 0.18047915399074554,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2849733861061826,
+      "grad_norm": 0.2030009776353836,
+      "learning_rate": 8e-05,
+      "loss": 1.5616,
+      "step": 2088
+    },
+    {
+      "epoch": 0.28510986761293844,
+      "grad_norm": 0.1893721967935562,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2089
+    },
+    {
+      "epoch": 0.28524634911969426,
+      "grad_norm": 0.2278795689344406,
+      "learning_rate": 8e-05,
+      "loss": 1.4578,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2853828306264501,
+      "grad_norm": 0.21405383944511414,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 2091
+    },
+    {
+      "epoch": 0.28551931213320597,
+      "grad_norm": 0.1927763819694519,
+      "learning_rate": 8e-05,
+      "loss": 1.5062,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2856557936399618,
+      "grad_norm": 0.20548342168331146,
+      "learning_rate": 8e-05,
+      "loss": 1.5398,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2857922751467176,
+      "grad_norm": 0.193179190158844,
+      "learning_rate": 8e-05,
+      "loss": 1.567,
+      "step": 2094
+    },
+    {
+      "epoch": 0.28592875665347345,
+      "grad_norm": 0.19033104181289673,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2860652381602293,
+      "grad_norm": 0.19091038405895233,
+      "learning_rate": 8e-05,
+      "loss": 1.5134,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2862017196669851,
+      "grad_norm": 0.18988966941833496,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 2097
+    },
+    {
+      "epoch": 0.286338201173741,
+      "grad_norm": 0.193196639418602,
+      "learning_rate": 8e-05,
+      "loss": 1.6583,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2864746826804968,
+      "grad_norm": 0.18785765767097473,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 2099
+    },
+    {
+      "epoch": 0.28661116418725263,
+      "grad_norm": 0.1824287623167038,
+      "learning_rate": 8e-05,
+      "loss": 1.5245,
+      "step": 2100
+    },
+    {
+      "epoch": 0.28674764569400846,
+      "grad_norm": 0.18832935392856598,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2868841272007643,
+      "grad_norm": 0.1972111016511917,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2870206087075201,
+      "grad_norm": 0.18275626003742218,
+      "learning_rate": 8e-05,
+      "loss": 1.5838,
+      "step": 2103
+    },
+    {
+      "epoch": 0.287157090214276,
+      "grad_norm": 0.18729160726070404,
+      "learning_rate": 8e-05,
+      "loss": 1.5443,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2872935717210318,
+      "grad_norm": 0.18493196368217468,
+      "learning_rate": 8e-05,
+      "loss": 1.5415,
+      "step": 2105
+    },
+    {
+      "epoch": 0.28743005322778764,
+      "grad_norm": 0.19562236964702606,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 2106
+    },
+    {
+      "epoch": 0.28756653473454347,
+      "grad_norm": 0.1914881318807602,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2877030162412993,
+      "grad_norm": 0.19250090420246124,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2878394977480551,
+      "grad_norm": 0.19696591794490814,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 2109
+    },
+    {
+      "epoch": 0.287975979254811,
+      "grad_norm": 0.19119641184806824,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2881124607615668,
+      "grad_norm": 0.1956411451101303,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 2111
+    },
+    {
+      "epoch": 0.28824894226832265,
+      "grad_norm": 0.19782638549804688,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2883854237750785,
+      "grad_norm": 0.193250373005867,
+      "learning_rate": 8e-05,
+      "loss": 1.571,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2885219052818343,
+      "grad_norm": 0.19214728474617004,
+      "learning_rate": 8e-05,
+      "loss": 1.5146,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2886583867885901,
+      "grad_norm": 0.19219325482845306,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 2115
+    },
+    {
+      "epoch": 0.288794868295346,
+      "grad_norm": 0.18980087339878082,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 2116
+    },
+    {
+      "epoch": 0.28893134980210183,
+      "grad_norm": 0.19851712882518768,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 2117
+    },
+    {
+      "epoch": 0.28906783130885766,
+      "grad_norm": 0.18184475600719452,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 2118
+    },
+    {
+      "epoch": 0.2892043128156135,
+      "grad_norm": 0.1804375797510147,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2893407943223693,
+      "grad_norm": 0.1868881732225418,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 2120
+    },
+    {
+      "epoch": 0.28947727582912514,
+      "grad_norm": 0.18092136085033417,
+      "learning_rate": 8e-05,
+      "loss": 1.4103,
+      "step": 2121
+    },
+    {
+      "epoch": 0.28961375733588096,
+      "grad_norm": 0.19184724986553192,
+      "learning_rate": 8e-05,
+      "loss": 1.5465,
+      "step": 2122
+    },
+    {
+      "epoch": 0.28975023884263684,
+      "grad_norm": 0.18744532763957977,
+      "learning_rate": 8e-05,
+      "loss": 1.5336,
+      "step": 2123
+    },
+    {
+      "epoch": 0.28988672034939267,
+      "grad_norm": 0.18454964458942413,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2900232018561485,
+      "grad_norm": 0.18746648728847504,
+      "learning_rate": 8e-05,
+      "loss": 1.5787,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2901596833629043,
+      "grad_norm": 0.1891246736049652,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 2126
+    },
+    {
+      "epoch": 0.29029616486966014,
+      "grad_norm": 0.19264520704746246,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 2127
+    },
+    {
+      "epoch": 0.29043264637641597,
+      "grad_norm": 0.1850605607032776,
+      "learning_rate": 8e-05,
+      "loss": 1.5125,
+      "step": 2128
+    },
+    {
+      "epoch": 0.29056912788317185,
+      "grad_norm": 0.18765035271644592,
+      "learning_rate": 8e-05,
+      "loss": 1.4998,
+      "step": 2129
+    },
+    {
+      "epoch": 0.2907056093899277,
+      "grad_norm": 0.18671131134033203,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2908420908966835,
+      "grad_norm": 0.18650102615356445,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 2131
+    },
+    {
+      "epoch": 0.29097857240343933,
+      "grad_norm": 0.1930171400308609,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 2132
+    },
+    {
+      "epoch": 0.29111505391019515,
+      "grad_norm": 0.17904651165008545,
+      "learning_rate": 8e-05,
+      "loss": 1.4884,
+      "step": 2133
+    },
+    {
+      "epoch": 0.291251535416951,
+      "grad_norm": 0.19207796454429626,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 2134
+    },
+    {
+      "epoch": 0.29138801692370686,
+      "grad_norm": 0.1889646202325821,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2915244984304627,
+      "grad_norm": 0.18964365124702454,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 2136
+    },
+    {
+      "epoch": 0.2916609799372185,
+      "grad_norm": 0.19761215150356293,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 2137
+    },
+    {
+      "epoch": 0.29179746144397434,
+      "grad_norm": 0.1873132437467575,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 2138
+    },
+    {
+      "epoch": 0.29193394295073016,
+      "grad_norm": 0.18819749355316162,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 2139
+    },
+    {
+      "epoch": 0.292070424457486,
+      "grad_norm": 0.1918659210205078,
+      "learning_rate": 8e-05,
+      "loss": 1.5342,
+      "step": 2140
+    },
+    {
+      "epoch": 0.29220690596424187,
+      "grad_norm": 0.18802089989185333,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2923433874709977,
+      "grad_norm": 0.19641587138175964,
+      "learning_rate": 8e-05,
+      "loss": 1.6525,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2924798689777535,
+      "grad_norm": 0.18942345678806305,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 2143
+    },
+    {
+      "epoch": 0.29261635048450935,
+      "grad_norm": 0.19834423065185547,
+      "learning_rate": 8e-05,
+      "loss": 1.5201,
+      "step": 2144
+    },
+    {
+      "epoch": 0.29275283199126517,
+      "grad_norm": 0.1920042335987091,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 2145
+    },
+    {
+      "epoch": 0.292889313498021,
+      "grad_norm": 0.19429340958595276,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2930257950047769,
+      "grad_norm": 0.19816112518310547,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 2147
+    },
+    {
+      "epoch": 0.2931622765115327,
+      "grad_norm": 0.1952446848154068,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 2148
+    },
+    {
+      "epoch": 0.29329875801828853,
+      "grad_norm": 0.19394353032112122,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 2149
+    },
+    {
+      "epoch": 0.29343523952504436,
+      "grad_norm": 0.19018422067165375,
+      "learning_rate": 8e-05,
+      "loss": 1.603,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2935717210318002,
+      "grad_norm": 0.20142124593257904,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 2151
+    },
+    {
+      "epoch": 0.293708202538556,
+      "grad_norm": 0.18389956653118134,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2152
+    },
+    {
+      "epoch": 0.29384468404531183,
+      "grad_norm": 0.20680299401283264,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2939811655520677,
+      "grad_norm": 0.1897064596414566,
+      "learning_rate": 8e-05,
+      "loss": 1.5477,
+      "step": 2154
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 0.19357186555862427,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 2155
+    },
+    {
+      "epoch": 0.29425412856557936,
+      "grad_norm": 0.1795782744884491,
+      "learning_rate": 8e-05,
+      "loss": 1.4079,
+      "step": 2156
+    },
+    {
+      "epoch": 0.2943906100723352,
+      "grad_norm": 0.18702712655067444,
+      "learning_rate": 8e-05,
+      "loss": 1.4603,
+      "step": 2157
+    },
+    {
+      "epoch": 0.294527091579091,
+      "grad_norm": 0.20010407269001007,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 2158
+    },
+    {
+      "epoch": 0.29466357308584684,
+      "grad_norm": 0.1988273561000824,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 2159
+    },
+    {
+      "epoch": 0.2948000545926027,
+      "grad_norm": 0.19437240064144135,
+      "learning_rate": 8e-05,
+      "loss": 1.5527,
+      "step": 2160
+    },
+    {
+      "epoch": 0.29493653609935855,
+      "grad_norm": 0.19792622327804565,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2950730176061144,
+      "grad_norm": 0.2051275074481964,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2952094991128702,
+      "grad_norm": 0.18476387858390808,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 2163
+    },
+    {
+      "epoch": 0.295345980619626,
+      "grad_norm": 0.18674908578395844,
+      "learning_rate": 8e-05,
+      "loss": 1.521,
+      "step": 2164
+    },
+    {
+      "epoch": 0.29548246212638185,
+      "grad_norm": 0.19958634674549103,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 2165
+    },
+    {
+      "epoch": 0.29561894363313773,
+      "grad_norm": 0.1926935911178589,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 2166
+    },
+    {
+      "epoch": 0.29575542513989356,
+      "grad_norm": 0.19321037828922272,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2958919066466494,
+      "grad_norm": 0.19959363341331482,
+      "learning_rate": 8e-05,
+      "loss": 1.5102,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2960283881534052,
+      "grad_norm": 0.19045016169548035,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 2169
+    },
+    {
+      "epoch": 0.29616486966016103,
+      "grad_norm": 0.18875819444656372,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 2170
+    },
+    {
+      "epoch": 0.29630135116691686,
+      "grad_norm": 0.19844792783260345,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 2171
+    },
+    {
+      "epoch": 0.29643783267367274,
+      "grad_norm": 0.19257670640945435,
+      "learning_rate": 8e-05,
+      "loss": 1.4932,
+      "step": 2172
+    },
+    {
+      "epoch": 0.29657431418042857,
+      "grad_norm": 0.20057834684848785,
+      "learning_rate": 8e-05,
+      "loss": 1.5459,
+      "step": 2173
+    },
+    {
+      "epoch": 0.2967107956871844,
+      "grad_norm": 0.18506813049316406,
+      "learning_rate": 8e-05,
+      "loss": 1.4849,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2968472771939402,
+      "grad_norm": 0.20173443853855133,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 2175
+    },
+    {
+      "epoch": 0.29698375870069604,
+      "grad_norm": 0.19865339994430542,
+      "learning_rate": 8e-05,
+      "loss": 1.5678,
+      "step": 2176
+    },
+    {
+      "epoch": 0.29712024020745187,
+      "grad_norm": 0.19193463027477264,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 2177
+    },
+    {
+      "epoch": 0.29725672171420775,
+      "grad_norm": 0.19100703299045563,
+      "learning_rate": 8e-05,
+      "loss": 1.4982,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2973932032209636,
+      "grad_norm": 0.18851928412914276,
+      "learning_rate": 8e-05,
+      "loss": 1.4989,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2975296847277194,
+      "grad_norm": 0.19007621705532074,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2976661662344752,
+      "grad_norm": 0.18618464469909668,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 2181
+    },
+    {
+      "epoch": 0.29780264774123105,
+      "grad_norm": 0.19135455787181854,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 2182
+    },
+    {
+      "epoch": 0.2979391292479869,
+      "grad_norm": 0.19191507995128632,
+      "learning_rate": 8e-05,
+      "loss": 1.5291,
+      "step": 2183
+    },
+    {
+      "epoch": 0.29807561075474276,
+      "grad_norm": 0.1900150030851364,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2982120922614986,
+      "grad_norm": 0.18471696972846985,
+      "learning_rate": 8e-05,
+      "loss": 1.4987,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2983485737682544,
+      "grad_norm": 0.18377231061458588,
+      "learning_rate": 8e-05,
+      "loss": 1.5939,
+      "step": 2186
+    },
+    {
+      "epoch": 0.29848505527501024,
+      "grad_norm": 0.18222413957118988,
+      "learning_rate": 8e-05,
+      "loss": 1.5185,
+      "step": 2187
+    },
+    {
+      "epoch": 0.29862153678176606,
+      "grad_norm": 0.18538598716259003,
+      "learning_rate": 8e-05,
+      "loss": 1.5332,
+      "step": 2188
+    },
+    {
+      "epoch": 0.2987580182885219,
+      "grad_norm": 0.18902656435966492,
+      "learning_rate": 8e-05,
+      "loss": 1.512,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2988944997952777,
+      "grad_norm": 0.19450733065605164,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2990309813020336,
+      "grad_norm": 0.19497349858283997,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2991674628087894,
+      "grad_norm": 0.1885165125131607,
+      "learning_rate": 8e-05,
+      "loss": 1.573,
+      "step": 2192
+    },
+    {
+      "epoch": 0.29930394431554525,
+      "grad_norm": 0.18749703466892242,
+      "learning_rate": 8e-05,
+      "loss": 1.5459,
+      "step": 2193
+    },
+    {
+      "epoch": 0.29944042582230107,
+      "grad_norm": 0.19850268959999084,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 2194
+    },
+    {
+      "epoch": 0.2995769073290569,
+      "grad_norm": 0.19352257251739502,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 2195
+    },
+    {
+      "epoch": 0.2997133888358127,
+      "grad_norm": 0.1824529618024826,
+      "learning_rate": 8e-05,
+      "loss": 1.5188,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2998498703425686,
+      "grad_norm": 0.19440414011478424,
+      "learning_rate": 8e-05,
+      "loss": 1.5499,
+      "step": 2197
+    },
+    {
+      "epoch": 0.29998635184932443,
+      "grad_norm": 0.2017410844564438,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 2198
+    },
+    {
+      "epoch": 0.30012283335608025,
+      "grad_norm": 0.18070609867572784,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3002593148628361,
+      "grad_norm": 0.1974911242723465,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3003957963695919,
+      "grad_norm": 0.18513643741607666,
+      "learning_rate": 8e-05,
+      "loss": 1.5047,
+      "step": 2201
+    },
+    {
+      "epoch": 0.30053227787634773,
+      "grad_norm": 0.1845792531967163,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3006687593831036,
+      "grad_norm": 0.19717486202716827,
+      "learning_rate": 8e-05,
+      "loss": 1.5362,
+      "step": 2203
+    },
+    {
+      "epoch": 0.30080524088985944,
+      "grad_norm": 0.18709112703800201,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 2204
+    },
+    {
+      "epoch": 0.30094172239661526,
+      "grad_norm": 0.19449973106384277,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3010782039033711,
+      "grad_norm": 0.1893892139196396,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3012146854101269,
+      "grad_norm": 0.18763096630573273,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 2207
+    },
+    {
+      "epoch": 0.30135116691688274,
+      "grad_norm": 0.20694661140441895,
+      "learning_rate": 8e-05,
+      "loss": 1.5444,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3014876484236386,
+      "grad_norm": 0.1910211741924286,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 2209
+    },
+    {
+      "epoch": 0.30162412993039445,
+      "grad_norm": 0.1971670240163803,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3017606114371503,
+      "grad_norm": 0.19217047095298767,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3018970929439061,
+      "grad_norm": 0.18745028972625732,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3020335744506619,
+      "grad_norm": 0.18405196070671082,
+      "learning_rate": 8e-05,
+      "loss": 1.4992,
+      "step": 2213
+    },
+    {
+      "epoch": 0.30217005595741775,
+      "grad_norm": 0.18875133991241455,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 2214
+    },
+    {
+      "epoch": 0.30230653746417363,
+      "grad_norm": 0.1918160319328308,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 2215
+    },
+    {
+      "epoch": 0.30244301897092946,
+      "grad_norm": 0.18814332783222198,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3025795004776853,
+      "grad_norm": 0.1866637021303177,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3027159819844411,
+      "grad_norm": 0.18924444913864136,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 2218
+    },
+    {
+      "epoch": 0.30285246349119693,
+      "grad_norm": 0.19841592013835907,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 2219
+    },
+    {
+      "epoch": 0.30298894499795276,
+      "grad_norm": 0.19458287954330444,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 2220
+    },
+    {
+      "epoch": 0.30312542650470864,
+      "grad_norm": 0.1910163015127182,
+      "learning_rate": 8e-05,
+      "loss": 1.5414,
+      "step": 2221
+    },
+    {
+      "epoch": 0.30326190801146446,
+      "grad_norm": 0.19162598252296448,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3033983895182203,
+      "grad_norm": 0.17796890437602997,
+      "learning_rate": 8e-05,
+      "loss": 1.44,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3035348710249761,
+      "grad_norm": 0.1886792629957199,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 2224
+    },
+    {
+      "epoch": 0.30367135253173194,
+      "grad_norm": 0.18323218822479248,
+      "learning_rate": 8e-05,
+      "loss": 1.4889,
+      "step": 2225
+    },
+    {
+      "epoch": 0.30380783403848777,
+      "grad_norm": 0.18252995610237122,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3039443155452436,
+      "grad_norm": 0.19099868834018707,
+      "learning_rate": 8e-05,
+      "loss": 1.5952,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3040807970519995,
+      "grad_norm": 0.19717931747436523,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3042172785587553,
+      "grad_norm": 0.18331508338451385,
+      "learning_rate": 8e-05,
+      "loss": 1.5418,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3043537600655111,
+      "grad_norm": 0.18861477077007294,
+      "learning_rate": 8e-05,
+      "loss": 1.5587,
+      "step": 2230
+    },
+    {
+      "epoch": 0.30449024157226695,
+      "grad_norm": 0.18971489369869232,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 2231
+    },
+    {
+      "epoch": 0.3046267230790228,
+      "grad_norm": 0.18543824553489685,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3047632045857786,
+      "grad_norm": 0.19533771276474,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3048996860925345,
+      "grad_norm": 0.18951739370822906,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3050361675992903,
+      "grad_norm": 0.19273875653743744,
+      "learning_rate": 8e-05,
+      "loss": 1.4793,
+      "step": 2235
+    },
+    {
+      "epoch": 0.30517264910604613,
+      "grad_norm": 0.18404525518417358,
+      "learning_rate": 8e-05,
+      "loss": 1.4385,
+      "step": 2236
+    },
+    {
+      "epoch": 0.30530913061280196,
+      "grad_norm": 0.18486222624778748,
+      "learning_rate": 8e-05,
+      "loss": 1.5183,
+      "step": 2237
+    },
+    {
+      "epoch": 0.3054456121195578,
+      "grad_norm": 0.19355586171150208,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3055820936263136,
+      "grad_norm": 0.19077162444591522,
+      "learning_rate": 8e-05,
+      "loss": 1.5393,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3057185751330695,
+      "grad_norm": 0.19492536783218384,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3058550566398253,
+      "grad_norm": 0.1930067539215088,
+      "learning_rate": 8e-05,
+      "loss": 1.5448,
+      "step": 2241
+    },
+    {
+      "epoch": 0.30599153814658114,
+      "grad_norm": 0.18975308537483215,
+      "learning_rate": 8e-05,
+      "loss": 1.4713,
+      "step": 2242
+    },
+    {
+      "epoch": 0.30612801965333697,
+      "grad_norm": 0.1962203085422516,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 2243
+    },
+    {
+      "epoch": 0.3062645011600928,
+      "grad_norm": 0.1905398666858673,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3064009826668486,
+      "grad_norm": 0.1920812577009201,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3065374641736045,
+      "grad_norm": 0.18745557963848114,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3066739456803603,
+      "grad_norm": 0.18615038692951202,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 2247
+    },
+    {
+      "epoch": 0.30681042718711615,
+      "grad_norm": 0.19473129510879517,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 2248
+    },
+    {
+      "epoch": 0.306946908693872,
+      "grad_norm": 0.19581641256809235,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3070833902006278,
+      "grad_norm": 0.19018195569515228,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 2250
+    },
+    {
+      "epoch": 0.30721987170738363,
+      "grad_norm": 0.1852266788482666,
+      "learning_rate": 8e-05,
+      "loss": 1.4906,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3073563532141395,
+      "grad_norm": 0.19332687556743622,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 2252
+    },
+    {
+      "epoch": 0.30749283472089534,
+      "grad_norm": 0.1825764924287796,
+      "learning_rate": 8e-05,
+      "loss": 1.4772,
+      "step": 2253
+    },
+    {
+      "epoch": 0.30762931622765116,
+      "grad_norm": 0.18047472834587097,
+      "learning_rate": 8e-05,
+      "loss": 1.473,
+      "step": 2254
+    },
+    {
+      "epoch": 0.307765797734407,
+      "grad_norm": 0.1946837157011032,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3079022792411628,
+      "grad_norm": 0.18675142526626587,
+      "learning_rate": 8e-05,
+      "loss": 1.5061,
+      "step": 2256
+    },
+    {
+      "epoch": 0.30803876074791864,
+      "grad_norm": 0.19220958650112152,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 2257
+    },
+    {
+      "epoch": 0.30817524225467446,
+      "grad_norm": 0.1952037811279297,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 2258
+    },
+    {
+      "epoch": 0.30831172376143035,
+      "grad_norm": 0.19373290240764618,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 2259
+    },
+    {
+      "epoch": 0.30844820526818617,
+      "grad_norm": 0.18846887350082397,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 2260
+    },
+    {
+      "epoch": 0.308584686774942,
+      "grad_norm": 0.19393399357795715,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3087211682816978,
+      "grad_norm": 0.1933368295431137,
+      "learning_rate": 8e-05,
+      "loss": 1.5388,
+      "step": 2262
+    },
+    {
+      "epoch": 0.30885764978845365,
+      "grad_norm": 0.1938529908657074,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 2263
+    },
+    {
+      "epoch": 0.3089941312952095,
+      "grad_norm": 0.19152048230171204,
+      "learning_rate": 8e-05,
+      "loss": 1.5858,
+      "step": 2264
+    },
+    {
+      "epoch": 0.30913061280196535,
+      "grad_norm": 0.19202856719493866,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3092670943087212,
+      "grad_norm": 0.18733921647071838,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 2266
+    },
+    {
+      "epoch": 0.309403575815477,
+      "grad_norm": 0.1896551251411438,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 2267
+    },
+    {
+      "epoch": 0.30954005732223283,
+      "grad_norm": 0.19032728672027588,
+      "learning_rate": 8e-05,
+      "loss": 1.5213,
+      "step": 2268
+    },
+    {
+      "epoch": 0.30967653882898866,
+      "grad_norm": 0.19229768216609955,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3098130203357445,
+      "grad_norm": 0.19707682728767395,
+      "learning_rate": 8e-05,
+      "loss": 1.5564,
+      "step": 2270
+    },
+    {
+      "epoch": 0.30994950184250036,
+      "grad_norm": 0.19649268686771393,
+      "learning_rate": 8e-05,
+      "loss": 1.6042,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3100859833492562,
+      "grad_norm": 0.18773022294044495,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 2272
+    },
+    {
+      "epoch": 0.310222464856012,
+      "grad_norm": 0.1910863220691681,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 2273
+    },
+    {
+      "epoch": 0.31035894636276784,
+      "grad_norm": 0.18872302770614624,
+      "learning_rate": 8e-05,
+      "loss": 1.5082,
+      "step": 2274
+    },
+    {
+      "epoch": 0.31049542786952367,
+      "grad_norm": 0.19204914569854736,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3106319093762795,
+      "grad_norm": 0.18933150172233582,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3107683908830354,
+      "grad_norm": 0.18659311532974243,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3109048723897912,
+      "grad_norm": 0.1954302191734314,
+      "learning_rate": 8e-05,
+      "loss": 1.5407,
+      "step": 2278
+    },
+    {
+      "epoch": 0.311041353896547,
+      "grad_norm": 0.19257852435112,
+      "learning_rate": 8e-05,
+      "loss": 1.5342,
+      "step": 2279
+    },
+    {
+      "epoch": 0.31117783540330285,
+      "grad_norm": 0.19051837921142578,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3113143169100587,
+      "grad_norm": 0.1838407814502716,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3114507984168145,
+      "grad_norm": 0.19048361480236053,
+      "learning_rate": 8e-05,
+      "loss": 1.518,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3115872799235704,
+      "grad_norm": 0.19328182935714722,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3117237614303262,
+      "grad_norm": 0.1882355958223343,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 2284
+    },
+    {
+      "epoch": 0.31186024293708203,
+      "grad_norm": 0.1923254132270813,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 2285
+    },
+    {
+      "epoch": 0.31199672444383786,
+      "grad_norm": 0.18768557906150818,
+      "learning_rate": 8e-05,
+      "loss": 1.4729,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3121332059505937,
+      "grad_norm": 0.18628625571727753,
+      "learning_rate": 8e-05,
+      "loss": 1.5578,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3122696874573495,
+      "grad_norm": 0.18642590939998627,
+      "learning_rate": 8e-05,
+      "loss": 1.5337,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3124061689641054,
+      "grad_norm": 0.19136059284210205,
+      "learning_rate": 8e-05,
+      "loss": 1.5309,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3125426504708612,
+      "grad_norm": 0.18829867243766785,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 2290
+    },
+    {
+      "epoch": 0.31267913197761704,
+      "grad_norm": 0.18478061258792877,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 2291
+    },
+    {
+      "epoch": 0.31281561348437287,
+      "grad_norm": 0.18714046478271484,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3129520949911287,
+      "grad_norm": 0.19507615268230438,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3130885764978845,
+      "grad_norm": 0.18988661468029022,
+      "learning_rate": 8e-05,
+      "loss": 1.5856,
+      "step": 2294
+    },
+    {
+      "epoch": 0.31322505800464034,
+      "grad_norm": 0.19226372241973877,
+      "learning_rate": 8e-05,
+      "loss": 1.5508,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3133615395113962,
+      "grad_norm": 0.19245581328868866,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 2296
+    },
+    {
+      "epoch": 0.31349802101815205,
+      "grad_norm": 0.18920768797397614,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3136345025249079,
+      "grad_norm": 0.19054719805717468,
+      "learning_rate": 8e-05,
+      "loss": 1.5213,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3137709840316637,
+      "grad_norm": 0.18515238165855408,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 2299
+    },
+    {
+      "epoch": 0.31390746553841953,
+      "grad_norm": 0.18925321102142334,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 2300
+    },
+    {
+      "epoch": 0.31404394704517535,
+      "grad_norm": 0.19498296082019806,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 2301
+    },
+    {
+      "epoch": 0.31418042855193123,
+      "grad_norm": 0.19261011481285095,
+      "learning_rate": 8e-05,
+      "loss": 1.539,
+      "step": 2302
+    },
+    {
+      "epoch": 0.31431691005868706,
+      "grad_norm": 0.19898930191993713,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3144533915654429,
+      "grad_norm": 0.1882028877735138,
+      "learning_rate": 8e-05,
+      "loss": 1.4938,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3145898730721987,
+      "grad_norm": 0.2005249708890915,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 2305
+    },
+    {
+      "epoch": 0.31472635457895454,
+      "grad_norm": 0.1920951008796692,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 2306
+    },
+    {
+      "epoch": 0.31486283608571036,
+      "grad_norm": 0.19342833757400513,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 2307
+    },
+    {
+      "epoch": 0.31499931759246624,
+      "grad_norm": 0.19189022481441498,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 2308
+    },
+    {
+      "epoch": 0.31513579909922207,
+      "grad_norm": 0.18880152702331543,
+      "learning_rate": 8e-05,
+      "loss": 1.5164,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3152722806059779,
+      "grad_norm": 0.18937347829341888,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3154087621127337,
+      "grad_norm": 0.1873825192451477,
+      "learning_rate": 8e-05,
+      "loss": 1.5834,
+      "step": 2311
+    },
+    {
+      "epoch": 0.31554524361948955,
+      "grad_norm": 0.19152410328388214,
+      "learning_rate": 8e-05,
+      "loss": 1.4531,
+      "step": 2312
+    },
+    {
+      "epoch": 0.31568172512624537,
+      "grad_norm": 0.18692977726459503,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 2313
+    },
+    {
+      "epoch": 0.31581820663300125,
+      "grad_norm": 0.1933252066373825,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3159546881397571,
+      "grad_norm": 0.194474995136261,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3160911696465129,
+      "grad_norm": 0.19405238330364227,
+      "learning_rate": 8e-05,
+      "loss": 1.5608,
+      "step": 2316
+    },
+    {
+      "epoch": 0.31622765115326873,
+      "grad_norm": 0.1938561648130417,
+      "learning_rate": 8e-05,
+      "loss": 1.537,
+      "step": 2317
+    },
+    {
+      "epoch": 0.31636413266002456,
+      "grad_norm": 0.19599539041519165,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3165006141667804,
+      "grad_norm": 0.18618819117546082,
+      "learning_rate": 8e-05,
+      "loss": 1.4821,
+      "step": 2319
+    },
+    {
+      "epoch": 0.31663709567353626,
+      "grad_norm": 0.20246824622154236,
+      "learning_rate": 8e-05,
+      "loss": 1.5812,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3167735771802921,
+      "grad_norm": 0.19695766270160675,
+      "learning_rate": 8e-05,
+      "loss": 1.5596,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3169100586870479,
+      "grad_norm": 0.19561611115932465,
+      "learning_rate": 8e-05,
+      "loss": 1.528,
+      "step": 2322
+    },
+    {
+      "epoch": 0.31704654019380374,
+      "grad_norm": 0.1915489286184311,
+      "learning_rate": 8e-05,
+      "loss": 1.6143,
+      "step": 2323
+    },
+    {
+      "epoch": 0.31718302170055956,
+      "grad_norm": 0.18535694479942322,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3173195032073154,
+      "grad_norm": 0.19556941092014313,
+      "learning_rate": 8e-05,
+      "loss": 1.5239,
+      "step": 2325
+    },
+    {
+      "epoch": 0.31745598471407127,
+      "grad_norm": 0.1962941735982895,
+      "learning_rate": 8e-05,
+      "loss": 1.5594,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3175924662208271,
+      "grad_norm": 0.19477158784866333,
+      "learning_rate": 8e-05,
+      "loss": 1.5218,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3177289477275829,
+      "grad_norm": 0.1849786788225174,
+      "learning_rate": 8e-05,
+      "loss": 1.4542,
+      "step": 2328
+    },
+    {
+      "epoch": 0.31786542923433875,
+      "grad_norm": 0.18511530756950378,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3180019107410946,
+      "grad_norm": 0.19234132766723633,
+      "learning_rate": 8e-05,
+      "loss": 1.5254,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3181383922478504,
+      "grad_norm": 0.20037028193473816,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3182748737546062,
+      "grad_norm": 0.20021872222423553,
+      "learning_rate": 8e-05,
+      "loss": 1.5144,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3184113552613621,
+      "grad_norm": 0.19810526072978973,
+      "learning_rate": 8e-05,
+      "loss": 1.5548,
+      "step": 2333
+    },
+    {
+      "epoch": 0.31854783676811793,
+      "grad_norm": 0.19990825653076172,
+      "learning_rate": 8e-05,
+      "loss": 1.5508,
+      "step": 2334
+    },
+    {
+      "epoch": 0.31868431827487376,
+      "grad_norm": 0.19812963902950287,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3188207997816296,
+      "grad_norm": 0.19568397104740143,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3189572812883854,
+      "grad_norm": 0.19830144941806793,
+      "learning_rate": 8e-05,
+      "loss": 1.5682,
+      "step": 2337
+    },
+    {
+      "epoch": 0.31909376279514123,
+      "grad_norm": 0.19387203454971313,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3192302443018971,
+      "grad_norm": 0.19514256715774536,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 2339
+    },
+    {
+      "epoch": 0.31936672580865294,
+      "grad_norm": 0.19740474224090576,
+      "learning_rate": 8e-05,
+      "loss": 1.5362,
+      "step": 2340
+    },
+    {
+      "epoch": 0.31950320731540877,
+      "grad_norm": 0.19206172227859497,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3196396888221646,
+      "grad_norm": 0.18451574444770813,
+      "learning_rate": 8e-05,
+      "loss": 1.542,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3197761703289204,
+      "grad_norm": 0.19501104950904846,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 2343
+    },
+    {
+      "epoch": 0.31991265183567624,
+      "grad_norm": 0.1942450851202011,
+      "learning_rate": 8e-05,
+      "loss": 1.5689,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3200491333424321,
+      "grad_norm": 0.18815119564533234,
+      "learning_rate": 8e-05,
+      "loss": 1.5291,
+      "step": 2345
+    },
+    {
+      "epoch": 0.32018561484918795,
+      "grad_norm": 0.20140774548053741,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3203220963559438,
+      "grad_norm": 0.2070215940475464,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3204585778626996,
+      "grad_norm": 0.1885538399219513,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3205950593694554,
+      "grad_norm": 0.19860891997814178,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 2349
+    },
+    {
+      "epoch": 0.32073154087621125,
+      "grad_norm": 0.20438644289970398,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 2350
+    },
+    {
+      "epoch": 0.32086802238296713,
+      "grad_norm": 0.1986912339925766,
+      "learning_rate": 8e-05,
+      "loss": 1.5103,
+      "step": 2351
+    },
+    {
+      "epoch": 0.32100450388972296,
+      "grad_norm": 0.20232762396335602,
+      "learning_rate": 8e-05,
+      "loss": 1.5187,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3211409853964788,
+      "grad_norm": 0.20045018196105957,
+      "learning_rate": 8e-05,
+      "loss": 1.5956,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3212774669032346,
+      "grad_norm": 0.19569602608680725,
+      "learning_rate": 8e-05,
+      "loss": 1.5431,
+      "step": 2354
+    },
+    {
+      "epoch": 0.32141394840999044,
+      "grad_norm": 0.1832921952009201,
+      "learning_rate": 8e-05,
+      "loss": 1.4396,
+      "step": 2355
+    },
+    {
+      "epoch": 0.32155042991674626,
+      "grad_norm": 0.19435308873653412,
+      "learning_rate": 8e-05,
+      "loss": 1.5759,
+      "step": 2356
+    },
+    {
+      "epoch": 0.32168691142350214,
+      "grad_norm": 0.205319344997406,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 2357
+    },
+    {
+      "epoch": 0.32182339293025797,
+      "grad_norm": 0.19843775033950806,
+      "learning_rate": 8e-05,
+      "loss": 1.5019,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3219598744370138,
+      "grad_norm": 0.18857765197753906,
+      "learning_rate": 8e-05,
+      "loss": 1.469,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3220963559437696,
+      "grad_norm": 0.19354255497455597,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 2360
+    },
+    {
+      "epoch": 0.32223283745052544,
+      "grad_norm": 0.19421203434467316,
+      "learning_rate": 8e-05,
+      "loss": 1.5682,
+      "step": 2361
+    },
+    {
+      "epoch": 0.32236931895728127,
+      "grad_norm": 0.19062180817127228,
+      "learning_rate": 8e-05,
+      "loss": 1.5688,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3225058004640371,
+      "grad_norm": 0.19428598880767822,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 2363
+    },
+    {
+      "epoch": 0.322642281970793,
+      "grad_norm": 0.19765454530715942,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3227787634775488,
+      "grad_norm": 0.19465826451778412,
+      "learning_rate": 8e-05,
+      "loss": 1.5461,
+      "step": 2365
+    },
+    {
+      "epoch": 0.32291524498430463,
+      "grad_norm": 0.1933916211128235,
+      "learning_rate": 8e-05,
+      "loss": 1.5787,
+      "step": 2366
+    },
+    {
+      "epoch": 0.32305172649106045,
+      "grad_norm": 0.1956445574760437,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3231882079978163,
+      "grad_norm": 0.1907869130373001,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3233246895045721,
+      "grad_norm": 0.1929887980222702,
+      "learning_rate": 8e-05,
+      "loss": 1.6289,
+      "step": 2369
+    },
+    {
+      "epoch": 0.323461171011328,
+      "grad_norm": 0.1880393922328949,
+      "learning_rate": 8e-05,
+      "loss": 1.4537,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3235976525180838,
+      "grad_norm": 0.1917516589164734,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 2371
+    },
+    {
+      "epoch": 0.32373413402483964,
+      "grad_norm": 0.19497986137866974,
+      "learning_rate": 8e-05,
+      "loss": 1.5737,
+      "step": 2372
+    },
+    {
+      "epoch": 0.32387061553159546,
+      "grad_norm": 0.1887574940919876,
+      "learning_rate": 8e-05,
+      "loss": 1.5907,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3240070970383513,
+      "grad_norm": 0.1917102038860321,
+      "learning_rate": 8e-05,
+      "loss": 1.5492,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3241435785451071,
+      "grad_norm": 0.1880488395690918,
+      "learning_rate": 8e-05,
+      "loss": 1.4248,
+      "step": 2375
+    },
+    {
+      "epoch": 0.324280060051863,
+      "grad_norm": 0.1989101618528366,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3244165415586188,
+      "grad_norm": 0.18295790255069733,
+      "learning_rate": 8e-05,
+      "loss": 1.4477,
+      "step": 2377
+    },
+    {
+      "epoch": 0.32455302306537465,
+      "grad_norm": 0.20745578408241272,
+      "learning_rate": 8e-05,
+      "loss": 1.512,
+      "step": 2378
+    },
+    {
+      "epoch": 0.32468950457213047,
+      "grad_norm": 0.19378052651882172,
+      "learning_rate": 8e-05,
+      "loss": 1.4894,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3248259860788863,
+      "grad_norm": 0.19062446057796478,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3249624675856421,
+      "grad_norm": 0.19509540498256683,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 2381
+    },
+    {
+      "epoch": 0.325098949092398,
+      "grad_norm": 0.18596932291984558,
+      "learning_rate": 8e-05,
+      "loss": 1.5445,
+      "step": 2382
+    },
+    {
+      "epoch": 0.32523543059915383,
+      "grad_norm": 0.18645717203617096,
+      "learning_rate": 8e-05,
+      "loss": 1.5341,
+      "step": 2383
+    },
+    {
+      "epoch": 0.32537191210590966,
+      "grad_norm": 0.19618059694766998,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3255083936126655,
+      "grad_norm": 0.19547946751117706,
+      "learning_rate": 8e-05,
+      "loss": 1.5788,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3256448751194213,
+      "grad_norm": 0.19641852378845215,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 2386
+    },
+    {
+      "epoch": 0.32578135662617713,
+      "grad_norm": 0.2065533697605133,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 2387
+    },
+    {
+      "epoch": 0.325917838132933,
+      "grad_norm": 0.1875258833169937,
+      "learning_rate": 8e-05,
+      "loss": 1.4827,
+      "step": 2388
+    },
+    {
+      "epoch": 0.32605431963968884,
+      "grad_norm": 0.19471386075019836,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 2389
+    },
+    {
+      "epoch": 0.32619080114644466,
+      "grad_norm": 0.20052587985992432,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3263272826532005,
+      "grad_norm": 0.188873291015625,
+      "learning_rate": 8e-05,
+      "loss": 1.511,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3264637641599563,
+      "grad_norm": 0.19497057795524597,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 2392
+    },
+    {
+      "epoch": 0.32660024566671214,
+      "grad_norm": 0.19966381788253784,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 2393
+    },
+    {
+      "epoch": 0.326736727173468,
+      "grad_norm": 0.1950003057718277,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 2394
+    },
+    {
+      "epoch": 0.32687320868022385,
+      "grad_norm": 0.20484621822834015,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3270096901869797,
+      "grad_norm": 0.19592083990573883,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 2396
+    },
+    {
+      "epoch": 0.3271461716937355,
+      "grad_norm": 0.18600867688655853,
+      "learning_rate": 8e-05,
+      "loss": 1.4156,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3272826532004913,
+      "grad_norm": 0.19322341680526733,
+      "learning_rate": 8e-05,
+      "loss": 1.5328,
+      "step": 2398
+    },
+    {
+      "epoch": 0.32741913470724715,
+      "grad_norm": 0.19816844165325165,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 2399
+    },
+    {
+      "epoch": 0.327555616214003,
+      "grad_norm": 0.19557923078536987,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 2400
+    },
+    {
+      "epoch": 0.32769209772075886,
+      "grad_norm": 0.19665782153606415,
+      "learning_rate": 8e-05,
+      "loss": 1.5693,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3278285792275147,
+      "grad_norm": 0.19420155882835388,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3279650607342705,
+      "grad_norm": 0.19276310503482819,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 2403
+    },
+    {
+      "epoch": 0.32810154224102633,
+      "grad_norm": 0.19227337837219238,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 2404
+    },
+    {
+      "epoch": 0.32823802374778216,
+      "grad_norm": 0.190183624625206,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 2405
+    },
+    {
+      "epoch": 0.328374505254538,
+      "grad_norm": 0.18371525406837463,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 2406
+    },
+    {
+      "epoch": 0.32851098676129387,
+      "grad_norm": 0.19337286055088043,
+      "learning_rate": 8e-05,
+      "loss": 1.4853,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3286474682680497,
+      "grad_norm": 0.19246435165405273,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3287839497748055,
+      "grad_norm": 0.19278155267238617,
+      "learning_rate": 8e-05,
+      "loss": 1.6183,
+      "step": 2409
+    },
+    {
+      "epoch": 0.32892043128156134,
+      "grad_norm": 0.1896626204252243,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 2410
+    },
+    {
+      "epoch": 0.32905691278831717,
+      "grad_norm": 0.19147953391075134,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 2411
+    },
+    {
+      "epoch": 0.329193394295073,
+      "grad_norm": 0.18816566467285156,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3293298758018289,
+      "grad_norm": 0.19373339414596558,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3294663573085847,
+      "grad_norm": 0.19565635919570923,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 2414
+    },
+    {
+      "epoch": 0.3296028388153405,
+      "grad_norm": 0.18949222564697266,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 2415
+    },
+    {
+      "epoch": 0.32973932032209635,
+      "grad_norm": 0.18624359369277954,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3298758018288522,
+      "grad_norm": 0.1924353688955307,
+      "learning_rate": 8e-05,
+      "loss": 1.6101,
+      "step": 2417
+    },
+    {
+      "epoch": 0.330012283335608,
+      "grad_norm": 0.18345238268375397,
+      "learning_rate": 8e-05,
+      "loss": 1.5016,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3301487648423639,
+      "grad_norm": 0.19848251342773438,
+      "learning_rate": 8e-05,
+      "loss": 1.5512,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3302852463491197,
+      "grad_norm": 0.1907501220703125,
+      "learning_rate": 8e-05,
+      "loss": 1.5866,
+      "step": 2420
+    },
+    {
+      "epoch": 0.33042172785587554,
+      "grad_norm": 0.18231217563152313,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 2421
+    },
+    {
+      "epoch": 0.33055820936263136,
+      "grad_norm": 0.1831728219985962,
+      "learning_rate": 8e-05,
+      "loss": 1.4766,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3306946908693872,
+      "grad_norm": 0.19155003130435944,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 2423
+    },
+    {
+      "epoch": 0.330831172376143,
+      "grad_norm": 0.19622866809368134,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3309676538828989,
+      "grad_norm": 0.18896323442459106,
+      "learning_rate": 8e-05,
+      "loss": 1.5218,
+      "step": 2425
+    },
+    {
+      "epoch": 0.3311041353896547,
+      "grad_norm": 0.1942889243364334,
+      "learning_rate": 8e-05,
+      "loss": 1.56,
+      "step": 2426
+    },
+    {
+      "epoch": 0.33124061689641054,
+      "grad_norm": 0.19754470884799957,
+      "learning_rate": 8e-05,
+      "loss": 1.5717,
+      "step": 2427
+    },
+    {
+      "epoch": 0.33137709840316637,
+      "grad_norm": 0.19484834372997284,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 2428
+    },
+    {
+      "epoch": 0.3315135799099222,
+      "grad_norm": 0.19175194203853607,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 2429
+    },
+    {
+      "epoch": 0.331650061416678,
+      "grad_norm": 0.20060524344444275,
+      "learning_rate": 8e-05,
+      "loss": 1.6009,
+      "step": 2430
+    },
+    {
+      "epoch": 0.33178654292343385,
+      "grad_norm": 0.19135399162769318,
+      "learning_rate": 8e-05,
+      "loss": 1.5493,
+      "step": 2431
+    },
+    {
+      "epoch": 0.33192302443018973,
+      "grad_norm": 0.1867528259754181,
+      "learning_rate": 8e-05,
+      "loss": 1.5305,
+      "step": 2432
+    },
+    {
+      "epoch": 0.33205950593694555,
+      "grad_norm": 0.19375182688236237,
+      "learning_rate": 8e-05,
+      "loss": 1.5204,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3321959874437014,
+      "grad_norm": 0.18884830176830292,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3323324689504572,
+      "grad_norm": 0.19130641222000122,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 2435
+    },
+    {
+      "epoch": 0.33246895045721303,
+      "grad_norm": 0.18871554732322693,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 2436
+    },
+    {
+      "epoch": 0.33260543196396886,
+      "grad_norm": 0.19300872087478638,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 2437
+    },
+    {
+      "epoch": 0.33274191347072474,
+      "grad_norm": 0.19003355503082275,
+      "learning_rate": 8e-05,
+      "loss": 1.5119,
+      "step": 2438
+    },
+    {
+      "epoch": 0.33287839497748056,
+      "grad_norm": 0.18634119629859924,
+      "learning_rate": 8e-05,
+      "loss": 1.4546,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3330148764842364,
+      "grad_norm": 0.1876678168773651,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3331513579909922,
+      "grad_norm": 0.19578248262405396,
+      "learning_rate": 8e-05,
+      "loss": 1.4987,
+      "step": 2441
+    },
+    {
+      "epoch": 0.33328783949774804,
+      "grad_norm": 0.192447230219841,
+      "learning_rate": 8e-05,
+      "loss": 1.4824,
+      "step": 2442
+    },
+    {
+      "epoch": 0.33342432100450387,
+      "grad_norm": 0.19464170932769775,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 2443
+    },
+    {
+      "epoch": 0.33356080251125975,
+      "grad_norm": 0.20000219345092773,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 2444
+    },
+    {
+      "epoch": 0.33369728401801557,
+      "grad_norm": 0.19101779162883759,
+      "learning_rate": 8e-05,
+      "loss": 1.4719,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3338337655247714,
+      "grad_norm": 0.19114328920841217,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3339702470315272,
+      "grad_norm": 0.19326002895832062,
+      "learning_rate": 8e-05,
+      "loss": 1.4871,
+      "step": 2447
+    },
+    {
+      "epoch": 0.33410672853828305,
+      "grad_norm": 0.1956133097410202,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 2448
+    },
+    {
+      "epoch": 0.3342432100450389,
+      "grad_norm": 0.1967337727546692,
+      "learning_rate": 8e-05,
+      "loss": 1.6301,
+      "step": 2449
+    },
+    {
+      "epoch": 0.33437969155179476,
+      "grad_norm": 0.1888309121131897,
+      "learning_rate": 8e-05,
+      "loss": 1.5053,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3345161730585506,
+      "grad_norm": 0.19888371229171753,
+      "learning_rate": 8e-05,
+      "loss": 1.5681,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3346526545653064,
+      "grad_norm": 0.19180575013160706,
+      "learning_rate": 8e-05,
+      "loss": 1.571,
+      "step": 2452
+    },
+    {
+      "epoch": 0.33478913607206223,
+      "grad_norm": 0.19496522843837738,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 2453
+    },
+    {
+      "epoch": 0.33492561757881806,
+      "grad_norm": 0.1954294741153717,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3350620990855739,
+      "grad_norm": 0.194271057844162,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 2455
+    },
+    {
+      "epoch": 0.33519858059232976,
+      "grad_norm": 0.19182570278644562,
+      "learning_rate": 8e-05,
+      "loss": 1.5984,
+      "step": 2456
+    },
+    {
+      "epoch": 0.3353350620990856,
+      "grad_norm": 0.1884409636259079,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 2457
+    },
+    {
+      "epoch": 0.3354715436058414,
+      "grad_norm": 0.1984720677137375,
+      "learning_rate": 8e-05,
+      "loss": 1.5801,
+      "step": 2458
+    },
+    {
+      "epoch": 0.33560802511259724,
+      "grad_norm": 0.1877586543560028,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 2459
+    },
+    {
+      "epoch": 0.33574450661935307,
+      "grad_norm": 0.1922798901796341,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3358809881261089,
+      "grad_norm": 0.18717573583126068,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 2461
+    },
+    {
+      "epoch": 0.3360174696328648,
+      "grad_norm": 0.19478553533554077,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 2462
+    },
+    {
+      "epoch": 0.3361539511396206,
+      "grad_norm": 0.18772883713245392,
+      "learning_rate": 8e-05,
+      "loss": 1.5364,
+      "step": 2463
+    },
+    {
+      "epoch": 0.3362904326463764,
+      "grad_norm": 0.19882705807685852,
+      "learning_rate": 8e-05,
+      "loss": 1.6195,
+      "step": 2464
+    },
+    {
+      "epoch": 0.33642691415313225,
+      "grad_norm": 0.19613197445869446,
+      "learning_rate": 8e-05,
+      "loss": 1.4862,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3365633956598881,
+      "grad_norm": 0.20750534534454346,
+      "learning_rate": 8e-05,
+      "loss": 1.5163,
+      "step": 2466
+    },
+    {
+      "epoch": 0.3366998771666439,
+      "grad_norm": 0.19459302723407745,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 2467
+    },
+    {
+      "epoch": 0.3368363586733997,
+      "grad_norm": 0.18895673751831055,
+      "learning_rate": 8e-05,
+      "loss": 1.5137,
+      "step": 2468
+    },
+    {
+      "epoch": 0.3369728401801556,
+      "grad_norm": 0.19213056564331055,
+      "learning_rate": 8e-05,
+      "loss": 1.5241,
+      "step": 2469
+    },
+    {
+      "epoch": 0.33710932168691143,
+      "grad_norm": 0.1933610886335373,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 2470
+    },
+    {
+      "epoch": 0.33724580319366726,
+      "grad_norm": 0.18792147934436798,
+      "learning_rate": 8e-05,
+      "loss": 1.5181,
+      "step": 2471
+    },
+    {
+      "epoch": 0.3373822847004231,
+      "grad_norm": 0.18653547763824463,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 2472
+    },
+    {
+      "epoch": 0.3375187662071789,
+      "grad_norm": 0.19106873869895935,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 2473
+    },
+    {
+      "epoch": 0.33765524771393474,
+      "grad_norm": 0.19438928365707397,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 2474
+    },
+    {
+      "epoch": 0.3377917292206906,
+      "grad_norm": 0.191884383559227,
+      "learning_rate": 8e-05,
+      "loss": 1.5381,
+      "step": 2475
+    },
+    {
+      "epoch": 0.33792821072744644,
+      "grad_norm": 0.19358599185943604,
+      "learning_rate": 8e-05,
+      "loss": 1.5264,
+      "step": 2476
+    },
+    {
+      "epoch": 0.33806469223420227,
+      "grad_norm": 0.1976454108953476,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 2477
+    },
+    {
+      "epoch": 0.3382011737409581,
+      "grad_norm": 0.1958320587873459,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3383376552477139,
+      "grad_norm": 0.1989540010690689,
+      "learning_rate": 8e-05,
+      "loss": 1.5225,
+      "step": 2479
+    },
+    {
+      "epoch": 0.33847413675446975,
+      "grad_norm": 0.18735961616039276,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3386106182612256,
+      "grad_norm": 0.18536771833896637,
+      "learning_rate": 8e-05,
+      "loss": 1.4572,
+      "step": 2481
+    },
+    {
+      "epoch": 0.33874709976798145,
+      "grad_norm": 0.19066278636455536,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 2482
+    },
+    {
+      "epoch": 0.3388835812747373,
+      "grad_norm": 0.19877173006534576,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 2483
+    },
+    {
+      "epoch": 0.3390200627814931,
+      "grad_norm": 0.19461213052272797,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 2484
+    },
+    {
+      "epoch": 0.33915654428824893,
+      "grad_norm": 0.19343744218349457,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 2485
+    },
+    {
+      "epoch": 0.33929302579500475,
+      "grad_norm": 0.20381417870521545,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 2486
+    },
+    {
+      "epoch": 0.33942950730176064,
+      "grad_norm": 0.19497017562389374,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 2487
+    },
+    {
+      "epoch": 0.33956598880851646,
+      "grad_norm": 0.198625385761261,
+      "learning_rate": 8e-05,
+      "loss": 1.5405,
+      "step": 2488
+    },
+    {
+      "epoch": 0.3397024703152723,
+      "grad_norm": 0.19684846699237823,
+      "learning_rate": 8e-05,
+      "loss": 1.5069,
+      "step": 2489
+    },
+    {
+      "epoch": 0.3398389518220281,
+      "grad_norm": 0.19157476723194122,
+      "learning_rate": 8e-05,
+      "loss": 1.4368,
+      "step": 2490
+    },
+    {
+      "epoch": 0.33997543332878394,
+      "grad_norm": 0.2119031548500061,
+      "learning_rate": 8e-05,
+      "loss": 1.5561,
+      "step": 2491
+    },
+    {
+      "epoch": 0.34011191483553976,
+      "grad_norm": 0.19901813566684723,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 2492
+    },
+    {
+      "epoch": 0.34024839634229564,
+      "grad_norm": 0.18913337588310242,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 2493
+    },
+    {
+      "epoch": 0.34038487784905147,
+      "grad_norm": 0.21063925325870514,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 2494
+    },
+    {
+      "epoch": 0.3405213593558073,
+      "grad_norm": 0.18714042007923126,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3406578408625631,
+      "grad_norm": 0.19401006400585175,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 2496
+    },
+    {
+      "epoch": 0.34079432236931895,
+      "grad_norm": 0.20214806497097015,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3409308038760748,
+      "grad_norm": 0.19135676324367523,
+      "learning_rate": 8e-05,
+      "loss": 1.5814,
+      "step": 2498
+    },
+    {
+      "epoch": 0.34106728538283065,
+      "grad_norm": 0.19156618416309357,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 2499
+    },
+    {
+      "epoch": 0.3412037668895865,
+      "grad_norm": 0.2019028216600418,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3413402483963423,
+      "grad_norm": 0.19800348579883575,
+      "learning_rate": 8e-05,
+      "loss": 1.5174,
+      "step": 2501
+    },
+    {
+      "epoch": 0.34147672990309813,
+      "grad_norm": 0.19926117360591888,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 2502
+    },
+    {
+      "epoch": 0.34161321140985396,
+      "grad_norm": 0.20413170754909515,
+      "learning_rate": 8e-05,
+      "loss": 1.5151,
+      "step": 2503
+    },
+    {
+      "epoch": 0.3417496929166098,
+      "grad_norm": 0.19088317453861237,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 2504
+    },
+    {
+      "epoch": 0.3418861744233656,
+      "grad_norm": 0.1939716935157776,
+      "learning_rate": 8e-05,
+      "loss": 1.5221,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3420226559301215,
+      "grad_norm": 0.2010374814271927,
+      "learning_rate": 8e-05,
+      "loss": 1.6089,
+      "step": 2506
+    },
+    {
+      "epoch": 0.3421591374368773,
+      "grad_norm": 0.19489942491054535,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 2507
+    },
+    {
+      "epoch": 0.34229561894363314,
+      "grad_norm": 0.19662658870220184,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 2508
+    },
+    {
+      "epoch": 0.34243210045038897,
+      "grad_norm": 0.19633057713508606,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 2509
+    },
+    {
+      "epoch": 0.3425685819571448,
+      "grad_norm": 0.1840714067220688,
+      "learning_rate": 8e-05,
+      "loss": 1.4578,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3427050634639006,
+      "grad_norm": 0.2019944041967392,
+      "learning_rate": 8e-05,
+      "loss": 1.5584,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3428415449706565,
+      "grad_norm": 0.1952175796031952,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 2512
+    },
+    {
+      "epoch": 0.3429780264774123,
+      "grad_norm": 0.195237398147583,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 2513
+    },
+    {
+      "epoch": 0.34311450798416815,
+      "grad_norm": 0.20003804564476013,
+      "learning_rate": 8e-05,
+      "loss": 1.5677,
+      "step": 2514
+    },
+    {
+      "epoch": 0.343250989490924,
+      "grad_norm": 0.19378447532653809,
+      "learning_rate": 8e-05,
+      "loss": 1.4833,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3433874709976798,
+      "grad_norm": 0.19758976995944977,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3435239525044356,
+      "grad_norm": 0.19025437533855438,
+      "learning_rate": 8e-05,
+      "loss": 1.4981,
+      "step": 2517
+    },
+    {
+      "epoch": 0.3436604340111915,
+      "grad_norm": 0.18906742334365845,
+      "learning_rate": 8e-05,
+      "loss": 1.5527,
+      "step": 2518
+    },
+    {
+      "epoch": 0.34379691551794733,
+      "grad_norm": 0.21416275203227997,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 2519
+    },
+    {
+      "epoch": 0.34393339702470316,
+      "grad_norm": 0.19798676669597626,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 2520
+    },
+    {
+      "epoch": 0.344069878531459,
+      "grad_norm": 0.21285538375377655,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 2521
+    },
+    {
+      "epoch": 0.3442063600382148,
+      "grad_norm": 0.20711740851402283,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 2522
+    },
+    {
+      "epoch": 0.34434284154497063,
+      "grad_norm": 0.20285910367965698,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 2523
+    },
+    {
+      "epoch": 0.3444793230517265,
+      "grad_norm": 0.19546018540859222,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 2524
+    },
+    {
+      "epoch": 0.34461580455848234,
+      "grad_norm": 0.18902599811553955,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 2525
+    },
+    {
+      "epoch": 0.34475228606523817,
+      "grad_norm": 0.19517406821250916,
+      "learning_rate": 8e-05,
+      "loss": 1.5003,
+      "step": 2526
+    },
+    {
+      "epoch": 0.344888767571994,
+      "grad_norm": 0.19044962525367737,
+      "learning_rate": 8e-05,
+      "loss": 1.5595,
+      "step": 2527
+    },
+    {
+      "epoch": 0.3450252490787498,
+      "grad_norm": 0.19572162628173828,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 2528
+    },
+    {
+      "epoch": 0.34516173058550564,
+      "grad_norm": 0.19989044964313507,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 2529
+    },
+    {
+      "epoch": 0.3452982120922615,
+      "grad_norm": 0.1954621970653534,
+      "learning_rate": 8e-05,
+      "loss": 1.4734,
+      "step": 2530
+    },
+    {
+      "epoch": 0.34543469359901735,
+      "grad_norm": 0.19431091845035553,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 2531
+    },
+    {
+      "epoch": 0.3455711751057732,
+      "grad_norm": 0.20303569734096527,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 2532
+    },
+    {
+      "epoch": 0.345707656612529,
+      "grad_norm": 0.1978529393672943,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 2533
+    },
+    {
+      "epoch": 0.3458441381192848,
+      "grad_norm": 0.1847924143075943,
+      "learning_rate": 8e-05,
+      "loss": 1.4782,
+      "step": 2534
+    },
+    {
+      "epoch": 0.34598061962604065,
+      "grad_norm": 0.19170251488685608,
+      "learning_rate": 8e-05,
+      "loss": 1.5808,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3461171011327965,
+      "grad_norm": 0.20397469401359558,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 2536
+    },
+    {
+      "epoch": 0.34625358263955236,
+      "grad_norm": 0.19734856486320496,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 2537
+    },
+    {
+      "epoch": 0.3463900641463082,
+      "grad_norm": 0.194797083735466,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 2538
+    },
+    {
+      "epoch": 0.346526545653064,
+      "grad_norm": 0.19497369229793549,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 2539
+    },
+    {
+      "epoch": 0.34666302715981984,
+      "grad_norm": 0.1882125437259674,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 2540
+    },
+    {
+      "epoch": 0.34679950866657566,
+      "grad_norm": 0.19456225633621216,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 2541
+    },
+    {
+      "epoch": 0.3469359901733315,
+      "grad_norm": 0.19356907904148102,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 2542
+    },
+    {
+      "epoch": 0.34707247168008737,
+      "grad_norm": 0.19334271550178528,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3472089531868432,
+      "grad_norm": 0.194456085562706,
+      "learning_rate": 8e-05,
+      "loss": 1.5894,
+      "step": 2544
+    },
+    {
+      "epoch": 0.347345434693599,
+      "grad_norm": 0.19849246740341187,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 2545
+    },
+    {
+      "epoch": 0.34748191620035485,
+      "grad_norm": 0.1903613805770874,
+      "learning_rate": 8e-05,
+      "loss": 1.5425,
+      "step": 2546
+    },
+    {
+      "epoch": 0.34761839770711067,
+      "grad_norm": 0.2045399248600006,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 2547
+    },
+    {
+      "epoch": 0.3477548792138665,
+      "grad_norm": 0.19214165210723877,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 2548
+    },
+    {
+      "epoch": 0.3478913607206224,
+      "grad_norm": 0.19689619541168213,
+      "learning_rate": 8e-05,
+      "loss": 1.5792,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3480278422273782,
+      "grad_norm": 0.2017962783575058,
+      "learning_rate": 8e-05,
+      "loss": 1.5829,
+      "step": 2550
+    },
+    {
+      "epoch": 0.34816432373413403,
+      "grad_norm": 0.1942274421453476,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 2551
+    },
+    {
+      "epoch": 0.34830080524088985,
+      "grad_norm": 0.20240655541419983,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 2552
+    },
+    {
+      "epoch": 0.3484372867476457,
+      "grad_norm": 0.1921750158071518,
+      "learning_rate": 8e-05,
+      "loss": 1.5286,
+      "step": 2553
+    },
+    {
+      "epoch": 0.3485737682544015,
+      "grad_norm": 0.19392859935760498,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3487102497611574,
+      "grad_norm": 0.19483475387096405,
+      "learning_rate": 8e-05,
+      "loss": 1.5094,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3488467312679132,
+      "grad_norm": 0.19146482646465302,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 2556
+    },
+    {
+      "epoch": 0.34898321277466904,
+      "grad_norm": 0.19937308132648468,
+      "learning_rate": 8e-05,
+      "loss": 1.5062,
+      "step": 2557
+    },
+    {
+      "epoch": 0.34911969428142486,
+      "grad_norm": 0.19719766080379486,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3492561757881807,
+      "grad_norm": 0.1888270080089569,
+      "learning_rate": 8e-05,
+      "loss": 1.5084,
+      "step": 2559
+    },
+    {
+      "epoch": 0.3493926572949365,
+      "grad_norm": 0.20241068303585052,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3495291388016924,
+      "grad_norm": 0.1963377445936203,
+      "learning_rate": 8e-05,
+      "loss": 1.5937,
+      "step": 2561
+    },
+    {
+      "epoch": 0.3496656203084482,
+      "grad_norm": 0.19205492734909058,
+      "learning_rate": 8e-05,
+      "loss": 1.4899,
+      "step": 2562
+    },
+    {
+      "epoch": 0.34980210181520405,
+      "grad_norm": 0.2034924328327179,
+      "learning_rate": 8e-05,
+      "loss": 1.6164,
+      "step": 2563
+    },
+    {
+      "epoch": 0.3499385833219599,
+      "grad_norm": 0.1940470039844513,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 2564
+    },
+    {
+      "epoch": 0.3500750648287157,
+      "grad_norm": 0.20668810606002808,
+      "learning_rate": 8e-05,
+      "loss": 1.6064,
+      "step": 2565
+    },
+    {
+      "epoch": 0.3502115463354715,
+      "grad_norm": 0.2019558995962143,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3503480278422274,
+      "grad_norm": 0.19554457068443298,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 2567
+    },
+    {
+      "epoch": 0.35048450934898323,
+      "grad_norm": 0.2013549506664276,
+      "learning_rate": 8e-05,
+      "loss": 1.6049,
+      "step": 2568
+    },
+    {
+      "epoch": 0.35062099085573906,
+      "grad_norm": 0.18695232272148132,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3507574723624949,
+      "grad_norm": 0.19889196753501892,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3508939538692507,
+      "grad_norm": 0.20030449330806732,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 2571
+    },
+    {
+      "epoch": 0.35103043537600653,
+      "grad_norm": 0.19805726408958435,
+      "learning_rate": 8e-05,
+      "loss": 1.6069,
+      "step": 2572
+    },
+    {
+      "epoch": 0.35116691688276236,
+      "grad_norm": 0.1995544582605362,
+      "learning_rate": 8e-05,
+      "loss": 1.4945,
+      "step": 2573
+    },
+    {
+      "epoch": 0.35130339838951824,
+      "grad_norm": 0.19294312596321106,
+      "learning_rate": 8e-05,
+      "loss": 1.4802,
+      "step": 2574
+    },
+    {
+      "epoch": 0.35143987989627407,
+      "grad_norm": 0.19078242778778076,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3515763614030299,
+      "grad_norm": 0.20172356069087982,
+      "learning_rate": 8e-05,
+      "loss": 1.509,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3517128429097857,
+      "grad_norm": 0.19706188142299652,
+      "learning_rate": 8e-05,
+      "loss": 1.5398,
+      "step": 2577
+    },
+    {
+      "epoch": 0.35184932441654154,
+      "grad_norm": 0.19191832840442657,
+      "learning_rate": 8e-05,
+      "loss": 1.4459,
+      "step": 2578
+    },
+    {
+      "epoch": 0.35198580592329737,
+      "grad_norm": 0.1973387449979782,
+      "learning_rate": 8e-05,
+      "loss": 1.5172,
+      "step": 2579
+    },
+    {
+      "epoch": 0.35212228743005325,
+      "grad_norm": 0.19400425255298615,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3522587689368091,
+      "grad_norm": 0.2051192671060562,
+      "learning_rate": 8e-05,
+      "loss": 1.5903,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3523952504435649,
+      "grad_norm": 0.18710793554782867,
+      "learning_rate": 8e-05,
+      "loss": 1.438,
+      "step": 2582
+    },
+    {
+      "epoch": 0.3525317319503207,
+      "grad_norm": 0.18687210977077484,
+      "learning_rate": 8e-05,
+      "loss": 1.4444,
+      "step": 2583
+    },
+    {
+      "epoch": 0.35266821345707655,
+      "grad_norm": 0.20916792750358582,
+      "learning_rate": 8e-05,
+      "loss": 1.579,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3528046949638324,
+      "grad_norm": 0.201734721660614,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 2585
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.19195617735385895,
+      "learning_rate": 8e-05,
+      "loss": 1.5549,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3530776579773441,
+      "grad_norm": 0.1995304673910141,
+      "learning_rate": 8e-05,
+      "loss": 1.4883,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3532141394840999,
+      "grad_norm": 0.20530566573143005,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 2588
+    },
+    {
+      "epoch": 0.35335062099085573,
+      "grad_norm": 0.1953418105840683,
+      "learning_rate": 8e-05,
+      "loss": 1.4764,
+      "step": 2589
+    },
+    {
+      "epoch": 0.35348710249761156,
+      "grad_norm": 0.19232945144176483,
+      "learning_rate": 8e-05,
+      "loss": 1.4842,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3536235840043674,
+      "grad_norm": 0.1948445439338684,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 2591
+    },
+    {
+      "epoch": 0.35376006551112327,
+      "grad_norm": 0.19241535663604736,
+      "learning_rate": 8e-05,
+      "loss": 1.5099,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3538965470178791,
+      "grad_norm": 0.19627416133880615,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3540330285246349,
+      "grad_norm": 0.19273103773593903,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 2594
+    },
+    {
+      "epoch": 0.35416951003139074,
+      "grad_norm": 0.19843879342079163,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 2595
+    },
+    {
+      "epoch": 0.35430599153814657,
+      "grad_norm": 0.19716261327266693,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3544424730449024,
+      "grad_norm": 0.1959685981273651,
+      "learning_rate": 8e-05,
+      "loss": 1.5541,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3545789545516583,
+      "grad_norm": 0.20762799680233002,
+      "learning_rate": 8e-05,
+      "loss": 1.5578,
+      "step": 2598
+    },
+    {
+      "epoch": 0.3547154360584141,
+      "grad_norm": 0.19284506142139435,
+      "learning_rate": 8e-05,
+      "loss": 1.5113,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3548519175651699,
+      "grad_norm": 0.19372038543224335,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 2600
+    },
+    {
+      "epoch": 0.35498839907192575,
+      "grad_norm": 0.1953934282064438,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3551248805786816,
+      "grad_norm": 0.20578987896442413,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3552613620854374,
+      "grad_norm": 0.20066507160663605,
+      "learning_rate": 8e-05,
+      "loss": 1.5113,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3553978435921933,
+      "grad_norm": 0.19259048998355865,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3555343250989491,
+      "grad_norm": 0.2026747316122055,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 2605
+    },
+    {
+      "epoch": 0.35567080660570494,
+      "grad_norm": 0.20289842784404755,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 2606
+    },
+    {
+      "epoch": 0.35580728811246076,
+      "grad_norm": 0.19875353574752808,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 2607
+    },
+    {
+      "epoch": 0.3559437696192166,
+      "grad_norm": 0.2012946903705597,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3560802511259724,
+      "grad_norm": 0.2037465125322342,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 2609
+    },
+    {
+      "epoch": 0.35621673263272824,
+      "grad_norm": 0.1972435861825943,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3563532141394841,
+      "grad_norm": 0.19260697066783905,
+      "learning_rate": 8e-05,
+      "loss": 1.5142,
+      "step": 2611
+    },
+    {
+      "epoch": 0.35648969564623995,
+      "grad_norm": 0.19583943486213684,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 2612
+    },
+    {
+      "epoch": 0.35662617715299577,
+      "grad_norm": 0.19478709995746613,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 2613
+    },
+    {
+      "epoch": 0.3567626586597516,
+      "grad_norm": 0.18682514131069183,
+      "learning_rate": 8e-05,
+      "loss": 1.5328,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3568991401665074,
+      "grad_norm": 0.19836457073688507,
+      "learning_rate": 8e-05,
+      "loss": 1.5571,
+      "step": 2615
+    },
+    {
+      "epoch": 0.35703562167326325,
+      "grad_norm": 0.19091983139514923,
+      "learning_rate": 8e-05,
+      "loss": 1.5105,
+      "step": 2616
+    },
+    {
+      "epoch": 0.35717210318001913,
+      "grad_norm": 0.197165384888649,
+      "learning_rate": 8e-05,
+      "loss": 1.5314,
+      "step": 2617
+    },
+    {
+      "epoch": 0.35730858468677495,
+      "grad_norm": 0.18971925973892212,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 2618
+    },
+    {
+      "epoch": 0.3574450661935308,
+      "grad_norm": 0.18598932027816772,
+      "learning_rate": 8e-05,
+      "loss": 1.4239,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3575815477002866,
+      "grad_norm": 0.19195757806301117,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 2620
+    },
+    {
+      "epoch": 0.35771802920704243,
+      "grad_norm": 0.19654802978038788,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 2621
+    },
+    {
+      "epoch": 0.35785451071379826,
+      "grad_norm": 0.18688727915287018,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 2622
+    },
+    {
+      "epoch": 0.35799099222055414,
+      "grad_norm": 0.20256516337394714,
+      "learning_rate": 8e-05,
+      "loss": 1.5267,
+      "step": 2623
+    },
+    {
+      "epoch": 0.35812747372730996,
+      "grad_norm": 0.19448009133338928,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 2624
+    },
+    {
+      "epoch": 0.3582639552340658,
+      "grad_norm": 0.19296544790267944,
+      "learning_rate": 8e-05,
+      "loss": 1.5376,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3584004367408216,
+      "grad_norm": 0.20121923089027405,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 2626
+    },
+    {
+      "epoch": 0.35853691824757744,
+      "grad_norm": 0.19750294089317322,
+      "learning_rate": 8e-05,
+      "loss": 1.4965,
+      "step": 2627
+    },
+    {
+      "epoch": 0.35867339975433327,
+      "grad_norm": 0.19986854493618011,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 2628
+    },
+    {
+      "epoch": 0.35880988126108915,
+      "grad_norm": 0.19554506242275238,
+      "learning_rate": 8e-05,
+      "loss": 1.5429,
+      "step": 2629
+    },
+    {
+      "epoch": 0.358946362767845,
+      "grad_norm": 0.19883698225021362,
+      "learning_rate": 8e-05,
+      "loss": 1.4906,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3590828442746008,
+      "grad_norm": 0.18505001068115234,
+      "learning_rate": 8e-05,
+      "loss": 1.5208,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3592193257813566,
+      "grad_norm": 0.20315483212471008,
+      "learning_rate": 8e-05,
+      "loss": 1.5348,
+      "step": 2632
+    },
+    {
+      "epoch": 0.35935580728811245,
+      "grad_norm": 0.18695324659347534,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3594922887948683,
+      "grad_norm": 0.18792392313480377,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 2634
+    },
+    {
+      "epoch": 0.35962877030162416,
+      "grad_norm": 0.1980956792831421,
+      "learning_rate": 8e-05,
+      "loss": 1.5276,
+      "step": 2635
+    },
+    {
+      "epoch": 0.35976525180838,
+      "grad_norm": 0.20917940139770508,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3599017333151358,
+      "grad_norm": 0.19428816437721252,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 2637
+    },
+    {
+      "epoch": 0.36003821482189163,
+      "grad_norm": 0.20210137963294983,
+      "learning_rate": 8e-05,
+      "loss": 1.4561,
+      "step": 2638
+    },
+    {
+      "epoch": 0.36017469632864746,
+      "grad_norm": 0.19680465757846832,
+      "learning_rate": 8e-05,
+      "loss": 1.55,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3603111778354033,
+      "grad_norm": 0.19674304127693176,
+      "learning_rate": 8e-05,
+      "loss": 1.5022,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3604476593421591,
+      "grad_norm": 0.20113793015480042,
+      "learning_rate": 8e-05,
+      "loss": 1.5228,
+      "step": 2641
+    },
+    {
+      "epoch": 0.360584140848915,
+      "grad_norm": 0.19191408157348633,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3607206223556708,
+      "grad_norm": 0.19506217539310455,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 2643
+    },
+    {
+      "epoch": 0.36085710386242664,
+      "grad_norm": 0.20981590449810028,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 2644
+    },
+    {
+      "epoch": 0.36099358536918247,
+      "grad_norm": 0.19451993703842163,
+      "learning_rate": 8e-05,
+      "loss": 1.563,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3611300668759383,
+      "grad_norm": 0.2060956358909607,
+      "learning_rate": 8e-05,
+      "loss": 1.4423,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3612665483826941,
+      "grad_norm": 0.19907425343990326,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 2647
+    },
+    {
+      "epoch": 0.36140302988945,
+      "grad_norm": 0.19366858899593353,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 2648
+    },
+    {
+      "epoch": 0.3615395113962058,
+      "grad_norm": 0.21619334816932678,
+      "learning_rate": 8e-05,
+      "loss": 1.5359,
+      "step": 2649
+    },
+    {
+      "epoch": 0.36167599290296165,
+      "grad_norm": 0.1992226541042328,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3618124744097175,
+      "grad_norm": 0.21054105460643768,
+      "learning_rate": 8e-05,
+      "loss": 1.4579,
+      "step": 2651
+    },
+    {
+      "epoch": 0.3619489559164733,
+      "grad_norm": 0.19776053726673126,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 2652
+    },
+    {
+      "epoch": 0.36208543742322913,
+      "grad_norm": 0.1988450139760971,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 2653
+    },
+    {
+      "epoch": 0.362221918929985,
+      "grad_norm": 0.19162483513355255,
+      "learning_rate": 8e-05,
+      "loss": 1.522,
+      "step": 2654
+    },
+    {
+      "epoch": 0.36235840043674084,
+      "grad_norm": 0.1950751692056656,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 2655
+    },
+    {
+      "epoch": 0.36249488194349666,
+      "grad_norm": 0.1985386461019516,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 2656
+    },
+    {
+      "epoch": 0.3626313634502525,
+      "grad_norm": 0.20346039533615112,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3627678449570083,
+      "grad_norm": 0.19609668850898743,
+      "learning_rate": 8e-05,
+      "loss": 1.5484,
+      "step": 2658
+    },
+    {
+      "epoch": 0.36290432646376414,
+      "grad_norm": 0.19500140845775604,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 2659
+    },
+    {
+      "epoch": 0.36304080797052,
+      "grad_norm": 0.1938134878873825,
+      "learning_rate": 8e-05,
+      "loss": 1.5155,
+      "step": 2660
+    },
+    {
+      "epoch": 0.36317728947727584,
+      "grad_norm": 0.19112981855869293,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 2661
+    },
+    {
+      "epoch": 0.36331377098403167,
+      "grad_norm": 0.18887364864349365,
+      "learning_rate": 8e-05,
+      "loss": 1.4811,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3634502524907875,
+      "grad_norm": 0.20011357963085175,
+      "learning_rate": 8e-05,
+      "loss": 1.5197,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3635867339975433,
+      "grad_norm": 0.18942269682884216,
+      "learning_rate": 8e-05,
+      "loss": 1.4658,
+      "step": 2664
+    },
+    {
+      "epoch": 0.36372321550429915,
+      "grad_norm": 0.19741171598434448,
+      "learning_rate": 8e-05,
+      "loss": 1.5095,
+      "step": 2665
+    },
+    {
+      "epoch": 0.363859697011055,
+      "grad_norm": 0.1949249655008316,
+      "learning_rate": 8e-05,
+      "loss": 1.496,
+      "step": 2666
+    },
+    {
+      "epoch": 0.36399617851781085,
+      "grad_norm": 0.1998748481273651,
+      "learning_rate": 8e-05,
+      "loss": 1.5745,
+      "step": 2667
+    },
+    {
+      "epoch": 0.3641326600245667,
+      "grad_norm": 0.19141504168510437,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3642691415313225,
+      "grad_norm": 0.19324477016925812,
+      "learning_rate": 8e-05,
+      "loss": 1.5638,
+      "step": 2669
+    },
+    {
+      "epoch": 0.36440562303807833,
+      "grad_norm": 0.1996828317642212,
+      "learning_rate": 8e-05,
+      "loss": 1.5698,
+      "step": 2670
+    },
+    {
+      "epoch": 0.36454210454483416,
+      "grad_norm": 0.195877805352211,
+      "learning_rate": 8e-05,
+      "loss": 1.5823,
+      "step": 2671
+    },
+    {
+      "epoch": 0.36467858605159004,
+      "grad_norm": 0.19857576489448547,
+      "learning_rate": 8e-05,
+      "loss": 1.5508,
+      "step": 2672
+    },
+    {
+      "epoch": 0.36481506755834586,
+      "grad_norm": 0.2116616815328598,
+      "learning_rate": 8e-05,
+      "loss": 1.5047,
+      "step": 2673
+    },
+    {
+      "epoch": 0.3649515490651017,
+      "grad_norm": 0.20244544744491577,
+      "learning_rate": 8e-05,
+      "loss": 1.6045,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3650880305718575,
+      "grad_norm": 0.19361169636249542,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 2675
+    },
+    {
+      "epoch": 0.36522451207861334,
+      "grad_norm": 0.19096067547798157,
+      "learning_rate": 8e-05,
+      "loss": 1.5076,
+      "step": 2676
+    },
+    {
+      "epoch": 0.36536099358536916,
+      "grad_norm": 0.20173348486423492,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 2677
+    },
+    {
+      "epoch": 0.365497475092125,
+      "grad_norm": 0.19586752355098724,
+      "learning_rate": 8e-05,
+      "loss": 1.4857,
+      "step": 2678
+    },
+    {
+      "epoch": 0.36563395659888087,
+      "grad_norm": 0.23996679484844208,
+      "learning_rate": 8e-05,
+      "loss": 1.537,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3657704381056367,
+      "grad_norm": 0.1985531747341156,
+      "learning_rate": 8e-05,
+      "loss": 1.4903,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3659069196123925,
+      "grad_norm": 0.19754116237163544,
+      "learning_rate": 8e-05,
+      "loss": 1.5281,
+      "step": 2681
+    },
+    {
+      "epoch": 0.36604340111914835,
+      "grad_norm": 0.20095433294773102,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3661798826259042,
+      "grad_norm": 0.19450584053993225,
+      "learning_rate": 8e-05,
+      "loss": 1.498,
+      "step": 2683
+    },
+    {
+      "epoch": 0.36631636413266,
+      "grad_norm": 0.19317109882831573,
+      "learning_rate": 8e-05,
+      "loss": 1.4674,
+      "step": 2684
+    },
+    {
+      "epoch": 0.3664528456394159,
+      "grad_norm": 0.20266908407211304,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3665893271461717,
+      "grad_norm": 0.20954738557338715,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 2686
+    },
+    {
+      "epoch": 0.36672580865292753,
+      "grad_norm": 0.19602203369140625,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 2687
+    },
+    {
+      "epoch": 0.36686229015968336,
+      "grad_norm": 0.18985100090503693,
+      "learning_rate": 8e-05,
+      "loss": 1.5467,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3669987716664392,
+      "grad_norm": 0.2018248736858368,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 2689
+    },
+    {
+      "epoch": 0.367135253173195,
+      "grad_norm": 0.19734995067119598,
+      "learning_rate": 8e-05,
+      "loss": 1.5446,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3672717346799509,
+      "grad_norm": 0.19626644253730774,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3674082161867067,
+      "grad_norm": 0.2009628862142563,
+      "learning_rate": 8e-05,
+      "loss": 1.5271,
+      "step": 2692
+    },
+    {
+      "epoch": 0.36754469769346254,
+      "grad_norm": 0.1883717179298401,
+      "learning_rate": 8e-05,
+      "loss": 1.4464,
+      "step": 2693
+    },
+    {
+      "epoch": 0.36768117920021837,
+      "grad_norm": 0.1963186413049698,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 2694
+    },
+    {
+      "epoch": 0.3678176607069742,
+      "grad_norm": 0.1977759301662445,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 2695
+    },
+    {
+      "epoch": 0.36795414221373,
+      "grad_norm": 0.1960255205631256,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3680906237204859,
+      "grad_norm": 0.19921021163463593,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 2697
+    },
+    {
+      "epoch": 0.3682271052272417,
+      "grad_norm": 0.19218194484710693,
+      "learning_rate": 8e-05,
+      "loss": 1.5088,
+      "step": 2698
+    },
+    {
+      "epoch": 0.36836358673399755,
+      "grad_norm": 0.19653165340423584,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3685000682407534,
+      "grad_norm": 0.19092877209186554,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3686365497475092,
+      "grad_norm": 0.19032913446426392,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 2701
+    },
+    {
+      "epoch": 0.368773031254265,
+      "grad_norm": 0.19808073341846466,
+      "learning_rate": 8e-05,
+      "loss": 1.5306,
+      "step": 2702
+    },
+    {
+      "epoch": 0.3689095127610209,
+      "grad_norm": 0.1957745999097824,
+      "learning_rate": 8e-05,
+      "loss": 1.5417,
+      "step": 2703
+    },
+    {
+      "epoch": 0.36904599426777673,
+      "grad_norm": 0.19573797285556793,
+      "learning_rate": 8e-05,
+      "loss": 1.5048,
+      "step": 2704
+    },
+    {
+      "epoch": 0.36918247577453256,
+      "grad_norm": 0.19019901752471924,
+      "learning_rate": 8e-05,
+      "loss": 1.5143,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3693189572812884,
+      "grad_norm": 0.19325849413871765,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3694554387880442,
+      "grad_norm": 0.19284597039222717,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 2707
+    },
+    {
+      "epoch": 0.36959192029480004,
+      "grad_norm": 0.20323938131332397,
+      "learning_rate": 8e-05,
+      "loss": 1.5677,
+      "step": 2708
+    },
+    {
+      "epoch": 0.36972840180155586,
+      "grad_norm": 0.20294398069381714,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 2709
+    },
+    {
+      "epoch": 0.36986488330831174,
+      "grad_norm": 0.19054478406906128,
+      "learning_rate": 8e-05,
+      "loss": 1.4954,
+      "step": 2710
+    },
+    {
+      "epoch": 0.37000136481506757,
+      "grad_norm": 0.194930300116539,
+      "learning_rate": 8e-05,
+      "loss": 1.4625,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3701378463218234,
+      "grad_norm": 0.1962171345949173,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 2712
+    },
+    {
+      "epoch": 0.3702743278285792,
+      "grad_norm": 0.1992320865392685,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 2713
+    },
+    {
+      "epoch": 0.37041080933533505,
+      "grad_norm": 0.19754379987716675,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 2714
+    },
+    {
+      "epoch": 0.37054729084209087,
+      "grad_norm": 0.19990310072898865,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 2715
+    },
+    {
+      "epoch": 0.37068377234884675,
+      "grad_norm": 0.20341627299785614,
+      "learning_rate": 8e-05,
+      "loss": 1.5973,
+      "step": 2716
+    },
+    {
+      "epoch": 0.3708202538556026,
+      "grad_norm": 0.19848500192165375,
+      "learning_rate": 8e-05,
+      "loss": 1.4709,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3709567353623584,
+      "grad_norm": 0.192295104265213,
+      "learning_rate": 8e-05,
+      "loss": 1.5477,
+      "step": 2718
+    },
+    {
+      "epoch": 0.37109321686911423,
+      "grad_norm": 0.19476161897182465,
+      "learning_rate": 8e-05,
+      "loss": 1.4509,
+      "step": 2719
+    },
+    {
+      "epoch": 0.37122969837587005,
+      "grad_norm": 0.19318239390850067,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3713661798826259,
+      "grad_norm": 0.19200479984283447,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 2721
+    },
+    {
+      "epoch": 0.37150266138938176,
+      "grad_norm": 0.19040045142173767,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3716391428961376,
+      "grad_norm": 0.19674921035766602,
+      "learning_rate": 8e-05,
+      "loss": 1.5137,
+      "step": 2723
+    },
+    {
+      "epoch": 0.3717756244028934,
+      "grad_norm": 0.1906231790781021,
+      "learning_rate": 8e-05,
+      "loss": 1.509,
+      "step": 2724
+    },
+    {
+      "epoch": 0.37191210590964924,
+      "grad_norm": 0.19204051792621613,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 2725
+    },
+    {
+      "epoch": 0.37204858741640506,
+      "grad_norm": 0.19326713681221008,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2726
+    },
+    {
+      "epoch": 0.3721850689231609,
+      "grad_norm": 0.1923893839120865,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 2727
+    },
+    {
+      "epoch": 0.37232155042991677,
+      "grad_norm": 0.19592903554439545,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3724580319366726,
+      "grad_norm": 0.19696220755577087,
+      "learning_rate": 8e-05,
+      "loss": 1.4988,
+      "step": 2729
+    },
+    {
+      "epoch": 0.3725945134434284,
+      "grad_norm": 0.19765207171440125,
+      "learning_rate": 8e-05,
+      "loss": 1.4931,
+      "step": 2730
+    },
+    {
+      "epoch": 0.37273099495018425,
+      "grad_norm": 0.19209659099578857,
+      "learning_rate": 8e-05,
+      "loss": 1.5018,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3728674764569401,
+      "grad_norm": 0.1987694799900055,
+      "learning_rate": 8e-05,
+      "loss": 1.5979,
+      "step": 2732
+    },
+    {
+      "epoch": 0.3730039579636959,
+      "grad_norm": 0.20281600952148438,
+      "learning_rate": 8e-05,
+      "loss": 1.578,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3731404394704518,
+      "grad_norm": 0.2019796520471573,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 2734
+    },
+    {
+      "epoch": 0.3732769209772076,
+      "grad_norm": 0.19378289580345154,
+      "learning_rate": 8e-05,
+      "loss": 1.5195,
+      "step": 2735
+    },
+    {
+      "epoch": 0.37341340248396343,
+      "grad_norm": 0.20103982090950012,
+      "learning_rate": 8e-05,
+      "loss": 1.4692,
+      "step": 2736
+    },
+    {
+      "epoch": 0.37354988399071926,
+      "grad_norm": 0.1952180415391922,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 2737
+    },
+    {
+      "epoch": 0.3736863654974751,
+      "grad_norm": 0.19639243185520172,
+      "learning_rate": 8e-05,
+      "loss": 1.5822,
+      "step": 2738
+    },
+    {
+      "epoch": 0.3738228470042309,
+      "grad_norm": 0.18987591564655304,
+      "learning_rate": 8e-05,
+      "loss": 1.4134,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3739593285109868,
+      "grad_norm": 0.20650343596935272,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3740958100177426,
+      "grad_norm": 0.19434286653995514,
+      "learning_rate": 8e-05,
+      "loss": 1.5034,
+      "step": 2741
+    },
+    {
+      "epoch": 0.37423229152449844,
+      "grad_norm": 0.20050287246704102,
+      "learning_rate": 8e-05,
+      "loss": 1.5615,
+      "step": 2742
+    },
+    {
+      "epoch": 0.37436877303125426,
+      "grad_norm": 0.19849415123462677,
+      "learning_rate": 8e-05,
+      "loss": 1.5767,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3745052545380101,
+      "grad_norm": 0.18816018104553223,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3746417360447659,
+      "grad_norm": 0.19758187234401703,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 2745
+    },
+    {
+      "epoch": 0.37477821755152174,
+      "grad_norm": 0.1868588775396347,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3749146990582776,
+      "grad_norm": 0.18856370449066162,
+      "learning_rate": 8e-05,
+      "loss": 1.3578,
+      "step": 2747
+    },
+    {
+      "epoch": 0.37505118056503345,
+      "grad_norm": 0.1910911649465561,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3751876620717893,
+      "grad_norm": 0.19655504822731018,
+      "learning_rate": 8e-05,
+      "loss": 1.5808,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3753241435785451,
+      "grad_norm": 0.196238175034523,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3754606250853009,
+      "grad_norm": 0.20389612019062042,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 2751
+    },
+    {
+      "epoch": 0.37559710659205675,
+      "grad_norm": 0.1911252737045288,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 2752
+    },
+    {
+      "epoch": 0.37573358809881263,
+      "grad_norm": 0.18786859512329102,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 2753
+    },
+    {
+      "epoch": 0.37587006960556846,
+      "grad_norm": 0.1957067847251892,
+      "learning_rate": 8e-05,
+      "loss": 1.4955,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3760065511123243,
+      "grad_norm": 0.19290585815906525,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3761430326190801,
+      "grad_norm": 0.19869758188724518,
+      "learning_rate": 8e-05,
+      "loss": 1.5123,
+      "step": 2756
+    },
+    {
+      "epoch": 0.37627951412583593,
+      "grad_norm": 0.19285526871681213,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 2757
+    },
+    {
+      "epoch": 0.37641599563259176,
+      "grad_norm": 0.1974519044160843,
+      "learning_rate": 8e-05,
+      "loss": 1.5903,
+      "step": 2758
+    },
+    {
+      "epoch": 0.37655247713934764,
+      "grad_norm": 0.2015846073627472,
+      "learning_rate": 8e-05,
+      "loss": 1.5061,
+      "step": 2759
+    },
+    {
+      "epoch": 0.37668895864610347,
+      "grad_norm": 0.21167241036891937,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3768254401528593,
+      "grad_norm": 0.1957661211490631,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3769619216596151,
+      "grad_norm": 0.20745979249477386,
+      "learning_rate": 8e-05,
+      "loss": 1.6164,
+      "step": 2762
+    },
+    {
+      "epoch": 0.37709840316637094,
+      "grad_norm": 0.20360028743743896,
+      "learning_rate": 8e-05,
+      "loss": 1.5006,
+      "step": 2763
+    },
+    {
+      "epoch": 0.37723488467312677,
+      "grad_norm": 0.2007840871810913,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 2764
+    },
+    {
+      "epoch": 0.37737136617988265,
+      "grad_norm": 0.197735995054245,
+      "learning_rate": 8e-05,
+      "loss": 1.4611,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3775078476866385,
+      "grad_norm": 0.20469868183135986,
+      "learning_rate": 8e-05,
+      "loss": 1.5672,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3776443291933943,
+      "grad_norm": 0.19251105189323425,
+      "learning_rate": 8e-05,
+      "loss": 1.483,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3777808107001501,
+      "grad_norm": 0.19543355703353882,
+      "learning_rate": 8e-05,
+      "loss": 1.5771,
+      "step": 2768
+    },
+    {
+      "epoch": 0.37791729220690595,
+      "grad_norm": 0.20735904574394226,
+      "learning_rate": 8e-05,
+      "loss": 1.5034,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3780537737136618,
+      "grad_norm": 0.19461093842983246,
+      "learning_rate": 8e-05,
+      "loss": 1.5222,
+      "step": 2770
+    },
+    {
+      "epoch": 0.37819025522041766,
+      "grad_norm": 0.2013184279203415,
+      "learning_rate": 8e-05,
+      "loss": 1.5441,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3783267367271735,
+      "grad_norm": 0.19306708872318268,
+      "learning_rate": 8e-05,
+      "loss": 1.4824,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3784632182339293,
+      "grad_norm": 0.20058812201023102,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 2773
+    },
+    {
+      "epoch": 0.37859969974068514,
+      "grad_norm": 0.19264377653598785,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 2774
+    },
+    {
+      "epoch": 0.37873618124744096,
+      "grad_norm": 0.1952289491891861,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3788726627541968,
+      "grad_norm": 0.19632276892662048,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 2776
+    },
+    {
+      "epoch": 0.37900914426095267,
+      "grad_norm": 0.18710362911224365,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3791456257677085,
+      "grad_norm": 0.19573870301246643,
+      "learning_rate": 8e-05,
+      "loss": 1.5084,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3792821072744643,
+      "grad_norm": 0.19529375433921814,
+      "learning_rate": 8e-05,
+      "loss": 1.5277,
+      "step": 2779
+    },
+    {
+      "epoch": 0.37941858878122015,
+      "grad_norm": 0.19804953038692474,
+      "learning_rate": 8e-05,
+      "loss": 1.4959,
+      "step": 2780
+    },
+    {
+      "epoch": 0.37955507028797597,
+      "grad_norm": 0.19311289489269257,
+      "learning_rate": 8e-05,
+      "loss": 1.5119,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3796915517947318,
+      "grad_norm": 0.1914122998714447,
+      "learning_rate": 8e-05,
+      "loss": 1.4941,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3798280333014876,
+      "grad_norm": 0.20707741379737854,
+      "learning_rate": 8e-05,
+      "loss": 1.6479,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3799645148082435,
+      "grad_norm": 0.19574078917503357,
+      "learning_rate": 8e-05,
+      "loss": 1.4836,
+      "step": 2784
+    },
+    {
+      "epoch": 0.38010099631499933,
+      "grad_norm": 0.19648070633411407,
+      "learning_rate": 8e-05,
+      "loss": 1.5115,
+      "step": 2785
+    },
+    {
+      "epoch": 0.38023747782175515,
+      "grad_norm": 0.19596077501773834,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 2786
+    },
+    {
+      "epoch": 0.380373959328511,
+      "grad_norm": 0.19718605279922485,
+      "learning_rate": 8e-05,
+      "loss": 1.5551,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3805104408352668,
+      "grad_norm": 0.20196565985679626,
+      "learning_rate": 8e-05,
+      "loss": 1.5213,
+      "step": 2788
+    },
+    {
+      "epoch": 0.38064692234202263,
+      "grad_norm": 0.19631926715373993,
+      "learning_rate": 8e-05,
+      "loss": 1.4902,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3807834038487785,
+      "grad_norm": 0.19461777806282043,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 2790
+    },
+    {
+      "epoch": 0.38091988535553434,
+      "grad_norm": 0.19680659472942352,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 2791
+    },
+    {
+      "epoch": 0.38105636686229016,
+      "grad_norm": 0.19592802226543427,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 2792
+    },
+    {
+      "epoch": 0.381192848369046,
+      "grad_norm": 0.20139016211032867,
+      "learning_rate": 8e-05,
+      "loss": 1.5193,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3813293298758018,
+      "grad_norm": 0.20303493738174438,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 2794
+    },
+    {
+      "epoch": 0.38146581138255764,
+      "grad_norm": 0.19889698922634125,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3816022928893135,
+      "grad_norm": 0.19793474674224854,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 2796
+    },
+    {
+      "epoch": 0.38173877439606935,
+      "grad_norm": 0.19721032679080963,
+      "learning_rate": 8e-05,
+      "loss": 1.5416,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3818752559028252,
+      "grad_norm": 0.1949787586927414,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 2798
+    },
+    {
+      "epoch": 0.382011737409581,
+      "grad_norm": 0.20098304748535156,
+      "learning_rate": 8e-05,
+      "loss": 1.5066,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3821482189163368,
+      "grad_norm": 0.1979992389678955,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 2800
+    },
+    {
+      "epoch": 0.38228470042309265,
+      "grad_norm": 0.19819101691246033,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 2801
+    },
+    {
+      "epoch": 0.38242118192984853,
+      "grad_norm": 0.19251833856105804,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 2802
+    },
+    {
+      "epoch": 0.38255766343660436,
+      "grad_norm": 0.19863919913768768,
+      "learning_rate": 8e-05,
+      "loss": 1.4531,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3826941449433602,
+      "grad_norm": 0.19388936460018158,
+      "learning_rate": 8e-05,
+      "loss": 1.4497,
+      "step": 2804
+    },
+    {
+      "epoch": 0.382830626450116,
+      "grad_norm": 0.19460538029670715,
+      "learning_rate": 8e-05,
+      "loss": 1.4978,
+      "step": 2805
+    },
+    {
+      "epoch": 0.38296710795687183,
+      "grad_norm": 0.1990063637495041,
+      "learning_rate": 8e-05,
+      "loss": 1.5652,
+      "step": 2806
+    },
+    {
+      "epoch": 0.38310358946362766,
+      "grad_norm": 0.19004380702972412,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 2807
+    },
+    {
+      "epoch": 0.38324007097038354,
+      "grad_norm": 0.20413538813591003,
+      "learning_rate": 8e-05,
+      "loss": 1.4504,
+      "step": 2808
+    },
+    {
+      "epoch": 0.38337655247713937,
+      "grad_norm": 0.1971040517091751,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3835130339838952,
+      "grad_norm": 0.20226643979549408,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 2810
+    },
+    {
+      "epoch": 0.383649515490651,
+      "grad_norm": 0.1960776299238205,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 2811
+    },
+    {
+      "epoch": 0.38378599699740684,
+      "grad_norm": 0.20299096405506134,
+      "learning_rate": 8e-05,
+      "loss": 1.5796,
+      "step": 2812
+    },
+    {
+      "epoch": 0.38392247850416267,
+      "grad_norm": 0.19002120196819305,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3840589600109185,
+      "grad_norm": 0.1999003142118454,
+      "learning_rate": 8e-05,
+      "loss": 1.5724,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3841954415176744,
+      "grad_norm": 0.19764094054698944,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3843319230244302,
+      "grad_norm": 0.19543305039405823,
+      "learning_rate": 8e-05,
+      "loss": 1.5514,
+      "step": 2816
+    },
+    {
+      "epoch": 0.384468404531186,
+      "grad_norm": 0.19510267674922943,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 2817
+    },
+    {
+      "epoch": 0.38460488603794185,
+      "grad_norm": 0.19557586312294006,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3847413675446977,
+      "grad_norm": 0.19730573892593384,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3848778490514535,
+      "grad_norm": 0.19240061938762665,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3850143305582094,
+      "grad_norm": 0.19733519852161407,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3851508120649652,
+      "grad_norm": 0.19816966354846954,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 2822
+    },
+    {
+      "epoch": 0.38528729357172103,
+      "grad_norm": 0.20153246819972992,
+      "learning_rate": 8e-05,
+      "loss": 1.4874,
+      "step": 2823
+    },
+    {
+      "epoch": 0.38542377507847686,
+      "grad_norm": 0.19149908423423767,
+      "learning_rate": 8e-05,
+      "loss": 1.4352,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3855602565852327,
+      "grad_norm": 0.19896867871284485,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3856967380919885,
+      "grad_norm": 0.1990615278482437,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3858332195987444,
+      "grad_norm": 0.19365647435188293,
+      "learning_rate": 8e-05,
+      "loss": 1.5248,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3859697011055002,
+      "grad_norm": 0.20246310532093048,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 2828
+    },
+    {
+      "epoch": 0.38610618261225604,
+      "grad_norm": 0.18818911910057068,
+      "learning_rate": 8e-05,
+      "loss": 1.4901,
+      "step": 2829
+    },
+    {
+      "epoch": 0.38624266411901187,
+      "grad_norm": 0.19576428830623627,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3863791456257677,
+      "grad_norm": 0.1970338076353073,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3865156271325235,
+      "grad_norm": 0.19619084894657135,
+      "learning_rate": 8e-05,
+      "loss": 1.5264,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3866521086392794,
+      "grad_norm": 0.1897464543581009,
+      "learning_rate": 8e-05,
+      "loss": 1.4905,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3867885901460352,
+      "grad_norm": 0.20004330575466156,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 2834
+    },
+    {
+      "epoch": 0.38692507165279105,
+      "grad_norm": 0.18595285713672638,
+      "learning_rate": 8e-05,
+      "loss": 1.4592,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3870615531595469,
+      "grad_norm": 0.1970224678516388,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3871980346663027,
+      "grad_norm": 0.20294629037380219,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 2837
+    },
+    {
+      "epoch": 0.38733451617305853,
+      "grad_norm": 0.193478062748909,
+      "learning_rate": 8e-05,
+      "loss": 1.4498,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3874709976798144,
+      "grad_norm": 0.20147845149040222,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 2839
+    },
+    {
+      "epoch": 0.38760747918657024,
+      "grad_norm": 0.1943940818309784,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 2840
+    },
+    {
+      "epoch": 0.38774396069332606,
+      "grad_norm": 0.20720873773097992,
+      "learning_rate": 8e-05,
+      "loss": 1.5658,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3878804422000819,
+      "grad_norm": 0.19790019094944,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3880169237068377,
+      "grad_norm": 0.1980382651090622,
+      "learning_rate": 8e-05,
+      "loss": 1.5056,
+      "step": 2843
+    },
+    {
+      "epoch": 0.38815340521359354,
+      "grad_norm": 0.20592498779296875,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3882898867203494,
+      "grad_norm": 0.20602920651435852,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 2845
+    },
+    {
+      "epoch": 0.38842636822710525,
+      "grad_norm": 0.19017846882343292,
+      "learning_rate": 8e-05,
+      "loss": 1.4984,
+      "step": 2846
+    },
+    {
+      "epoch": 0.38856284973386107,
+      "grad_norm": 0.202789768576622,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 2847
+    },
+    {
+      "epoch": 0.3886993312406169,
+      "grad_norm": 0.1992550641298294,
+      "learning_rate": 8e-05,
+      "loss": 1.4842,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3888358127473727,
+      "grad_norm": 0.19333533942699432,
+      "learning_rate": 8e-05,
+      "loss": 1.4712,
+      "step": 2849
+    },
+    {
+      "epoch": 0.38897229425412855,
+      "grad_norm": 0.2050512284040451,
+      "learning_rate": 8e-05,
+      "loss": 1.5065,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3891087757608844,
+      "grad_norm": 0.20399445295333862,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2851
+    },
+    {
+      "epoch": 0.38924525726764025,
+      "grad_norm": 0.1994365155696869,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3893817387743961,
+      "grad_norm": 0.1944725066423416,
+      "learning_rate": 8e-05,
+      "loss": 1.5247,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3895182202811519,
+      "grad_norm": 0.20120403170585632,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 2854
+    },
+    {
+      "epoch": 0.38965470178790773,
+      "grad_norm": 0.19254420697689056,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 2855
+    },
+    {
+      "epoch": 0.38979118329466356,
+      "grad_norm": 0.19902515411376953,
+      "learning_rate": 8e-05,
+      "loss": 1.514,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3899276648014194,
+      "grad_norm": 0.2014210969209671,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 2857
+    },
+    {
+      "epoch": 0.39006414630817526,
+      "grad_norm": 0.19931615889072418,
+      "learning_rate": 8e-05,
+      "loss": 1.5179,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3902006278149311,
+      "grad_norm": 0.19515490531921387,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3903371093216869,
+      "grad_norm": 0.217337965965271,
+      "learning_rate": 8e-05,
+      "loss": 1.5818,
+      "step": 2860
+    },
+    {
+      "epoch": 0.39047359082844274,
+      "grad_norm": 0.19456127285957336,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 2861
+    },
+    {
+      "epoch": 0.39061007233519857,
+      "grad_norm": 0.19887696206569672,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3907465538419544,
+      "grad_norm": 0.19483307003974915,
+      "learning_rate": 8e-05,
+      "loss": 1.4973,
+      "step": 2863
+    },
+    {
+      "epoch": 0.3908830353487103,
+      "grad_norm": 0.19631604850292206,
+      "learning_rate": 8e-05,
+      "loss": 1.4889,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3910195168554661,
+      "grad_norm": 0.19995272159576416,
+      "learning_rate": 8e-05,
+      "loss": 1.5028,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3911559983622219,
+      "grad_norm": 0.19609548151493073,
+      "learning_rate": 8e-05,
+      "loss": 1.4999,
+      "step": 2866
+    },
+    {
+      "epoch": 0.39129247986897775,
+      "grad_norm": 0.19483570754528046,
+      "learning_rate": 8e-05,
+      "loss": 1.4794,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3914289613757336,
+      "grad_norm": 0.20292863249778748,
+      "learning_rate": 8e-05,
+      "loss": 1.5696,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3915654428824894,
+      "grad_norm": 0.2035442590713501,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3917019243892453,
+      "grad_norm": 0.19722416996955872,
+      "learning_rate": 8e-05,
+      "loss": 1.5133,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3918384058960011,
+      "grad_norm": 0.1925741732120514,
+      "learning_rate": 8e-05,
+      "loss": 1.4329,
+      "step": 2871
+    },
+    {
+      "epoch": 0.39197488740275693,
+      "grad_norm": 0.1932852864265442,
+      "learning_rate": 8e-05,
+      "loss": 1.5362,
+      "step": 2872
+    },
+    {
+      "epoch": 0.39211136890951276,
+      "grad_norm": 0.21710823476314545,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3922478504162686,
+      "grad_norm": 0.19802843034267426,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3923843319230244,
+      "grad_norm": 0.19122307002544403,
+      "learning_rate": 8e-05,
+      "loss": 1.4646,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3925208134297803,
+      "grad_norm": 0.1979711800813675,
+      "learning_rate": 8e-05,
+      "loss": 1.4536,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3926572949365361,
+      "grad_norm": 0.19692827761173248,
+      "learning_rate": 8e-05,
+      "loss": 1.4841,
+      "step": 2877
+    },
+    {
+      "epoch": 0.39279377644329194,
+      "grad_norm": 0.2021845132112503,
+      "learning_rate": 8e-05,
+      "loss": 1.5376,
+      "step": 2878
+    },
+    {
+      "epoch": 0.39293025795004777,
+      "grad_norm": 0.19492624700069427,
+      "learning_rate": 8e-05,
+      "loss": 1.5264,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3930667394568036,
+      "grad_norm": 0.1994514763355255,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3932032209635594,
+      "grad_norm": 0.19496627151966095,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3933397024703153,
+      "grad_norm": 0.19734826683998108,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3934761839770711,
+      "grad_norm": 0.200940802693367,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 2883
+    },
+    {
+      "epoch": 0.39361266548382695,
+      "grad_norm": 0.1994246244430542,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3937491469905828,
+      "grad_norm": 0.19973237812519073,
+      "learning_rate": 8e-05,
+      "loss": 1.5419,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3938856284973386,
+      "grad_norm": 0.20020098984241486,
+      "learning_rate": 8e-05,
+      "loss": 1.4667,
+      "step": 2886
+    },
+    {
+      "epoch": 0.39402211000409443,
+      "grad_norm": 0.20032110810279846,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 2887
+    },
+    {
+      "epoch": 0.39415859151085025,
+      "grad_norm": 0.197679340839386,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 2888
+    },
+    {
+      "epoch": 0.39429507301760613,
+      "grad_norm": 0.19777023792266846,
+      "learning_rate": 8e-05,
+      "loss": 1.5246,
+      "step": 2889
+    },
+    {
+      "epoch": 0.39443155452436196,
+      "grad_norm": 0.18900208175182343,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3945680360311178,
+      "grad_norm": 0.20367109775543213,
+      "learning_rate": 8e-05,
+      "loss": 1.4749,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3947045175378736,
+      "grad_norm": 0.19022367894649506,
+      "learning_rate": 8e-05,
+      "loss": 1.4721,
+      "step": 2892
+    },
+    {
+      "epoch": 0.39484099904462944,
+      "grad_norm": 0.1938842386007309,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 2893
+    },
+    {
+      "epoch": 0.39497748055138526,
+      "grad_norm": 0.19535568356513977,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 2894
+    },
+    {
+      "epoch": 0.39511396205814114,
+      "grad_norm": 0.19387255609035492,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2895
+    },
+    {
+      "epoch": 0.39525044356489697,
+      "grad_norm": 0.1997697800397873,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3953869250716528,
+      "grad_norm": 0.19633899629116058,
+      "learning_rate": 8e-05,
+      "loss": 1.5448,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3955234065784086,
+      "grad_norm": 0.19447049498558044,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 2898
+    },
+    {
+      "epoch": 0.39565988808516445,
+      "grad_norm": 0.19827061891555786,
+      "learning_rate": 8e-05,
+      "loss": 1.5021,
+      "step": 2899
+    },
+    {
+      "epoch": 0.39579636959192027,
+      "grad_norm": 0.2008560746908188,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 2900
+    },
+    {
+      "epoch": 0.39593285109867615,
+      "grad_norm": 0.20060548186302185,
+      "learning_rate": 8e-05,
+      "loss": 1.551,
+      "step": 2901
+    },
+    {
+      "epoch": 0.396069332605432,
+      "grad_norm": 0.20607057213783264,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3962058141121878,
+      "grad_norm": 0.19841592013835907,
+      "learning_rate": 8e-05,
+      "loss": 1.4703,
+      "step": 2903
+    },
+    {
+      "epoch": 0.39634229561894363,
+      "grad_norm": 0.20533297955989838,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 2904
+    },
+    {
+      "epoch": 0.39647877712569946,
+      "grad_norm": 0.19957555830478668,
+      "learning_rate": 8e-05,
+      "loss": 1.4912,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3966152586324553,
+      "grad_norm": 0.20700012147426605,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 2906
+    },
+    {
+      "epoch": 0.39675174013921116,
+      "grad_norm": 0.1938125193119049,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 2907
+    },
+    {
+      "epoch": 0.396888221645967,
+      "grad_norm": 0.1980900913476944,
+      "learning_rate": 8e-05,
+      "loss": 1.5737,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3970247031527228,
+      "grad_norm": 0.1976432502269745,
+      "learning_rate": 8e-05,
+      "loss": 1.5376,
+      "step": 2909
+    },
+    {
+      "epoch": 0.39716118465947864,
+      "grad_norm": 0.19101516902446747,
+      "learning_rate": 8e-05,
+      "loss": 1.4588,
+      "step": 2910
+    },
+    {
+      "epoch": 0.39729766616623446,
+      "grad_norm": 0.20482368767261505,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3974341476729903,
+      "grad_norm": 0.19892390072345734,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 2912
+    },
+    {
+      "epoch": 0.39757062917974617,
+      "grad_norm": 0.1995740383863449,
+      "learning_rate": 8e-05,
+      "loss": 1.5065,
+      "step": 2913
+    },
+    {
+      "epoch": 0.397707110686502,
+      "grad_norm": 0.20146548748016357,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3978435921932578,
+      "grad_norm": 0.19357001781463623,
+      "learning_rate": 8e-05,
+      "loss": 1.4879,
+      "step": 2915
+    },
+    {
+      "epoch": 0.39798007370001365,
+      "grad_norm": 0.19232109189033508,
+      "learning_rate": 8e-05,
+      "loss": 1.5136,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3981165552067695,
+      "grad_norm": 0.20586858689785004,
+      "learning_rate": 8e-05,
+      "loss": 1.5242,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3982530367135253,
+      "grad_norm": 0.2016007900238037,
+      "learning_rate": 8e-05,
+      "loss": 1.4635,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3983895182202811,
+      "grad_norm": 0.1968727707862854,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 2919
+    },
+    {
+      "epoch": 0.398525999727037,
+      "grad_norm": 0.1950712502002716,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 2920
+    },
+    {
+      "epoch": 0.39866248123379283,
+      "grad_norm": 0.19620060920715332,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 2921
+    },
+    {
+      "epoch": 0.39879896274054866,
+      "grad_norm": 0.19629189372062683,
+      "learning_rate": 8e-05,
+      "loss": 1.4995,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3989354442473045,
+      "grad_norm": 0.18881890177726746,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3990719257540603,
+      "grad_norm": 0.2067987322807312,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 2924
+    },
+    {
+      "epoch": 0.39920840726081613,
+      "grad_norm": 0.20668795704841614,
+      "learning_rate": 8e-05,
+      "loss": 1.5181,
+      "step": 2925
+    },
+    {
+      "epoch": 0.399344888767572,
+      "grad_norm": 0.1990017592906952,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 2926
+    },
+    {
+      "epoch": 0.39948137027432784,
+      "grad_norm": 0.21066772937774658,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 2927
+    },
+    {
+      "epoch": 0.39961785178108367,
+      "grad_norm": 0.2047015279531479,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3997543332878395,
+      "grad_norm": 0.20325912535190582,
+      "learning_rate": 8e-05,
+      "loss": 1.5252,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3998908147945953,
+      "grad_norm": 0.19412583112716675,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 2930
+    },
+    {
+      "epoch": 0.40002729630135114,
+      "grad_norm": 0.18911629915237427,
+      "learning_rate": 8e-05,
+      "loss": 1.4648,
+      "step": 2931
+    },
+    {
+      "epoch": 0.400163777808107,
+      "grad_norm": 0.19966891407966614,
+      "learning_rate": 8e-05,
+      "loss": 1.4965,
+      "step": 2932
+    },
+    {
+      "epoch": 0.40030025931486285,
+      "grad_norm": 0.1976897418498993,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 2933
+    },
+    {
+      "epoch": 0.4004367408216187,
+      "grad_norm": 0.20238204300403595,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4005732223283745,
+      "grad_norm": 0.19850894808769226,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 2935
+    },
+    {
+      "epoch": 0.4007097038351303,
+      "grad_norm": 0.19660159945487976,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 2936
+    },
+    {
+      "epoch": 0.40084618534188615,
+      "grad_norm": 0.19738981127738953,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 2937
+    },
+    {
+      "epoch": 0.40098266684864203,
+      "grad_norm": 0.20087642967700958,
+      "learning_rate": 8e-05,
+      "loss": 1.532,
+      "step": 2938
+    },
+    {
+      "epoch": 0.40111914835539786,
+      "grad_norm": 0.2034512609243393,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 2939
+    },
+    {
+      "epoch": 0.4012556298621537,
+      "grad_norm": 0.2034153938293457,
+      "learning_rate": 8e-05,
+      "loss": 1.5602,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4013921113689095,
+      "grad_norm": 0.2082194983959198,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 2941
+    },
+    {
+      "epoch": 0.40152859287566534,
+      "grad_norm": 0.2005544900894165,
+      "learning_rate": 8e-05,
+      "loss": 1.5428,
+      "step": 2942
+    },
+    {
+      "epoch": 0.40166507438242116,
+      "grad_norm": 0.1988706886768341,
+      "learning_rate": 8e-05,
+      "loss": 1.5231,
+      "step": 2943
+    },
+    {
+      "epoch": 0.40180155588917704,
+      "grad_norm": 0.19264404475688934,
+      "learning_rate": 8e-05,
+      "loss": 1.479,
+      "step": 2944
+    },
+    {
+      "epoch": 0.40193803739593287,
+      "grad_norm": 0.1900002509355545,
+      "learning_rate": 8e-05,
+      "loss": 1.4628,
+      "step": 2945
+    },
+    {
+      "epoch": 0.4020745189026887,
+      "grad_norm": 0.19762036204338074,
+      "learning_rate": 8e-05,
+      "loss": 1.5035,
+      "step": 2946
+    },
+    {
+      "epoch": 0.4022110004094445,
+      "grad_norm": 0.20186229050159454,
+      "learning_rate": 8e-05,
+      "loss": 1.5615,
+      "step": 2947
+    },
+    {
+      "epoch": 0.40234748191620034,
+      "grad_norm": 0.19748632609844208,
+      "learning_rate": 8e-05,
+      "loss": 1.5256,
+      "step": 2948
+    },
+    {
+      "epoch": 0.40248396342295617,
+      "grad_norm": 0.198756605386734,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 2949
+    },
+    {
+      "epoch": 0.40262044492971205,
+      "grad_norm": 0.21259327232837677,
+      "learning_rate": 8e-05,
+      "loss": 1.5608,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4027569264364679,
+      "grad_norm": 0.2042684704065323,
+      "learning_rate": 8e-05,
+      "loss": 1.544,
+      "step": 2951
+    },
+    {
+      "epoch": 0.4028934079432237,
+      "grad_norm": 0.2009604573249817,
+      "learning_rate": 8e-05,
+      "loss": 1.5201,
+      "step": 2952
+    },
+    {
+      "epoch": 0.40302988944997953,
+      "grad_norm": 0.1993720531463623,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 2953
+    },
+    {
+      "epoch": 0.40316637095673535,
+      "grad_norm": 0.20310159027576447,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 2954
+    },
+    {
+      "epoch": 0.4033028524634912,
+      "grad_norm": 0.19351494312286377,
+      "learning_rate": 8e-05,
+      "loss": 1.501,
+      "step": 2955
+    },
+    {
+      "epoch": 0.403439333970247,
+      "grad_norm": 0.20704621076583862,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 2956
+    },
+    {
+      "epoch": 0.4035758154770029,
+      "grad_norm": 0.19627414643764496,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4037122969837587,
+      "grad_norm": 0.1970394402742386,
+      "learning_rate": 8e-05,
+      "loss": 1.5328,
+      "step": 2958
+    },
+    {
+      "epoch": 0.40384877849051454,
+      "grad_norm": 0.2014622986316681,
+      "learning_rate": 8e-05,
+      "loss": 1.5313,
+      "step": 2959
+    },
+    {
+      "epoch": 0.40398525999727036,
+      "grad_norm": 0.20270584523677826,
+      "learning_rate": 8e-05,
+      "loss": 1.4908,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4041217415040262,
+      "grad_norm": 0.19967079162597656,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 2961
+    },
+    {
+      "epoch": 0.404258223010782,
+      "grad_norm": 0.21107113361358643,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 2962
+    },
+    {
+      "epoch": 0.4043947045175379,
+      "grad_norm": 0.19966532289981842,
+      "learning_rate": 8e-05,
+      "loss": 1.4921,
+      "step": 2963
+    },
+    {
+      "epoch": 0.4045311860242937,
+      "grad_norm": 0.1987009048461914,
+      "learning_rate": 8e-05,
+      "loss": 1.5128,
+      "step": 2964
+    },
+    {
+      "epoch": 0.40466766753104955,
+      "grad_norm": 0.1970715969800949,
+      "learning_rate": 8e-05,
+      "loss": 1.5314,
+      "step": 2965
+    },
+    {
+      "epoch": 0.40480414903780537,
+      "grad_norm": 0.1987844705581665,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4049406305445612,
+      "grad_norm": 0.2008032202720642,
+      "learning_rate": 8e-05,
+      "loss": 1.566,
+      "step": 2967
+    },
+    {
+      "epoch": 0.405077112051317,
+      "grad_norm": 0.2003677785396576,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 2968
+    },
+    {
+      "epoch": 0.4052135935580729,
+      "grad_norm": 0.19556944072246552,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 2969
+    },
+    {
+      "epoch": 0.40535007506482873,
+      "grad_norm": 0.1971994936466217,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 2970
+    },
+    {
+      "epoch": 0.40548655657158456,
+      "grad_norm": 0.1930166780948639,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 2971
+    },
+    {
+      "epoch": 0.4056230380783404,
+      "grad_norm": 0.19386087357997894,
+      "learning_rate": 8e-05,
+      "loss": 1.4588,
+      "step": 2972
+    },
+    {
+      "epoch": 0.4057595195850962,
+      "grad_norm": 0.2058456838130951,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 2973
+    },
+    {
+      "epoch": 0.40589600109185203,
+      "grad_norm": 0.19446761906147003,
+      "learning_rate": 8e-05,
+      "loss": 1.4449,
+      "step": 2974
+    },
+    {
+      "epoch": 0.4060324825986079,
+      "grad_norm": 0.1961190402507782,
+      "learning_rate": 8e-05,
+      "loss": 1.4753,
+      "step": 2975
+    },
+    {
+      "epoch": 0.40616896410536374,
+      "grad_norm": 0.20437094569206238,
+      "learning_rate": 8e-05,
+      "loss": 1.6177,
+      "step": 2976
+    },
+    {
+      "epoch": 0.40630544561211956,
+      "grad_norm": 0.19437143206596375,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 2977
+    },
+    {
+      "epoch": 0.4064419271188754,
+      "grad_norm": 0.19836843013763428,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 2978
+    },
+    {
+      "epoch": 0.4065784086256312,
+      "grad_norm": 0.206318661570549,
+      "learning_rate": 8e-05,
+      "loss": 1.5858,
+      "step": 2979
+    },
+    {
+      "epoch": 0.40671489013238704,
+      "grad_norm": 0.1988915205001831,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4068513716391429,
+      "grad_norm": 0.20537327229976654,
+      "learning_rate": 8e-05,
+      "loss": 1.5248,
+      "step": 2981
+    },
+    {
+      "epoch": 0.40698785314589875,
+      "grad_norm": 0.19781246781349182,
+      "learning_rate": 8e-05,
+      "loss": 1.5081,
+      "step": 2982
+    },
+    {
+      "epoch": 0.4071243346526546,
+      "grad_norm": 0.20263095200061798,
+      "learning_rate": 8e-05,
+      "loss": 1.514,
+      "step": 2983
+    },
+    {
+      "epoch": 0.4072608161594104,
+      "grad_norm": 0.20989759266376495,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4073972976661662,
+      "grad_norm": 0.1970755010843277,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 2985
+    },
+    {
+      "epoch": 0.40753377917292205,
+      "grad_norm": 0.2051224708557129,
+      "learning_rate": 8e-05,
+      "loss": 1.5323,
+      "step": 2986
+    },
+    {
+      "epoch": 0.4076702606796779,
+      "grad_norm": 0.2038838118314743,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 2987
+    },
+    {
+      "epoch": 0.40780674218643376,
+      "grad_norm": 0.20259542763233185,
+      "learning_rate": 8e-05,
+      "loss": 1.5166,
+      "step": 2988
+    },
+    {
+      "epoch": 0.4079432236931896,
+      "grad_norm": 0.2053993046283722,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4080797051999454,
+      "grad_norm": 0.20370444655418396,
+      "learning_rate": 8e-05,
+      "loss": 1.5167,
+      "step": 2990
+    },
+    {
+      "epoch": 0.40821618670670123,
+      "grad_norm": 0.20440323650836945,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 2991
+    },
+    {
+      "epoch": 0.40835266821345706,
+      "grad_norm": 0.199214905500412,
+      "learning_rate": 8e-05,
+      "loss": 1.5572,
+      "step": 2992
+    },
+    {
+      "epoch": 0.4084891497202129,
+      "grad_norm": 0.20004010200500488,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 2993
+    },
+    {
+      "epoch": 0.40862563122696877,
+      "grad_norm": 0.19804321229457855,
+      "learning_rate": 8e-05,
+      "loss": 1.5475,
+      "step": 2994
+    },
+    {
+      "epoch": 0.4087621127337246,
+      "grad_norm": 0.19671191275119781,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4088985942404804,
+      "grad_norm": 0.19805675745010376,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 2996
+    },
+    {
+      "epoch": 0.40903507574723624,
+      "grad_norm": 0.19692322611808777,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 2997
+    },
+    {
+      "epoch": 0.40917155725399207,
+      "grad_norm": 0.20234470069408417,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4093080387607479,
+      "grad_norm": 0.1951553076505661,
+      "learning_rate": 8e-05,
+      "loss": 1.5432,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4094445202675038,
+      "grad_norm": 0.20264188945293427,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4095810017742596,
+      "grad_norm": 0.20976346731185913,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 3001
+    },
+    {
+      "epoch": 0.4097174832810154,
+      "grad_norm": 0.19725334644317627,
+      "learning_rate": 8e-05,
+      "loss": 1.4706,
+      "step": 3002
+    },
+    {
+      "epoch": 0.40985396478777125,
+      "grad_norm": 0.19273890554904938,
+      "learning_rate": 8e-05,
+      "loss": 1.5066,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4099904462945271,
+      "grad_norm": 0.1956668496131897,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 3004
+    },
+    {
+      "epoch": 0.4101269278012829,
+      "grad_norm": 0.20118993520736694,
+      "learning_rate": 8e-05,
+      "loss": 1.4773,
+      "step": 3005
+    },
+    {
+      "epoch": 0.4102634093080388,
+      "grad_norm": 0.20344021916389465,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 3006
+    },
+    {
+      "epoch": 0.4103998908147946,
+      "grad_norm": 0.19675013422966003,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 3007
+    },
+    {
+      "epoch": 0.41053637232155044,
+      "grad_norm": 0.2034749686717987,
+      "learning_rate": 8e-05,
+      "loss": 1.5023,
+      "step": 3008
+    },
+    {
+      "epoch": 0.41067285382830626,
+      "grad_norm": 0.20168781280517578,
+      "learning_rate": 8e-05,
+      "loss": 1.5403,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4108093353350621,
+      "grad_norm": 0.19701792299747467,
+      "learning_rate": 8e-05,
+      "loss": 1.424,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4109458168418179,
+      "grad_norm": 0.21370914578437805,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 3011
+    },
+    {
+      "epoch": 0.4110822983485738,
+      "grad_norm": 0.19638139009475708,
+      "learning_rate": 8e-05,
+      "loss": 1.4563,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4112187798553296,
+      "grad_norm": 0.21868322789669037,
+      "learning_rate": 8e-05,
+      "loss": 1.5132,
+      "step": 3013
+    },
+    {
+      "epoch": 0.41135526136208544,
+      "grad_norm": 0.20473575592041016,
+      "learning_rate": 8e-05,
+      "loss": 1.4399,
+      "step": 3014
+    },
+    {
+      "epoch": 0.41149174286884127,
+      "grad_norm": 0.20193174481391907,
+      "learning_rate": 8e-05,
+      "loss": 1.5402,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4116282243755971,
+      "grad_norm": 0.21881115436553955,
+      "learning_rate": 8e-05,
+      "loss": 1.5054,
+      "step": 3016
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 0.2003670632839203,
+      "learning_rate": 8e-05,
+      "loss": 1.5642,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4119011873891088,
+      "grad_norm": 0.1960163414478302,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 3018
+    },
+    {
+      "epoch": 0.41203766889586463,
+      "grad_norm": 0.19543088972568512,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 3019
+    },
+    {
+      "epoch": 0.41217415040262045,
+      "grad_norm": 0.20008787512779236,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4123106319093763,
+      "grad_norm": 0.19972200691699982,
+      "learning_rate": 8e-05,
+      "loss": 1.5241,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4124471134161321,
+      "grad_norm": 0.2094513475894928,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 3022
+    },
+    {
+      "epoch": 0.41258359492288793,
+      "grad_norm": 0.1957295536994934,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 3023
+    },
+    {
+      "epoch": 0.41272007642964376,
+      "grad_norm": 0.19742745161056519,
+      "learning_rate": 8e-05,
+      "loss": 1.408,
+      "step": 3024
+    },
+    {
+      "epoch": 0.41285655793639964,
+      "grad_norm": 0.20839804410934448,
+      "learning_rate": 8e-05,
+      "loss": 1.5112,
+      "step": 3025
+    },
+    {
+      "epoch": 0.41299303944315546,
+      "grad_norm": 0.19152940809726715,
+      "learning_rate": 8e-05,
+      "loss": 1.472,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4131295209499113,
+      "grad_norm": 0.2101462185382843,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4132660024566671,
+      "grad_norm": 0.21248017251491547,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 3028
+    },
+    {
+      "epoch": 0.41340248396342294,
+      "grad_norm": 0.2032170593738556,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3029
+    },
+    {
+      "epoch": 0.41353896547017877,
+      "grad_norm": 0.1994853913784027,
+      "learning_rate": 8e-05,
+      "loss": 1.5385,
+      "step": 3030
+    },
+    {
+      "epoch": 0.41367544697693465,
+      "grad_norm": 0.19901813566684723,
+      "learning_rate": 8e-05,
+      "loss": 1.5269,
+      "step": 3031
+    },
+    {
+      "epoch": 0.41381192848369047,
+      "grad_norm": 0.19795392453670502,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4139484099904463,
+      "grad_norm": 0.1975666731595993,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4140848914972021,
+      "grad_norm": 0.20246610045433044,
+      "learning_rate": 8e-05,
+      "loss": 1.5156,
+      "step": 3034
+    },
+    {
+      "epoch": 0.41422137300395795,
+      "grad_norm": 0.20654650032520294,
+      "learning_rate": 8e-05,
+      "loss": 1.5134,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4143578545107138,
+      "grad_norm": 0.19502303004264832,
+      "learning_rate": 8e-05,
+      "loss": 1.4803,
+      "step": 3036
+    },
+    {
+      "epoch": 0.41449433601746966,
+      "grad_norm": 0.1982269287109375,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4146308175242255,
+      "grad_norm": 0.202792227268219,
+      "learning_rate": 8e-05,
+      "loss": 1.524,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4147672990309813,
+      "grad_norm": 0.19526930153369904,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 3039
+    },
+    {
+      "epoch": 0.41490378053773713,
+      "grad_norm": 0.19401291012763977,
+      "learning_rate": 8e-05,
+      "loss": 1.4669,
+      "step": 3040
+    },
+    {
+      "epoch": 0.41504026204449296,
+      "grad_norm": 0.20006267726421356,
+      "learning_rate": 8e-05,
+      "loss": 1.5017,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4151767435512488,
+      "grad_norm": 0.20156654715538025,
+      "learning_rate": 8e-05,
+      "loss": 1.5939,
+      "step": 3042
+    },
+    {
+      "epoch": 0.41531322505800466,
+      "grad_norm": 0.20333774387836456,
+      "learning_rate": 8e-05,
+      "loss": 1.4993,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4154497065647605,
+      "grad_norm": 0.20235462486743927,
+      "learning_rate": 8e-05,
+      "loss": 1.5341,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4155861880715163,
+      "grad_norm": 0.19902227818965912,
+      "learning_rate": 8e-05,
+      "loss": 1.488,
+      "step": 3045
+    },
+    {
+      "epoch": 0.41572266957827214,
+      "grad_norm": 0.19731463491916656,
+      "learning_rate": 8e-05,
+      "loss": 1.5076,
+      "step": 3046
+    },
+    {
+      "epoch": 0.41585915108502797,
+      "grad_norm": 0.20182174444198608,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4159956325917838,
+      "grad_norm": 0.19677038490772247,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4161321140985397,
+      "grad_norm": 0.2003851681947708,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4162685956052955,
+      "grad_norm": 0.2017872929573059,
+      "learning_rate": 8e-05,
+      "loss": 1.5434,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4164050771120513,
+      "grad_norm": 0.18890967965126038,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3051
+    },
+    {
+      "epoch": 0.41654155861880715,
+      "grad_norm": 0.20344127714633942,
+      "learning_rate": 8e-05,
+      "loss": 1.5306,
+      "step": 3052
+    },
+    {
+      "epoch": 0.416678040125563,
+      "grad_norm": 0.1985037922859192,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4168145216323188,
+      "grad_norm": 0.20234611630439758,
+      "learning_rate": 8e-05,
+      "loss": 1.5208,
+      "step": 3054
+    },
+    {
+      "epoch": 0.4169510031390747,
+      "grad_norm": 0.20751430094242096,
+      "learning_rate": 8e-05,
+      "loss": 1.5102,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4170874846458305,
+      "grad_norm": 0.19542154669761658,
+      "learning_rate": 8e-05,
+      "loss": 1.5077,
+      "step": 3056
+    },
+    {
+      "epoch": 0.41722396615258633,
+      "grad_norm": 0.2110983282327652,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 3057
+    },
+    {
+      "epoch": 0.41736044765934216,
+      "grad_norm": 0.2074403464794159,
+      "learning_rate": 8e-05,
+      "loss": 1.467,
+      "step": 3058
+    },
+    {
+      "epoch": 0.417496929166098,
+      "grad_norm": 0.19831030070781708,
+      "learning_rate": 8e-05,
+      "loss": 1.5048,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4176334106728538,
+      "grad_norm": 0.20088212192058563,
+      "learning_rate": 8e-05,
+      "loss": 1.4959,
+      "step": 3060
+    },
+    {
+      "epoch": 0.41776989217960964,
+      "grad_norm": 0.2006976455450058,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4179063736863655,
+      "grad_norm": 0.20028850436210632,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 3062
+    },
+    {
+      "epoch": 0.41804285519312134,
+      "grad_norm": 0.20760108530521393,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 3063
+    },
+    {
+      "epoch": 0.41817933669987717,
+      "grad_norm": 0.20036286115646362,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 3064
+    },
+    {
+      "epoch": 0.418315818206633,
+      "grad_norm": 0.20541973412036896,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4184522997133888,
+      "grad_norm": 0.2129804939031601,
+      "learning_rate": 8e-05,
+      "loss": 1.5762,
+      "step": 3066
+    },
+    {
+      "epoch": 0.41858878122014465,
+      "grad_norm": 0.2024850994348526,
+      "learning_rate": 8e-05,
+      "loss": 1.4776,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4187252627269005,
+      "grad_norm": 0.1978427767753601,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 3068
+    },
+    {
+      "epoch": 0.41886174423365635,
+      "grad_norm": 0.2182529717683792,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4189982257404122,
+      "grad_norm": 0.2051420360803604,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 3070
+    },
+    {
+      "epoch": 0.419134707247168,
+      "grad_norm": 0.18921470642089844,
+      "learning_rate": 8e-05,
+      "loss": 1.3556,
+      "step": 3071
+    },
+    {
+      "epoch": 0.41927118875392383,
+      "grad_norm": 0.19818145036697388,
+      "learning_rate": 8e-05,
+      "loss": 1.473,
+      "step": 3072
+    },
+    {
+      "epoch": 0.41940767026067965,
+      "grad_norm": 0.20514588057994843,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 3073
+    },
+    {
+      "epoch": 0.41954415176743554,
+      "grad_norm": 0.2012423723936081,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 3074
+    },
+    {
+      "epoch": 0.41968063327419136,
+      "grad_norm": 0.19921784102916718,
+      "learning_rate": 8e-05,
+      "loss": 1.5449,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4198171147809472,
+      "grad_norm": 0.2028958648443222,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 3076
+    },
+    {
+      "epoch": 0.419953596287703,
+      "grad_norm": 0.19983601570129395,
+      "learning_rate": 8e-05,
+      "loss": 1.4525,
+      "step": 3077
+    },
+    {
+      "epoch": 0.42009007779445884,
+      "grad_norm": 0.19572323560714722,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 3078
+    },
+    {
+      "epoch": 0.42022655930121466,
+      "grad_norm": 0.19779036939144135,
+      "learning_rate": 8e-05,
+      "loss": 1.5401,
+      "step": 3079
+    },
+    {
+      "epoch": 0.42036304080797054,
+      "grad_norm": 0.20040729641914368,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 3080
+    },
+    {
+      "epoch": 0.42049952231472637,
+      "grad_norm": 0.19912153482437134,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4206360038214822,
+      "grad_norm": 0.18947699666023254,
+      "learning_rate": 8e-05,
+      "loss": 1.369,
+      "step": 3082
+    },
+    {
+      "epoch": 0.420772485328238,
+      "grad_norm": 0.20511123538017273,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 3083
+    },
+    {
+      "epoch": 0.42090896683499385,
+      "grad_norm": 0.20144593715667725,
+      "learning_rate": 8e-05,
+      "loss": 1.5145,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4210454483417497,
+      "grad_norm": 0.2135966420173645,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 3085
+    },
+    {
+      "epoch": 0.42118192984850555,
+      "grad_norm": 0.20007885992527008,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4213184113552614,
+      "grad_norm": 0.2059982866048813,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4214548928620172,
+      "grad_norm": 0.20596514642238617,
+      "learning_rate": 8e-05,
+      "loss": 1.5241,
+      "step": 3088
+    },
+    {
+      "epoch": 0.42159137436877303,
+      "grad_norm": 0.19343741238117218,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 3089
+    },
+    {
+      "epoch": 0.42172785587552886,
+      "grad_norm": 0.21002456545829773,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4218643373822847,
+      "grad_norm": 0.19752705097198486,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4220008188890405,
+      "grad_norm": 0.19870814681053162,
+      "learning_rate": 8e-05,
+      "loss": 1.5724,
+      "step": 3092
+    },
+    {
+      "epoch": 0.4221373003957964,
+      "grad_norm": 0.20339490473270416,
+      "learning_rate": 8e-05,
+      "loss": 1.5534,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4222737819025522,
+      "grad_norm": 0.19840386509895325,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 3094
+    },
+    {
+      "epoch": 0.42241026340930804,
+      "grad_norm": 0.19879502058029175,
+      "learning_rate": 8e-05,
+      "loss": 1.5096,
+      "step": 3095
+    },
+    {
+      "epoch": 0.42254674491606387,
+      "grad_norm": 0.20178432762622833,
+      "learning_rate": 8e-05,
+      "loss": 1.5249,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4226832264228197,
+      "grad_norm": 0.20104891061782837,
+      "learning_rate": 8e-05,
+      "loss": 1.527,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4228197079295755,
+      "grad_norm": 0.2035837173461914,
+      "learning_rate": 8e-05,
+      "loss": 1.4853,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4229561894363314,
+      "grad_norm": 0.19916000962257385,
+      "learning_rate": 8e-05,
+      "loss": 1.469,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4230926709430872,
+      "grad_norm": 0.20036402344703674,
+      "learning_rate": 8e-05,
+      "loss": 1.4276,
+      "step": 3100
+    },
+    {
+      "epoch": 0.42322915244984305,
+      "grad_norm": 0.21880033612251282,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4233656339565989,
+      "grad_norm": 0.19910287857055664,
+      "learning_rate": 8e-05,
+      "loss": 1.5119,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4235021154633547,
+      "grad_norm": 0.1978716254234314,
+      "learning_rate": 8e-05,
+      "loss": 1.4507,
+      "step": 3103
+    },
+    {
+      "epoch": 0.4236385969701105,
+      "grad_norm": 0.20327045023441315,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4237750784768664,
+      "grad_norm": 0.1991620510816574,
+      "learning_rate": 8e-05,
+      "loss": 1.4018,
+      "step": 3105
+    },
+    {
+      "epoch": 0.42391155998362223,
+      "grad_norm": 0.2157338559627533,
+      "learning_rate": 8e-05,
+      "loss": 1.5252,
+      "step": 3106
+    },
+    {
+      "epoch": 0.42404804149037806,
+      "grad_norm": 0.197748064994812,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4241845229971339,
+      "grad_norm": 0.19764789938926697,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4243210045038897,
+      "grad_norm": 0.19249731302261353,
+      "learning_rate": 8e-05,
+      "loss": 1.4729,
+      "step": 3109
+    },
+    {
+      "epoch": 0.42445748601064553,
+      "grad_norm": 0.200058713555336,
+      "learning_rate": 8e-05,
+      "loss": 1.5313,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4245939675174014,
+      "grad_norm": 0.20205599069595337,
+      "learning_rate": 8e-05,
+      "loss": 1.507,
+      "step": 3111
+    },
+    {
+      "epoch": 0.42473044902415724,
+      "grad_norm": 0.20448777079582214,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 3112
+    },
+    {
+      "epoch": 0.42486693053091307,
+      "grad_norm": 0.20205707848072052,
+      "learning_rate": 8e-05,
+      "loss": 1.5148,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4250034120376689,
+      "grad_norm": 0.20858852565288544,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4251398935444247,
+      "grad_norm": 0.20042209327220917,
+      "learning_rate": 8e-05,
+      "loss": 1.5403,
+      "step": 3115
+    },
+    {
+      "epoch": 0.42527637505118054,
+      "grad_norm": 0.19518731534481049,
+      "learning_rate": 8e-05,
+      "loss": 1.5434,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4254128565579364,
+      "grad_norm": 0.19998310506343842,
+      "learning_rate": 8e-05,
+      "loss": 1.5566,
+      "step": 3117
+    },
+    {
+      "epoch": 0.42554933806469225,
+      "grad_norm": 0.19976671040058136,
+      "learning_rate": 8e-05,
+      "loss": 1.4858,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4256858195714481,
+      "grad_norm": 0.1955524981021881,
+      "learning_rate": 8e-05,
+      "loss": 1.4889,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4258223010782039,
+      "grad_norm": 0.209818497300148,
+      "learning_rate": 8e-05,
+      "loss": 1.5221,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4259587825849597,
+      "grad_norm": 0.19949398934841156,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 3121
+    },
+    {
+      "epoch": 0.42609526409171555,
+      "grad_norm": 0.1976507008075714,
+      "learning_rate": 8e-05,
+      "loss": 1.5398,
+      "step": 3122
+    },
+    {
+      "epoch": 0.42623174559847143,
+      "grad_norm": 0.19486378133296967,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 3123
+    },
+    {
+      "epoch": 0.42636822710522726,
+      "grad_norm": 0.19989556074142456,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4265047086119831,
+      "grad_norm": 0.19861209392547607,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4266411901187389,
+      "grad_norm": 0.19279126822948456,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 3126
+    },
+    {
+      "epoch": 0.42677767162549474,
+      "grad_norm": 0.19014571607112885,
+      "learning_rate": 8e-05,
+      "loss": 1.463,
+      "step": 3127
+    },
+    {
+      "epoch": 0.42691415313225056,
+      "grad_norm": 0.19941695034503937,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4270506346390064,
+      "grad_norm": 0.20443876087665558,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3129
+    },
+    {
+      "epoch": 0.42718711614576227,
+      "grad_norm": 0.20131316781044006,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4273235976525181,
+      "grad_norm": 0.1999267339706421,
+      "learning_rate": 8e-05,
+      "loss": 1.5957,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4274600791592739,
+      "grad_norm": 0.20583860576152802,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 3132
+    },
+    {
+      "epoch": 0.42759656066602975,
+      "grad_norm": 0.20933708548545837,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 3133
+    },
+    {
+      "epoch": 0.42773304217278557,
+      "grad_norm": 0.19596390426158905,
+      "learning_rate": 8e-05,
+      "loss": 1.4283,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4278695236795414,
+      "grad_norm": 0.2023884356021881,
+      "learning_rate": 8e-05,
+      "loss": 1.5254,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4280060051862973,
+      "grad_norm": 0.2096606343984604,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4281424866930531,
+      "grad_norm": 0.19961294531822205,
+      "learning_rate": 8e-05,
+      "loss": 1.5022,
+      "step": 3137
+    },
+    {
+      "epoch": 0.42827896819980893,
+      "grad_norm": 0.205382838845253,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 3138
+    },
+    {
+      "epoch": 0.42841544970656475,
+      "grad_norm": 0.21725013852119446,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4285519312133206,
+      "grad_norm": 0.19860441982746124,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4286884127200764,
+      "grad_norm": 0.20583562552928925,
+      "learning_rate": 8e-05,
+      "loss": 1.4873,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4288248942268323,
+      "grad_norm": 0.21374842524528503,
+      "learning_rate": 8e-05,
+      "loss": 1.4914,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4289613757335881,
+      "grad_norm": 0.20345541834831238,
+      "learning_rate": 8e-05,
+      "loss": 1.5,
+      "step": 3143
+    },
+    {
+      "epoch": 0.42909785724034394,
+      "grad_norm": 0.20387053489685059,
+      "learning_rate": 8e-05,
+      "loss": 1.4675,
+      "step": 3144
+    },
+    {
+      "epoch": 0.42923433874709976,
+      "grad_norm": 0.2053491473197937,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4293708202538556,
+      "grad_norm": 0.2038285881280899,
+      "learning_rate": 8e-05,
+      "loss": 1.4863,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4295073017606114,
+      "grad_norm": 0.19712214171886444,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4296437832673673,
+      "grad_norm": 0.20303410291671753,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4297802647741231,
+      "grad_norm": 0.2036440223455429,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 3149
+    },
+    {
+      "epoch": 0.42991674628087895,
+      "grad_norm": 0.2042427361011505,
+      "learning_rate": 8e-05,
+      "loss": 1.5281,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4300532277876348,
+      "grad_norm": 0.21023394167423248,
+      "learning_rate": 8e-05,
+      "loss": 1.5282,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4301897092943906,
+      "grad_norm": 0.19869254529476166,
+      "learning_rate": 8e-05,
+      "loss": 1.4866,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4303261908011464,
+      "grad_norm": 0.2015899270772934,
+      "learning_rate": 8e-05,
+      "loss": 1.5495,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4304626723079023,
+      "grad_norm": 0.20118282735347748,
+      "learning_rate": 8e-05,
+      "loss": 1.4714,
+      "step": 3154
+    },
+    {
+      "epoch": 0.43059915381465813,
+      "grad_norm": 0.2037975788116455,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 3155
+    },
+    {
+      "epoch": 0.43073563532141396,
+      "grad_norm": 0.19756707549095154,
+      "learning_rate": 8e-05,
+      "loss": 1.501,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4308721168281698,
+      "grad_norm": 0.20184838771820068,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4310085983349256,
+      "grad_norm": 0.20584118366241455,
+      "learning_rate": 8e-05,
+      "loss": 1.5376,
+      "step": 3158
+    },
+    {
+      "epoch": 0.43114507984168143,
+      "grad_norm": 0.2120056301355362,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 3159
+    },
+    {
+      "epoch": 0.43128156134843726,
+      "grad_norm": 0.20122769474983215,
+      "learning_rate": 8e-05,
+      "loss": 1.5268,
+      "step": 3160
+    },
+    {
+      "epoch": 0.43141804285519314,
+      "grad_norm": 0.20029057562351227,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 3161
+    },
+    {
+      "epoch": 0.43155452436194897,
+      "grad_norm": 0.21150970458984375,
+      "learning_rate": 8e-05,
+      "loss": 1.4944,
+      "step": 3162
+    },
+    {
+      "epoch": 0.4316910058687048,
+      "grad_norm": 0.20189321041107178,
+      "learning_rate": 8e-05,
+      "loss": 1.48,
+      "step": 3163
+    },
+    {
+      "epoch": 0.4318274873754606,
+      "grad_norm": 0.1946241557598114,
+      "learning_rate": 8e-05,
+      "loss": 1.4505,
+      "step": 3164
+    },
+    {
+      "epoch": 0.43196396888221644,
+      "grad_norm": 0.19614176452159882,
+      "learning_rate": 8e-05,
+      "loss": 1.4734,
+      "step": 3165
+    },
+    {
+      "epoch": 0.43210045038897227,
+      "grad_norm": 0.19670329988002777,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 3166
+    },
+    {
+      "epoch": 0.43223693189572815,
+      "grad_norm": 0.1972903311252594,
+      "learning_rate": 8e-05,
+      "loss": 1.5028,
+      "step": 3167
+    },
+    {
+      "epoch": 0.432373413402484,
+      "grad_norm": 0.208711639046669,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4325098949092398,
+      "grad_norm": 0.20804809033870697,
+      "learning_rate": 8e-05,
+      "loss": 1.5582,
+      "step": 3169
+    },
+    {
+      "epoch": 0.4326463764159956,
+      "grad_norm": 0.20560160279273987,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 3170
+    },
+    {
+      "epoch": 0.43278285792275145,
+      "grad_norm": 0.20082396268844604,
+      "learning_rate": 8e-05,
+      "loss": 1.5562,
+      "step": 3171
+    },
+    {
+      "epoch": 0.4329193394295073,
+      "grad_norm": 0.20229803025722504,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 3172
+    },
+    {
+      "epoch": 0.43305582093626316,
+      "grad_norm": 0.19297587871551514,
+      "learning_rate": 8e-05,
+      "loss": 1.4546,
+      "step": 3173
+    },
+    {
+      "epoch": 0.433192302443019,
+      "grad_norm": 0.20060156285762787,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4333287839497748,
+      "grad_norm": 0.20060162246227264,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 3175
+    },
+    {
+      "epoch": 0.43346526545653064,
+      "grad_norm": 0.1981133371591568,
+      "learning_rate": 8e-05,
+      "loss": 1.4664,
+      "step": 3176
+    },
+    {
+      "epoch": 0.43360174696328646,
+      "grad_norm": 0.20163919031620026,
+      "learning_rate": 8e-05,
+      "loss": 1.4975,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4337382284700423,
+      "grad_norm": 0.19982893764972687,
+      "learning_rate": 8e-05,
+      "loss": 1.4712,
+      "step": 3178
+    },
+    {
+      "epoch": 0.43387470997679817,
+      "grad_norm": 0.19722945988178253,
+      "learning_rate": 8e-05,
+      "loss": 1.5226,
+      "step": 3179
+    },
+    {
+      "epoch": 0.434011191483554,
+      "grad_norm": 0.19788241386413574,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4341476729903098,
+      "grad_norm": 0.19868123531341553,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 3181
+    },
+    {
+      "epoch": 0.43428415449706564,
+      "grad_norm": 0.1970725655555725,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 3182
+    },
+    {
+      "epoch": 0.43442063600382147,
+      "grad_norm": 0.21243047714233398,
+      "learning_rate": 8e-05,
+      "loss": 1.5468,
+      "step": 3183
+    },
+    {
+      "epoch": 0.4345571175105773,
+      "grad_norm": 0.19795763492584229,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 3184
+    },
+    {
+      "epoch": 0.4346935990173332,
+      "grad_norm": 0.20356614887714386,
+      "learning_rate": 8e-05,
+      "loss": 1.4669,
+      "step": 3185
+    },
+    {
+      "epoch": 0.434830080524089,
+      "grad_norm": 0.20835043489933014,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4349665620308448,
+      "grad_norm": 0.2024518996477127,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 3187
+    },
+    {
+      "epoch": 0.43510304353760065,
+      "grad_norm": 0.204525887966156,
+      "learning_rate": 8e-05,
+      "loss": 1.4834,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4352395250443565,
+      "grad_norm": 0.20044328272342682,
+      "learning_rate": 8e-05,
+      "loss": 1.4579,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4353760065511123,
+      "grad_norm": 0.19587454199790955,
+      "learning_rate": 8e-05,
+      "loss": 1.4651,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4355124880578682,
+      "grad_norm": 0.20442837476730347,
+      "learning_rate": 8e-05,
+      "loss": 1.5035,
+      "step": 3191
+    },
+    {
+      "epoch": 0.435648969564624,
+      "grad_norm": 0.1999359279870987,
+      "learning_rate": 8e-05,
+      "loss": 1.4941,
+      "step": 3192
+    },
+    {
+      "epoch": 0.43578545107137984,
+      "grad_norm": 0.20011280477046967,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 3193
+    },
+    {
+      "epoch": 0.43592193257813566,
+      "grad_norm": 0.20265041291713715,
+      "learning_rate": 8e-05,
+      "loss": 1.4994,
+      "step": 3194
+    },
+    {
+      "epoch": 0.4360584140848915,
+      "grad_norm": 0.20954529941082,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 3195
+    },
+    {
+      "epoch": 0.4361948955916473,
+      "grad_norm": 0.21581964194774628,
+      "learning_rate": 8e-05,
+      "loss": 1.5333,
+      "step": 3196
+    },
+    {
+      "epoch": 0.43633137709840314,
+      "grad_norm": 0.21604877710342407,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 3197
+    },
+    {
+      "epoch": 0.436467858605159,
+      "grad_norm": 0.20408225059509277,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 3198
+    },
+    {
+      "epoch": 0.43660434011191485,
+      "grad_norm": 0.21442236006259918,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 3199
+    },
+    {
+      "epoch": 0.43674082161867067,
+      "grad_norm": 0.19689717888832092,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4368773031254265,
+      "grad_norm": 0.20440372824668884,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 3201
+    },
+    {
+      "epoch": 0.4370137846321823,
+      "grad_norm": 0.20694170892238617,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 3202
+    },
+    {
+      "epoch": 0.43715026613893815,
+      "grad_norm": 0.2014540135860443,
+      "learning_rate": 8e-05,
+      "loss": 1.5509,
+      "step": 3203
+    },
+    {
+      "epoch": 0.43728674764569403,
+      "grad_norm": 0.19631561636924744,
+      "learning_rate": 8e-05,
+      "loss": 1.4357,
+      "step": 3204
+    },
+    {
+      "epoch": 0.43742322915244986,
+      "grad_norm": 0.19948531687259674,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 3205
+    },
+    {
+      "epoch": 0.4375597106592057,
+      "grad_norm": 0.20264436304569244,
+      "learning_rate": 8e-05,
+      "loss": 1.5159,
+      "step": 3206
+    },
+    {
+      "epoch": 0.4376961921659615,
+      "grad_norm": 0.21580493450164795,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 3207
+    },
+    {
+      "epoch": 0.43783267367271733,
+      "grad_norm": 0.20941375195980072,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 3208
+    },
+    {
+      "epoch": 0.43796915517947316,
+      "grad_norm": 0.2025696188211441,
+      "learning_rate": 8e-05,
+      "loss": 1.4622,
+      "step": 3209
+    },
+    {
+      "epoch": 0.43810563668622904,
+      "grad_norm": 0.20156575739383698,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 3210
+    },
+    {
+      "epoch": 0.43824211819298486,
+      "grad_norm": 0.20984598994255066,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 3211
+    },
+    {
+      "epoch": 0.4383785996997407,
+      "grad_norm": 0.21122805774211884,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4385150812064965,
+      "grad_norm": 0.1980714350938797,
+      "learning_rate": 8e-05,
+      "loss": 1.4614,
+      "step": 3213
+    },
+    {
+      "epoch": 0.43865156271325234,
+      "grad_norm": 0.20132118463516235,
+      "learning_rate": 8e-05,
+      "loss": 1.5208,
+      "step": 3214
+    },
+    {
+      "epoch": 0.43878804422000817,
+      "grad_norm": 0.1983589380979538,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 3215
+    },
+    {
+      "epoch": 0.43892452572676405,
+      "grad_norm": 0.19686248898506165,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 3216
+    },
+    {
+      "epoch": 0.4390610072335199,
+      "grad_norm": 0.2037227302789688,
+      "learning_rate": 8e-05,
+      "loss": 1.5326,
+      "step": 3217
+    },
+    {
+      "epoch": 0.4391974887402757,
+      "grad_norm": 0.20650477707386017,
+      "learning_rate": 8e-05,
+      "loss": 1.5419,
+      "step": 3218
+    },
+    {
+      "epoch": 0.4393339702470315,
+      "grad_norm": 0.2068139910697937,
+      "learning_rate": 8e-05,
+      "loss": 1.5189,
+      "step": 3219
+    },
+    {
+      "epoch": 0.43947045175378735,
+      "grad_norm": 0.20991960167884827,
+      "learning_rate": 8e-05,
+      "loss": 1.4901,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4396069332605432,
+      "grad_norm": 0.20221896469593048,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 3221
+    },
+    {
+      "epoch": 0.43974341476729906,
+      "grad_norm": 0.20129981637001038,
+      "learning_rate": 8e-05,
+      "loss": 1.4262,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4398798962740549,
+      "grad_norm": 0.2061825692653656,
+      "learning_rate": 8e-05,
+      "loss": 1.514,
+      "step": 3223
+    },
+    {
+      "epoch": 0.4400163777808107,
+      "grad_norm": 0.20163790881633759,
+      "learning_rate": 8e-05,
+      "loss": 1.5286,
+      "step": 3224
+    },
+    {
+      "epoch": 0.44015285928756653,
+      "grad_norm": 0.20273712277412415,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 3225
+    },
+    {
+      "epoch": 0.44028934079432236,
+      "grad_norm": 0.20562024414539337,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4404258223010782,
+      "grad_norm": 0.19715231657028198,
+      "learning_rate": 8e-05,
+      "loss": 1.5121,
+      "step": 3227
+    },
+    {
+      "epoch": 0.44056230380783407,
+      "grad_norm": 0.20362688601016998,
+      "learning_rate": 8e-05,
+      "loss": 1.4977,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4406987853145899,
+      "grad_norm": 0.20189505815505981,
+      "learning_rate": 8e-05,
+      "loss": 1.4321,
+      "step": 3229
+    },
+    {
+      "epoch": 0.4408352668213457,
+      "grad_norm": 0.2063058763742447,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 3230
+    },
+    {
+      "epoch": 0.44097174832810154,
+      "grad_norm": 0.20270198583602905,
+      "learning_rate": 8e-05,
+      "loss": 1.4638,
+      "step": 3231
+    },
+    {
+      "epoch": 0.44110822983485737,
+      "grad_norm": 0.20998196303844452,
+      "learning_rate": 8e-05,
+      "loss": 1.5371,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4412447113416132,
+      "grad_norm": 0.20822195708751678,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 3233
+    },
+    {
+      "epoch": 0.441381192848369,
+      "grad_norm": 0.20829296112060547,
+      "learning_rate": 8e-05,
+      "loss": 1.5192,
+      "step": 3234
+    },
+    {
+      "epoch": 0.4415176743551249,
+      "grad_norm": 0.20105651021003723,
+      "learning_rate": 8e-05,
+      "loss": 1.5134,
+      "step": 3235
+    },
+    {
+      "epoch": 0.4416541558618807,
+      "grad_norm": 0.1972811073064804,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 3236
+    },
+    {
+      "epoch": 0.44179063736863655,
+      "grad_norm": 0.196736142039299,
+      "learning_rate": 8e-05,
+      "loss": 1.4542,
+      "step": 3237
+    },
+    {
+      "epoch": 0.4419271188753924,
+      "grad_norm": 0.19351115822792053,
+      "learning_rate": 8e-05,
+      "loss": 1.487,
+      "step": 3238
+    },
+    {
+      "epoch": 0.4420636003821482,
+      "grad_norm": 0.20947715640068054,
+      "learning_rate": 8e-05,
+      "loss": 1.5274,
+      "step": 3239
+    },
+    {
+      "epoch": 0.44220008188890403,
+      "grad_norm": 0.20681138336658478,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4423365633956599,
+      "grad_norm": 0.2098737210035324,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 3241
+    },
+    {
+      "epoch": 0.44247304490241574,
+      "grad_norm": 0.21125026047229767,
+      "learning_rate": 8e-05,
+      "loss": 1.5756,
+      "step": 3242
+    },
+    {
+      "epoch": 0.44260952640917156,
+      "grad_norm": 0.19590415060520172,
+      "learning_rate": 8e-05,
+      "loss": 1.5496,
+      "step": 3243
+    },
+    {
+      "epoch": 0.4427460079159274,
+      "grad_norm": 0.21082034707069397,
+      "learning_rate": 8e-05,
+      "loss": 1.5465,
+      "step": 3244
+    },
+    {
+      "epoch": 0.4428824894226832,
+      "grad_norm": 0.20047715306282043,
+      "learning_rate": 8e-05,
+      "loss": 1.4536,
+      "step": 3245
+    },
+    {
+      "epoch": 0.44301897092943904,
+      "grad_norm": 0.19345968961715698,
+      "learning_rate": 8e-05,
+      "loss": 1.4444,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4431554524361949,
+      "grad_norm": 0.1998324692249298,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 3247
+    },
+    {
+      "epoch": 0.44329193394295074,
+      "grad_norm": 0.20281356573104858,
+      "learning_rate": 8e-05,
+      "loss": 1.5028,
+      "step": 3248
+    },
+    {
+      "epoch": 0.44342841544970657,
+      "grad_norm": 0.2019728124141693,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 3249
+    },
+    {
+      "epoch": 0.4435648969564624,
+      "grad_norm": 0.20388513803482056,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4437013784632182,
+      "grad_norm": 0.21065255999565125,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 3251
+    },
+    {
+      "epoch": 0.44383785996997405,
+      "grad_norm": 0.20478476583957672,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 3252
+    },
+    {
+      "epoch": 0.44397434147672993,
+      "grad_norm": 0.20145496726036072,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 3253
+    },
+    {
+      "epoch": 0.44411082298348575,
+      "grad_norm": 0.20837576687335968,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 3254
+    },
+    {
+      "epoch": 0.4442473044902416,
+      "grad_norm": 0.20676518976688385,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 3255
+    },
+    {
+      "epoch": 0.4443837859969974,
+      "grad_norm": 0.20005717873573303,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 3256
+    },
+    {
+      "epoch": 0.44452026750375323,
+      "grad_norm": 0.20589189231395721,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 3257
+    },
+    {
+      "epoch": 0.44465674901050906,
+      "grad_norm": 0.19458532333374023,
+      "learning_rate": 8e-05,
+      "loss": 1.4664,
+      "step": 3258
+    },
+    {
+      "epoch": 0.44479323051726494,
+      "grad_norm": 0.2031342089176178,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 3259
+    },
+    {
+      "epoch": 0.44492971202402076,
+      "grad_norm": 0.20186376571655273,
+      "learning_rate": 8e-05,
+      "loss": 1.5496,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4450661935307766,
+      "grad_norm": 0.21457123756408691,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4452026750375324,
+      "grad_norm": 0.20058996975421906,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 3262
+    },
+    {
+      "epoch": 0.44533915654428824,
+      "grad_norm": 0.19569219648838043,
+      "learning_rate": 8e-05,
+      "loss": 1.4615,
+      "step": 3263
+    },
+    {
+      "epoch": 0.44547563805104406,
+      "grad_norm": 0.20526716113090515,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 3264
+    },
+    {
+      "epoch": 0.4456121195577999,
+      "grad_norm": 0.2020854502916336,
+      "learning_rate": 8e-05,
+      "loss": 1.4831,
+      "step": 3265
+    },
+    {
+      "epoch": 0.44574860106455577,
+      "grad_norm": 0.20917055010795593,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 3266
+    },
+    {
+      "epoch": 0.4458850825713116,
+      "grad_norm": 0.19959941506385803,
+      "learning_rate": 8e-05,
+      "loss": 1.4752,
+      "step": 3267
+    },
+    {
+      "epoch": 0.4460215640780674,
+      "grad_norm": 0.1964586079120636,
+      "learning_rate": 8e-05,
+      "loss": 1.4329,
+      "step": 3268
+    },
+    {
+      "epoch": 0.44615804558482325,
+      "grad_norm": 0.20622660219669342,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 3269
+    },
+    {
+      "epoch": 0.4462945270915791,
+      "grad_norm": 0.20728552341461182,
+      "learning_rate": 8e-05,
+      "loss": 1.4974,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4464310085983349,
+      "grad_norm": 0.20910215377807617,
+      "learning_rate": 8e-05,
+      "loss": 1.4808,
+      "step": 3271
+    },
+    {
+      "epoch": 0.4465674901050908,
+      "grad_norm": 0.20853453874588013,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 3272
+    },
+    {
+      "epoch": 0.4467039716118466,
+      "grad_norm": 0.20444753766059875,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 3273
+    },
+    {
+      "epoch": 0.44684045311860243,
+      "grad_norm": 0.2087266594171524,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 3274
+    },
+    {
+      "epoch": 0.44697693462535826,
+      "grad_norm": 0.2099228948354721,
+      "learning_rate": 8e-05,
+      "loss": 1.5106,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4471134161321141,
+      "grad_norm": 0.20607341825962067,
+      "learning_rate": 8e-05,
+      "loss": 1.5667,
+      "step": 3276
+    },
+    {
+      "epoch": 0.4472498976388699,
+      "grad_norm": 0.2116592824459076,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 3277
+    },
+    {
+      "epoch": 0.4473863791456258,
+      "grad_norm": 0.20507965981960297,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 3278
+    },
+    {
+      "epoch": 0.4475228606523816,
+      "grad_norm": 0.19957639276981354,
+      "learning_rate": 8e-05,
+      "loss": 1.4904,
+      "step": 3279
+    },
+    {
+      "epoch": 0.44765934215913744,
+      "grad_norm": 0.20068909227848053,
+      "learning_rate": 8e-05,
+      "loss": 1.4915,
+      "step": 3280
+    },
+    {
+      "epoch": 0.44779582366589327,
+      "grad_norm": 0.19837771356105804,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4479323051726491,
+      "grad_norm": 0.20921039581298828,
+      "learning_rate": 8e-05,
+      "loss": 1.5068,
+      "step": 3282
+    },
+    {
+      "epoch": 0.4480687866794049,
+      "grad_norm": 0.19557024538516998,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 3283
+    },
+    {
+      "epoch": 0.4482052681861608,
+      "grad_norm": 0.2009022831916809,
+      "learning_rate": 8e-05,
+      "loss": 1.5053,
+      "step": 3284
+    },
+    {
+      "epoch": 0.4483417496929166,
+      "grad_norm": 0.19988054037094116,
+      "learning_rate": 8e-05,
+      "loss": 1.5538,
+      "step": 3285
+    },
+    {
+      "epoch": 0.44847823119967245,
+      "grad_norm": 0.20481422543525696,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 3286
+    },
+    {
+      "epoch": 0.4486147127064283,
+      "grad_norm": 0.20532605051994324,
+      "learning_rate": 8e-05,
+      "loss": 1.5212,
+      "step": 3287
+    },
+    {
+      "epoch": 0.4487511942131841,
+      "grad_norm": 0.19546401500701904,
+      "learning_rate": 8e-05,
+      "loss": 1.4819,
+      "step": 3288
+    },
+    {
+      "epoch": 0.4488876757199399,
+      "grad_norm": 0.2069421410560608,
+      "learning_rate": 8e-05,
+      "loss": 1.4446,
+      "step": 3289
+    },
+    {
+      "epoch": 0.4490241572266958,
+      "grad_norm": 0.21116097271442413,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 3290
+    },
+    {
+      "epoch": 0.44916063873345163,
+      "grad_norm": 0.20897267758846283,
+      "learning_rate": 8e-05,
+      "loss": 1.5194,
+      "step": 3291
+    },
+    {
+      "epoch": 0.44929712024020746,
+      "grad_norm": 0.204427108168602,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 3292
+    },
+    {
+      "epoch": 0.4494336017469633,
+      "grad_norm": 0.20322507619857788,
+      "learning_rate": 8e-05,
+      "loss": 1.5059,
+      "step": 3293
+    },
+    {
+      "epoch": 0.4495700832537191,
+      "grad_norm": 0.20074470341205597,
+      "learning_rate": 8e-05,
+      "loss": 1.4903,
+      "step": 3294
+    },
+    {
+      "epoch": 0.44970656476047494,
+      "grad_norm": 0.20163731276988983,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 3295
+    },
+    {
+      "epoch": 0.4498430462672308,
+      "grad_norm": 0.20294135808944702,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 3296
+    },
+    {
+      "epoch": 0.44997952777398664,
+      "grad_norm": 0.19917891919612885,
+      "learning_rate": 8e-05,
+      "loss": 1.5193,
+      "step": 3297
+    },
+    {
+      "epoch": 0.45011600928074247,
+      "grad_norm": 0.20442698895931244,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 3298
+    },
+    {
+      "epoch": 0.4502524907874983,
+      "grad_norm": 0.21351264417171478,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 3299
+    },
+    {
+      "epoch": 0.4503889722942541,
+      "grad_norm": 0.20064903795719147,
+      "learning_rate": 8e-05,
+      "loss": 1.4876,
+      "step": 3300
+    },
+    {
+      "epoch": 0.45052545380100995,
+      "grad_norm": 0.20760200917720795,
+      "learning_rate": 8e-05,
+      "loss": 1.5109,
+      "step": 3301
+    },
+    {
+      "epoch": 0.45066193530776577,
+      "grad_norm": 0.21604247391223907,
+      "learning_rate": 8e-05,
+      "loss": 1.5034,
+      "step": 3302
+    },
+    {
+      "epoch": 0.45079841681452165,
+      "grad_norm": 0.20299625396728516,
+      "learning_rate": 8e-05,
+      "loss": 1.5225,
+      "step": 3303
+    },
+    {
+      "epoch": 0.4509348983212775,
+      "grad_norm": 0.20287269353866577,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 3304
+    },
+    {
+      "epoch": 0.4510713798280333,
+      "grad_norm": 0.20292556285858154,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 3305
+    },
+    {
+      "epoch": 0.45120786133478913,
+      "grad_norm": 0.20476415753364563,
+      "learning_rate": 8e-05,
+      "loss": 1.4859,
+      "step": 3306
+    },
+    {
+      "epoch": 0.45134434284154495,
+      "grad_norm": 0.20652766525745392,
+      "learning_rate": 8e-05,
+      "loss": 1.5395,
+      "step": 3307
+    },
+    {
+      "epoch": 0.4514808243483008,
+      "grad_norm": 0.1977231353521347,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 3308
+    },
+    {
+      "epoch": 0.45161730585505666,
+      "grad_norm": 0.20224037766456604,
+      "learning_rate": 8e-05,
+      "loss": 1.5549,
+      "step": 3309
+    },
+    {
+      "epoch": 0.4517537873618125,
+      "grad_norm": 0.20821750164031982,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4518902688685683,
+      "grad_norm": 0.20505215227603912,
+      "learning_rate": 8e-05,
+      "loss": 1.5094,
+      "step": 3311
+    },
+    {
+      "epoch": 0.45202675037532414,
+      "grad_norm": 0.20222346484661102,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 3312
+    },
+    {
+      "epoch": 0.45216323188207996,
+      "grad_norm": 0.20393307507038116,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 3313
+    },
+    {
+      "epoch": 0.4522997133888358,
+      "grad_norm": 0.20658648014068604,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 3314
+    },
+    {
+      "epoch": 0.45243619489559167,
+      "grad_norm": 0.2030039280653,
+      "learning_rate": 8e-05,
+      "loss": 1.5381,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4525726764023475,
+      "grad_norm": 0.19749687612056732,
+      "learning_rate": 8e-05,
+      "loss": 1.5239,
+      "step": 3316
+    },
+    {
+      "epoch": 0.4527091579091033,
+      "grad_norm": 0.20338688790798187,
+      "learning_rate": 8e-05,
+      "loss": 1.5507,
+      "step": 3317
+    },
+    {
+      "epoch": 0.45284563941585915,
+      "grad_norm": 0.20814357697963715,
+      "learning_rate": 8e-05,
+      "loss": 1.487,
+      "step": 3318
+    },
+    {
+      "epoch": 0.452982120922615,
+      "grad_norm": 0.20874108374118805,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 3319
+    },
+    {
+      "epoch": 0.4531186024293708,
+      "grad_norm": 0.1981109231710434,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4532550839361267,
+      "grad_norm": 0.19380518794059753,
+      "learning_rate": 8e-05,
+      "loss": 1.3919,
+      "step": 3321
+    },
+    {
+      "epoch": 0.4533915654428825,
+      "grad_norm": 0.19989387691020966,
+      "learning_rate": 8e-05,
+      "loss": 1.4703,
+      "step": 3322
+    },
+    {
+      "epoch": 0.45352804694963833,
+      "grad_norm": 0.2048611342906952,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 3323
+    },
+    {
+      "epoch": 0.45366452845639416,
+      "grad_norm": 0.19949188828468323,
+      "learning_rate": 8e-05,
+      "loss": 1.5242,
+      "step": 3324
+    },
+    {
+      "epoch": 0.45380100996315,
+      "grad_norm": 0.20602813363075256,
+      "learning_rate": 8e-05,
+      "loss": 1.4855,
+      "step": 3325
+    },
+    {
+      "epoch": 0.4539374914699058,
+      "grad_norm": 0.2010207623243332,
+      "learning_rate": 8e-05,
+      "loss": 1.487,
+      "step": 3326
+    },
+    {
+      "epoch": 0.4540739729766617,
+      "grad_norm": 0.19941088557243347,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4542104544834175,
+      "grad_norm": 0.21096745133399963,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 3328
+    },
+    {
+      "epoch": 0.45434693599017334,
+      "grad_norm": 0.20490902662277222,
+      "learning_rate": 8e-05,
+      "loss": 1.4486,
+      "step": 3329
+    },
+    {
+      "epoch": 0.45448341749692917,
+      "grad_norm": 0.20350489020347595,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 3330
+    },
+    {
+      "epoch": 0.454619899003685,
+      "grad_norm": 0.19724179804325104,
+      "learning_rate": 8e-05,
+      "loss": 1.4632,
+      "step": 3331
+    },
+    {
+      "epoch": 0.4547563805104408,
+      "grad_norm": 0.20504145324230194,
+      "learning_rate": 8e-05,
+      "loss": 1.4499,
+      "step": 3332
+    },
+    {
+      "epoch": 0.4548928620171967,
+      "grad_norm": 0.20698674023151398,
+      "learning_rate": 8e-05,
+      "loss": 1.5428,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4550293435239525,
+      "grad_norm": 0.20949898660182953,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 3334
+    },
+    {
+      "epoch": 0.45516582503070835,
+      "grad_norm": 0.2054181545972824,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4553023065374642,
+      "grad_norm": 0.20110256969928741,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 3336
+    },
+    {
+      "epoch": 0.45543878804422,
+      "grad_norm": 0.21335738897323608,
+      "learning_rate": 8e-05,
+      "loss": 1.4973,
+      "step": 3337
+    },
+    {
+      "epoch": 0.4555752695509758,
+      "grad_norm": 0.22037526965141296,
+      "learning_rate": 8e-05,
+      "loss": 1.5179,
+      "step": 3338
+    },
+    {
+      "epoch": 0.45571175105773165,
+      "grad_norm": 0.2023409754037857,
+      "learning_rate": 8e-05,
+      "loss": 1.4262,
+      "step": 3339
+    },
+    {
+      "epoch": 0.45584823256448753,
+      "grad_norm": 0.2147899866104126,
+      "learning_rate": 8e-05,
+      "loss": 1.5505,
+      "step": 3340
+    },
+    {
+      "epoch": 0.45598471407124336,
+      "grad_norm": 0.20637552440166473,
+      "learning_rate": 8e-05,
+      "loss": 1.4571,
+      "step": 3341
+    },
+    {
+      "epoch": 0.4561211955779992,
+      "grad_norm": 0.20200565457344055,
+      "learning_rate": 8e-05,
+      "loss": 1.5192,
+      "step": 3342
+    },
+    {
+      "epoch": 0.456257677084755,
+      "grad_norm": 0.2167416363954544,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 3343
+    },
+    {
+      "epoch": 0.45639415859151083,
+      "grad_norm": 0.2135801911354065,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 3344
+    },
+    {
+      "epoch": 0.45653064009826666,
+      "grad_norm": 0.19715051352977753,
+      "learning_rate": 8e-05,
+      "loss": 1.4314,
+      "step": 3345
+    },
+    {
+      "epoch": 0.45666712160502254,
+      "grad_norm": 0.21061591804027557,
+      "learning_rate": 8e-05,
+      "loss": 1.4466,
+      "step": 3346
+    },
+    {
+      "epoch": 0.45680360311177837,
+      "grad_norm": 0.22038885951042175,
+      "learning_rate": 8e-05,
+      "loss": 1.4793,
+      "step": 3347
+    },
+    {
+      "epoch": 0.4569400846185342,
+      "grad_norm": 0.21247152984142303,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 3348
+    },
+    {
+      "epoch": 0.45707656612529,
+      "grad_norm": 0.20761588215827942,
+      "learning_rate": 8e-05,
+      "loss": 1.4878,
+      "step": 3349
+    },
+    {
+      "epoch": 0.45721304763204584,
+      "grad_norm": 0.20544196665287018,
+      "learning_rate": 8e-05,
+      "loss": 1.4723,
+      "step": 3350
+    },
+    {
+      "epoch": 0.45734952913880167,
+      "grad_norm": 0.1954910159111023,
+      "learning_rate": 8e-05,
+      "loss": 1.5197,
+      "step": 3351
+    },
+    {
+      "epoch": 0.45748601064555755,
+      "grad_norm": 0.21049875020980835,
+      "learning_rate": 8e-05,
+      "loss": 1.4816,
+      "step": 3352
+    },
+    {
+      "epoch": 0.4576224921523134,
+      "grad_norm": 0.20720385015010834,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 3353
+    },
+    {
+      "epoch": 0.4577589736590692,
+      "grad_norm": 0.19467219710350037,
+      "learning_rate": 8e-05,
+      "loss": 1.4886,
+      "step": 3354
+    },
+    {
+      "epoch": 0.457895455165825,
+      "grad_norm": 0.21409712731838226,
+      "learning_rate": 8e-05,
+      "loss": 1.5023,
+      "step": 3355
+    },
+    {
+      "epoch": 0.45803193667258085,
+      "grad_norm": 0.20801040530204773,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 3356
+    },
+    {
+      "epoch": 0.4581684181793367,
+      "grad_norm": 0.2032715082168579,
+      "learning_rate": 8e-05,
+      "loss": 1.5405,
+      "step": 3357
+    },
+    {
+      "epoch": 0.45830489968609256,
+      "grad_norm": 0.2053171992301941,
+      "learning_rate": 8e-05,
+      "loss": 1.49,
+      "step": 3358
+    },
+    {
+      "epoch": 0.4584413811928484,
+      "grad_norm": 0.2031218409538269,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 3359
+    },
+    {
+      "epoch": 0.4585778626996042,
+      "grad_norm": 0.20945163071155548,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 3360
+    },
+    {
+      "epoch": 0.45871434420636004,
+      "grad_norm": 0.20190584659576416,
+      "learning_rate": 8e-05,
+      "loss": 1.5562,
+      "step": 3361
+    },
+    {
+      "epoch": 0.45885082571311586,
+      "grad_norm": 0.21301424503326416,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 3362
+    },
+    {
+      "epoch": 0.4589873072198717,
+      "grad_norm": 0.2128937691450119,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 3363
+    },
+    {
+      "epoch": 0.45912378872662757,
+      "grad_norm": 0.210130974650383,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4592602702333834,
+      "grad_norm": 0.2120717614889145,
+      "learning_rate": 8e-05,
+      "loss": 1.5912,
+      "step": 3365
+    },
+    {
+      "epoch": 0.4593967517401392,
+      "grad_norm": 0.19900810718536377,
+      "learning_rate": 8e-05,
+      "loss": 1.5107,
+      "step": 3366
+    },
+    {
+      "epoch": 0.45953323324689505,
+      "grad_norm": 0.20568035542964935,
+      "learning_rate": 8e-05,
+      "loss": 1.426,
+      "step": 3367
+    },
+    {
+      "epoch": 0.45966971475365087,
+      "grad_norm": 0.19921958446502686,
+      "learning_rate": 8e-05,
+      "loss": 1.4982,
+      "step": 3368
+    },
+    {
+      "epoch": 0.4598061962604067,
+      "grad_norm": 0.19632309675216675,
+      "learning_rate": 8e-05,
+      "loss": 1.486,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4599426777671625,
+      "grad_norm": 0.202032670378685,
+      "learning_rate": 8e-05,
+      "loss": 1.5204,
+      "step": 3370
+    },
+    {
+      "epoch": 0.4600791592739184,
+      "grad_norm": 0.20294281840324402,
+      "learning_rate": 8e-05,
+      "loss": 1.448,
+      "step": 3371
+    },
+    {
+      "epoch": 0.46021564078067423,
+      "grad_norm": 0.204691082239151,
+      "learning_rate": 8e-05,
+      "loss": 1.5305,
+      "step": 3372
+    },
+    {
+      "epoch": 0.46035212228743005,
+      "grad_norm": 0.223212331533432,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 3373
+    },
+    {
+      "epoch": 0.4604886037941859,
+      "grad_norm": 0.202362522482872,
+      "learning_rate": 8e-05,
+      "loss": 1.428,
+      "step": 3374
+    },
+    {
+      "epoch": 0.4606250853009417,
+      "grad_norm": 0.20770061016082764,
+      "learning_rate": 8e-05,
+      "loss": 1.5372,
+      "step": 3375
+    },
+    {
+      "epoch": 0.46076156680769753,
+      "grad_norm": 0.20081312954425812,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 3376
+    },
+    {
+      "epoch": 0.4608980483144534,
+      "grad_norm": 0.21202722191810608,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 3377
+    },
+    {
+      "epoch": 0.46103452982120924,
+      "grad_norm": 0.20443178713321686,
+      "learning_rate": 8e-05,
+      "loss": 1.4857,
+      "step": 3378
+    },
+    {
+      "epoch": 0.46117101132796506,
+      "grad_norm": 0.19714051485061646,
+      "learning_rate": 8e-05,
+      "loss": 1.4355,
+      "step": 3379
+    },
+    {
+      "epoch": 0.4613074928347209,
+      "grad_norm": 0.20318607985973358,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 3380
+    },
+    {
+      "epoch": 0.4614439743414767,
+      "grad_norm": 0.21267586946487427,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 3381
+    },
+    {
+      "epoch": 0.46158045584823254,
+      "grad_norm": 0.21322016417980194,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 3382
+    },
+    {
+      "epoch": 0.4617169373549884,
+      "grad_norm": 0.20762303471565247,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 3383
+    },
+    {
+      "epoch": 0.46185341886174425,
+      "grad_norm": 0.19948901236057281,
+      "learning_rate": 8e-05,
+      "loss": 1.4694,
+      "step": 3384
+    },
+    {
+      "epoch": 0.4619899003685001,
+      "grad_norm": 0.20969977974891663,
+      "learning_rate": 8e-05,
+      "loss": 1.5382,
+      "step": 3385
+    },
+    {
+      "epoch": 0.4621263818752559,
+      "grad_norm": 0.20358628034591675,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 3386
+    },
+    {
+      "epoch": 0.4622628633820117,
+      "grad_norm": 0.20342092216014862,
+      "learning_rate": 8e-05,
+      "loss": 1.4972,
+      "step": 3387
+    },
+    {
+      "epoch": 0.46239934488876755,
+      "grad_norm": 0.20850619673728943,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 3388
+    },
+    {
+      "epoch": 0.46253582639552343,
+      "grad_norm": 0.19653436541557312,
+      "learning_rate": 8e-05,
+      "loss": 1.4729,
+      "step": 3389
+    },
+    {
+      "epoch": 0.46267230790227926,
+      "grad_norm": 0.2011420726776123,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4628087894090351,
+      "grad_norm": 0.21052059531211853,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 3391
+    },
+    {
+      "epoch": 0.4629452709157909,
+      "grad_norm": 0.21375620365142822,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 3392
+    },
+    {
+      "epoch": 0.46308175242254673,
+      "grad_norm": 0.20165874063968658,
+      "learning_rate": 8e-05,
+      "loss": 1.4789,
+      "step": 3393
+    },
+    {
+      "epoch": 0.46321823392930256,
+      "grad_norm": 0.21167610585689545,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 3394
+    },
+    {
+      "epoch": 0.46335471543605844,
+      "grad_norm": 0.20535144209861755,
+      "learning_rate": 8e-05,
+      "loss": 1.5265,
+      "step": 3395
+    },
+    {
+      "epoch": 0.46349119694281427,
+      "grad_norm": 0.20090624690055847,
+      "learning_rate": 8e-05,
+      "loss": 1.4745,
+      "step": 3396
+    },
+    {
+      "epoch": 0.4636276784495701,
+      "grad_norm": 0.20968738198280334,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 3397
+    },
+    {
+      "epoch": 0.4637641599563259,
+      "grad_norm": 0.20477834343910217,
+      "learning_rate": 8e-05,
+      "loss": 1.4472,
+      "step": 3398
+    },
+    {
+      "epoch": 0.46390064146308174,
+      "grad_norm": 0.1990976184606552,
+      "learning_rate": 8e-05,
+      "loss": 1.433,
+      "step": 3399
+    },
+    {
+      "epoch": 0.46403712296983757,
+      "grad_norm": 0.20355604588985443,
+      "learning_rate": 8e-05,
+      "loss": 1.5164,
+      "step": 3400
+    },
+    {
+      "epoch": 0.46417360447659345,
+      "grad_norm": 0.2029486745595932,
+      "learning_rate": 8e-05,
+      "loss": 1.5304,
+      "step": 3401
+    },
+    {
+      "epoch": 0.4643100859833493,
+      "grad_norm": 0.20830398797988892,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4644465674901051,
+      "grad_norm": 0.20323744416236877,
+      "learning_rate": 8e-05,
+      "loss": 1.4365,
+      "step": 3403
+    },
+    {
+      "epoch": 0.4645830489968609,
+      "grad_norm": 0.20358824729919434,
+      "learning_rate": 8e-05,
+      "loss": 1.5638,
+      "step": 3404
+    },
+    {
+      "epoch": 0.46471953050361675,
+      "grad_norm": 0.2191862314939499,
+      "learning_rate": 8e-05,
+      "loss": 1.395,
+      "step": 3405
+    },
+    {
+      "epoch": 0.4648560120103726,
+      "grad_norm": 0.2157006561756134,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 3406
+    },
+    {
+      "epoch": 0.4649924935171284,
+      "grad_norm": 0.20152826607227325,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 3407
+    },
+    {
+      "epoch": 0.4651289750238843,
+      "grad_norm": 0.23860777914524078,
+      "learning_rate": 8e-05,
+      "loss": 1.564,
+      "step": 3408
+    },
+    {
+      "epoch": 0.4652654565306401,
+      "grad_norm": 0.21131755411624908,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3409
+    },
+    {
+      "epoch": 0.46540193803739593,
+      "grad_norm": 0.21757672727108002,
+      "learning_rate": 8e-05,
+      "loss": 1.4683,
+      "step": 3410
+    },
+    {
+      "epoch": 0.46553841954415176,
+      "grad_norm": 0.22972005605697632,
+      "learning_rate": 8e-05,
+      "loss": 1.471,
+      "step": 3411
+    },
+    {
+      "epoch": 0.4656749010509076,
+      "grad_norm": 0.21489542722702026,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 3412
+    },
+    {
+      "epoch": 0.4658113825576634,
+      "grad_norm": 0.21335068345069885,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 3413
+    },
+    {
+      "epoch": 0.4659478640644193,
+      "grad_norm": 0.23898380994796753,
+      "learning_rate": 8e-05,
+      "loss": 1.4957,
+      "step": 3414
+    },
+    {
+      "epoch": 0.4660843455711751,
+      "grad_norm": 0.2167702615261078,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 3415
+    },
+    {
+      "epoch": 0.46622082707793094,
+      "grad_norm": 0.22000667452812195,
+      "learning_rate": 8e-05,
+      "loss": 1.5579,
+      "step": 3416
+    },
+    {
+      "epoch": 0.46635730858468677,
+      "grad_norm": 0.22867894172668457,
+      "learning_rate": 8e-05,
+      "loss": 1.4241,
+      "step": 3417
+    },
+    {
+      "epoch": 0.4664937900914426,
+      "grad_norm": 0.20478494465351105,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 3418
+    },
+    {
+      "epoch": 0.4666302715981984,
+      "grad_norm": 0.20832200348377228,
+      "learning_rate": 8e-05,
+      "loss": 1.4789,
+      "step": 3419
+    },
+    {
+      "epoch": 0.4667667531049543,
+      "grad_norm": 0.23364369571208954,
+      "learning_rate": 8e-05,
+      "loss": 1.479,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4669032346117101,
+      "grad_norm": 0.20809462666511536,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 3421
+    },
+    {
+      "epoch": 0.46703971611846595,
+      "grad_norm": 0.2016485035419464,
+      "learning_rate": 8e-05,
+      "loss": 1.5433,
+      "step": 3422
+    },
+    {
+      "epoch": 0.4671761976252218,
+      "grad_norm": 0.2142065316438675,
+      "learning_rate": 8e-05,
+      "loss": 1.4763,
+      "step": 3423
+    },
+    {
+      "epoch": 0.4673126791319776,
+      "grad_norm": 0.21977971494197845,
+      "learning_rate": 8e-05,
+      "loss": 1.4527,
+      "step": 3424
+    },
+    {
+      "epoch": 0.46744916063873343,
+      "grad_norm": 0.21415258944034576,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4675856421454893,
+      "grad_norm": 0.23099614679813385,
+      "learning_rate": 8e-05,
+      "loss": 1.4281,
+      "step": 3426
+    },
+    {
+      "epoch": 0.46772212365224514,
+      "grad_norm": 0.2302975356578827,
+      "learning_rate": 8e-05,
+      "loss": 1.5825,
+      "step": 3427
+    },
+    {
+      "epoch": 0.46785860515900096,
+      "grad_norm": 0.2032427042722702,
+      "learning_rate": 8e-05,
+      "loss": 1.423,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4679950866657568,
+      "grad_norm": 0.21455858647823334,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4681315681725126,
+      "grad_norm": 0.22139950096607208,
+      "learning_rate": 8e-05,
+      "loss": 1.4743,
+      "step": 3430
+    },
+    {
+      "epoch": 0.46826804967926844,
+      "grad_norm": 0.20664794743061066,
+      "learning_rate": 8e-05,
+      "loss": 1.5245,
+      "step": 3431
+    },
+    {
+      "epoch": 0.4684045311860243,
+      "grad_norm": 0.21289144456386566,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 3432
+    },
+    {
+      "epoch": 0.46854101269278015,
+      "grad_norm": 0.20248781144618988,
+      "learning_rate": 8e-05,
+      "loss": 1.5166,
+      "step": 3433
+    },
+    {
+      "epoch": 0.46867749419953597,
+      "grad_norm": 0.20343711972236633,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 3434
+    },
+    {
+      "epoch": 0.4688139757062918,
+      "grad_norm": 0.20081031322479248,
+      "learning_rate": 8e-05,
+      "loss": 1.4039,
+      "step": 3435
+    },
+    {
+      "epoch": 0.4689504572130476,
+      "grad_norm": 0.2264920473098755,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 3436
+    },
+    {
+      "epoch": 0.46908693871980345,
+      "grad_norm": 0.20996640622615814,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 3437
+    },
+    {
+      "epoch": 0.4692234202265593,
+      "grad_norm": 0.2003016173839569,
+      "learning_rate": 8e-05,
+      "loss": 1.4972,
+      "step": 3438
+    },
+    {
+      "epoch": 0.46935990173331515,
+      "grad_norm": 0.20201997458934784,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 3439
+    },
+    {
+      "epoch": 0.469496383240071,
+      "grad_norm": 0.21266897022724152,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4696328647468268,
+      "grad_norm": 0.20544934272766113,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 3441
+    },
+    {
+      "epoch": 0.46976934625358263,
+      "grad_norm": 0.2060510665178299,
+      "learning_rate": 8e-05,
+      "loss": 1.4759,
+      "step": 3442
+    },
+    {
+      "epoch": 0.46990582776033846,
+      "grad_norm": 0.20486944913864136,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4700423092670943,
+      "grad_norm": 0.20600253343582153,
+      "learning_rate": 8e-05,
+      "loss": 1.5459,
+      "step": 3444
+    },
+    {
+      "epoch": 0.47017879077385016,
+      "grad_norm": 0.20423419773578644,
+      "learning_rate": 8e-05,
+      "loss": 1.4899,
+      "step": 3445
+    },
+    {
+      "epoch": 0.470315272280606,
+      "grad_norm": 0.20719143748283386,
+      "learning_rate": 8e-05,
+      "loss": 1.4813,
+      "step": 3446
+    },
+    {
+      "epoch": 0.4704517537873618,
+      "grad_norm": 0.20277182757854462,
+      "learning_rate": 8e-05,
+      "loss": 1.4333,
+      "step": 3447
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.2142658680677414,
+      "learning_rate": 8e-05,
+      "loss": 1.5602,
+      "step": 3448
+    },
+    {
+      "epoch": 0.47072471680087347,
+      "grad_norm": 0.20700852572917938,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4708611983076293,
+      "grad_norm": 0.20282982289791107,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4709976798143852,
+      "grad_norm": 0.21157848834991455,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 3451
+    },
+    {
+      "epoch": 0.471134161321141,
+      "grad_norm": 0.20206362009048462,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 3452
+    },
+    {
+      "epoch": 0.4712706428278968,
+      "grad_norm": 0.21035797894001007,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 3453
+    },
+    {
+      "epoch": 0.47140712433465265,
+      "grad_norm": 0.2178732007741928,
+      "learning_rate": 8e-05,
+      "loss": 1.5156,
+      "step": 3454
+    },
+    {
+      "epoch": 0.4715436058414085,
+      "grad_norm": 0.19992229342460632,
+      "learning_rate": 8e-05,
+      "loss": 1.4092,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4716800873481643,
+      "grad_norm": 0.20279943943023682,
+      "learning_rate": 8e-05,
+      "loss": 1.4586,
+      "step": 3456
+    },
+    {
+      "epoch": 0.4718165688549202,
+      "grad_norm": 0.21036101877689362,
+      "learning_rate": 8e-05,
+      "loss": 1.5242,
+      "step": 3457
+    },
+    {
+      "epoch": 0.471953050361676,
+      "grad_norm": 0.1957864910364151,
+      "learning_rate": 8e-05,
+      "loss": 1.4407,
+      "step": 3458
+    },
+    {
+      "epoch": 0.47208953186843183,
+      "grad_norm": 0.2215675711631775,
+      "learning_rate": 8e-05,
+      "loss": 1.5938,
+      "step": 3459
+    },
+    {
+      "epoch": 0.47222601337518766,
+      "grad_norm": 0.2071194052696228,
+      "learning_rate": 8e-05,
+      "loss": 1.4956,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4723624948819435,
+      "grad_norm": 0.20393699407577515,
+      "learning_rate": 8e-05,
+      "loss": 1.4817,
+      "step": 3461
+    },
+    {
+      "epoch": 0.4724989763886993,
+      "grad_norm": 0.21055908501148224,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4726354578954552,
+      "grad_norm": 0.20912297070026398,
+      "learning_rate": 8e-05,
+      "loss": 1.5247,
+      "step": 3463
+    },
+    {
+      "epoch": 0.472771939402211,
+      "grad_norm": 0.2044554054737091,
+      "learning_rate": 8e-05,
+      "loss": 1.4924,
+      "step": 3464
+    },
+    {
+      "epoch": 0.47290842090896684,
+      "grad_norm": 0.21440665423870087,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 3465
+    },
+    {
+      "epoch": 0.47304490241572267,
+      "grad_norm": 0.2028266042470932,
+      "learning_rate": 8e-05,
+      "loss": 1.5522,
+      "step": 3466
+    },
+    {
+      "epoch": 0.4731813839224785,
+      "grad_norm": 0.20227700471878052,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 3467
+    },
+    {
+      "epoch": 0.4733178654292343,
+      "grad_norm": 0.2044088989496231,
+      "learning_rate": 8e-05,
+      "loss": 1.5612,
+      "step": 3468
+    },
+    {
+      "epoch": 0.4734543469359902,
+      "grad_norm": 0.21253597736358643,
+      "learning_rate": 8e-05,
+      "loss": 1.5493,
+      "step": 3469
+    },
+    {
+      "epoch": 0.473590828442746,
+      "grad_norm": 0.21428847312927246,
+      "learning_rate": 8e-05,
+      "loss": 1.5311,
+      "step": 3470
+    },
+    {
+      "epoch": 0.47372730994950185,
+      "grad_norm": 0.21106933057308197,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 3471
+    },
+    {
+      "epoch": 0.4738637914562577,
+      "grad_norm": 0.20953063666820526,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 3472
+    },
+    {
+      "epoch": 0.4740002729630135,
+      "grad_norm": 0.20550869405269623,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3473
+    },
+    {
+      "epoch": 0.47413675446976933,
+      "grad_norm": 0.20663829147815704,
+      "learning_rate": 8e-05,
+      "loss": 1.5143,
+      "step": 3474
+    },
+    {
+      "epoch": 0.47427323597652515,
+      "grad_norm": 0.20458601415157318,
+      "learning_rate": 8e-05,
+      "loss": 1.5665,
+      "step": 3475
+    },
+    {
+      "epoch": 0.47440971748328103,
+      "grad_norm": 0.21060355007648468,
+      "learning_rate": 8e-05,
+      "loss": 1.5437,
+      "step": 3476
+    },
+    {
+      "epoch": 0.47454619899003686,
+      "grad_norm": 0.20915454626083374,
+      "learning_rate": 8e-05,
+      "loss": 1.431,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4746826804967927,
+      "grad_norm": 0.20820870995521545,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 3478
+    },
+    {
+      "epoch": 0.4748191620035485,
+      "grad_norm": 0.20602834224700928,
+      "learning_rate": 8e-05,
+      "loss": 1.5115,
+      "step": 3479
+    },
+    {
+      "epoch": 0.47495564351030434,
+      "grad_norm": 0.20320291817188263,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 3480
+    },
+    {
+      "epoch": 0.47509212501706016,
+      "grad_norm": 0.19657786190509796,
+      "learning_rate": 8e-05,
+      "loss": 1.4439,
+      "step": 3481
+    },
+    {
+      "epoch": 0.47522860652381604,
+      "grad_norm": 0.2016376107931137,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 3482
+    },
+    {
+      "epoch": 0.47536508803057187,
+      "grad_norm": 0.20283100008964539,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 3483
+    },
+    {
+      "epoch": 0.4755015695373277,
+      "grad_norm": 0.20347493886947632,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3484
+    },
+    {
+      "epoch": 0.4756380510440835,
+      "grad_norm": 0.20947591960430145,
+      "learning_rate": 8e-05,
+      "loss": 1.52,
+      "step": 3485
+    },
+    {
+      "epoch": 0.47577453255083935,
+      "grad_norm": 0.2025967836380005,
+      "learning_rate": 8e-05,
+      "loss": 1.5315,
+      "step": 3486
+    },
+    {
+      "epoch": 0.47591101405759517,
+      "grad_norm": 0.20849086344242096,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 3487
+    },
+    {
+      "epoch": 0.47604749556435105,
+      "grad_norm": 0.20108284056186676,
+      "learning_rate": 8e-05,
+      "loss": 1.5088,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4761839770711069,
+      "grad_norm": 0.1997194141149521,
+      "learning_rate": 8e-05,
+      "loss": 1.4483,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4763204585778627,
+      "grad_norm": 0.20717458426952362,
+      "learning_rate": 8e-05,
+      "loss": 1.5491,
+      "step": 3490
+    },
+    {
+      "epoch": 0.47645694008461853,
+      "grad_norm": 0.198810875415802,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 3491
+    },
+    {
+      "epoch": 0.47659342159137436,
+      "grad_norm": 0.1989993304014206,
+      "learning_rate": 8e-05,
+      "loss": 1.4878,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4767299030981302,
+      "grad_norm": 0.19711193442344666,
+      "learning_rate": 8e-05,
+      "loss": 1.4602,
+      "step": 3493
+    },
+    {
+      "epoch": 0.47686638460488606,
+      "grad_norm": 0.20564322173595428,
+      "learning_rate": 8e-05,
+      "loss": 1.5432,
+      "step": 3494
+    },
+    {
+      "epoch": 0.4770028661116419,
+      "grad_norm": 0.2079828977584839,
+      "learning_rate": 8e-05,
+      "loss": 1.5008,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4771393476183977,
+      "grad_norm": 0.19164466857910156,
+      "learning_rate": 8e-05,
+      "loss": 1.3999,
+      "step": 3496
+    },
+    {
+      "epoch": 0.47727582912515354,
+      "grad_norm": 0.2024288922548294,
+      "learning_rate": 8e-05,
+      "loss": 1.4604,
+      "step": 3497
+    },
+    {
+      "epoch": 0.47741231063190936,
+      "grad_norm": 0.19840653240680695,
+      "learning_rate": 8e-05,
+      "loss": 1.4869,
+      "step": 3498
+    },
+    {
+      "epoch": 0.4775487921386652,
+      "grad_norm": 0.20264117419719696,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 3499
+    },
+    {
+      "epoch": 0.47768527364542107,
+      "grad_norm": 0.2120170295238495,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4778217551521769,
+      "grad_norm": 0.2042725682258606,
+      "learning_rate": 8e-05,
+      "loss": 1.4688,
+      "step": 3501
+    },
+    {
+      "epoch": 0.4779582366589327,
+      "grad_norm": 0.20023949444293976,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 3502
+    },
+    {
+      "epoch": 0.47809471816568855,
+      "grad_norm": 0.2013224959373474,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 3503
+    },
+    {
+      "epoch": 0.4782311996724444,
+      "grad_norm": 0.202875018119812,
+      "learning_rate": 8e-05,
+      "loss": 1.4423,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4783676811792002,
+      "grad_norm": 0.2079051434993744,
+      "learning_rate": 8e-05,
+      "loss": 1.527,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4785041626859561,
+      "grad_norm": 0.21132634580135345,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 3506
+    },
+    {
+      "epoch": 0.4786406441927119,
+      "grad_norm": 0.20469678938388824,
+      "learning_rate": 8e-05,
+      "loss": 1.4491,
+      "step": 3507
+    },
+    {
+      "epoch": 0.47877712569946773,
+      "grad_norm": 0.20625104010105133,
+      "learning_rate": 8e-05,
+      "loss": 1.5204,
+      "step": 3508
+    },
+    {
+      "epoch": 0.47891360720622356,
+      "grad_norm": 0.1978519707918167,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 3509
+    },
+    {
+      "epoch": 0.4790500887129794,
+      "grad_norm": 0.20960263907909393,
+      "learning_rate": 8e-05,
+      "loss": 1.5739,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4791865702197352,
+      "grad_norm": 0.19807633757591248,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 3511
+    },
+    {
+      "epoch": 0.47932305172649103,
+      "grad_norm": 0.20979318022727966,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4794595332332469,
+      "grad_norm": 0.20573720335960388,
+      "learning_rate": 8e-05,
+      "loss": 1.5056,
+      "step": 3513
+    },
+    {
+      "epoch": 0.47959601474000274,
+      "grad_norm": 0.19832788407802582,
+      "learning_rate": 8e-05,
+      "loss": 1.4183,
+      "step": 3514
+    },
+    {
+      "epoch": 0.47973249624675857,
+      "grad_norm": 0.20368269085884094,
+      "learning_rate": 8e-05,
+      "loss": 1.5001,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4798689777535144,
+      "grad_norm": 0.20572879910469055,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4800054592602702,
+      "grad_norm": 0.2067234367132187,
+      "learning_rate": 8e-05,
+      "loss": 1.5398,
+      "step": 3517
+    },
+    {
+      "epoch": 0.48014194076702604,
+      "grad_norm": 0.21909917891025543,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 3518
+    },
+    {
+      "epoch": 0.4802784222737819,
+      "grad_norm": 0.2101248800754547,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 3519
+    },
+    {
+      "epoch": 0.48041490378053775,
+      "grad_norm": 0.2082640677690506,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4805513852872936,
+      "grad_norm": 0.20257088541984558,
+      "learning_rate": 8e-05,
+      "loss": 1.5267,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4806878667940494,
+      "grad_norm": 0.20171722769737244,
+      "learning_rate": 8e-05,
+      "loss": 1.4967,
+      "step": 3522
+    },
+    {
+      "epoch": 0.4808243483008052,
+      "grad_norm": 0.20478981733322144,
+      "learning_rate": 8e-05,
+      "loss": 1.4735,
+      "step": 3523
+    },
+    {
+      "epoch": 0.48096082980756105,
+      "grad_norm": 0.21053843200206757,
+      "learning_rate": 8e-05,
+      "loss": 1.537,
+      "step": 3524
+    },
+    {
+      "epoch": 0.48109731131431693,
+      "grad_norm": 0.1977248340845108,
+      "learning_rate": 8e-05,
+      "loss": 1.4505,
+      "step": 3525
+    },
+    {
+      "epoch": 0.48123379282107276,
+      "grad_norm": 0.20767652988433838,
+      "learning_rate": 8e-05,
+      "loss": 1.5107,
+      "step": 3526
+    },
+    {
+      "epoch": 0.4813702743278286,
+      "grad_norm": 0.21315601468086243,
+      "learning_rate": 8e-05,
+      "loss": 1.4999,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4815067558345844,
+      "grad_norm": 0.1964656114578247,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 3528
+    },
+    {
+      "epoch": 0.48164323734134024,
+      "grad_norm": 0.217056006193161,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 3529
+    },
+    {
+      "epoch": 0.48177971884809606,
+      "grad_norm": 0.20948797464370728,
+      "learning_rate": 8e-05,
+      "loss": 1.5437,
+      "step": 3530
+    },
+    {
+      "epoch": 0.48191620035485194,
+      "grad_norm": 0.20981857180595398,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 3531
+    },
+    {
+      "epoch": 0.48205268186160777,
+      "grad_norm": 0.21023494005203247,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4821891633683636,
+      "grad_norm": 0.21750009059906006,
+      "learning_rate": 8e-05,
+      "loss": 1.5158,
+      "step": 3533
+    },
+    {
+      "epoch": 0.4823256448751194,
+      "grad_norm": 0.21467478573322296,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 3534
+    },
+    {
+      "epoch": 0.48246212638187524,
+      "grad_norm": 0.21620170772075653,
+      "learning_rate": 8e-05,
+      "loss": 1.4308,
+      "step": 3535
+    },
+    {
+      "epoch": 0.48259860788863107,
+      "grad_norm": 0.21362650394439697,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 3536
+    },
+    {
+      "epoch": 0.48273508939538695,
+      "grad_norm": 0.19977182149887085,
+      "learning_rate": 8e-05,
+      "loss": 1.3927,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4828715709021428,
+      "grad_norm": 0.21269384026527405,
+      "learning_rate": 8e-05,
+      "loss": 1.5823,
+      "step": 3538
+    },
+    {
+      "epoch": 0.4830080524088986,
+      "grad_norm": 0.22291666269302368,
+      "learning_rate": 8e-05,
+      "loss": 1.4763,
+      "step": 3539
+    },
+    {
+      "epoch": 0.48314453391565443,
+      "grad_norm": 0.20446087419986725,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 3540
+    },
+    {
+      "epoch": 0.48328101542241025,
+      "grad_norm": 0.21411104500293732,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 3541
+    },
+    {
+      "epoch": 0.4834174969291661,
+      "grad_norm": 0.20731037855148315,
+      "learning_rate": 8e-05,
+      "loss": 1.4386,
+      "step": 3542
+    },
+    {
+      "epoch": 0.4835539784359219,
+      "grad_norm": 0.20868165791034698,
+      "learning_rate": 8e-05,
+      "loss": 1.5401,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4836904599426778,
+      "grad_norm": 0.21335147321224213,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 3544
+    },
+    {
+      "epoch": 0.4838269414494336,
+      "grad_norm": 0.21525166928768158,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 3545
+    },
+    {
+      "epoch": 0.48396342295618944,
+      "grad_norm": 0.21303965151309967,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 3546
+    },
+    {
+      "epoch": 0.48409990446294526,
+      "grad_norm": 0.19725294411182404,
+      "learning_rate": 8e-05,
+      "loss": 1.4294,
+      "step": 3547
+    },
+    {
+      "epoch": 0.4842363859697011,
+      "grad_norm": 0.20208528637886047,
+      "learning_rate": 8e-05,
+      "loss": 1.4196,
+      "step": 3548
+    },
+    {
+      "epoch": 0.4843728674764569,
+      "grad_norm": 0.20230498909950256,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4845093489832128,
+      "grad_norm": 0.20581379532814026,
+      "learning_rate": 8e-05,
+      "loss": 1.5385,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4846458304899686,
+      "grad_norm": 0.2000308632850647,
+      "learning_rate": 8e-05,
+      "loss": 1.5019,
+      "step": 3551
+    },
+    {
+      "epoch": 0.48478231199672445,
+      "grad_norm": 0.20568643510341644,
+      "learning_rate": 8e-05,
+      "loss": 1.4417,
+      "step": 3552
+    },
+    {
+      "epoch": 0.48491879350348027,
+      "grad_norm": 0.20863598585128784,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 3553
+    },
+    {
+      "epoch": 0.4850552750102361,
+      "grad_norm": 0.208083838224411,
+      "learning_rate": 8e-05,
+      "loss": 1.4835,
+      "step": 3554
+    },
+    {
+      "epoch": 0.4851917565169919,
+      "grad_norm": 0.20798850059509277,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 3555
+    },
+    {
+      "epoch": 0.4853282380237478,
+      "grad_norm": 0.21109066903591156,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 3556
+    },
+    {
+      "epoch": 0.48546471953050363,
+      "grad_norm": 0.2202380746603012,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 3557
+    },
+    {
+      "epoch": 0.48560120103725946,
+      "grad_norm": 0.20535895228385925,
+      "learning_rate": 8e-05,
+      "loss": 1.5163,
+      "step": 3558
+    },
+    {
+      "epoch": 0.4857376825440153,
+      "grad_norm": 0.2063886970281601,
+      "learning_rate": 8e-05,
+      "loss": 1.4956,
+      "step": 3559
+    },
+    {
+      "epoch": 0.4858741640507711,
+      "grad_norm": 0.21430860459804535,
+      "learning_rate": 8e-05,
+      "loss": 1.5613,
+      "step": 3560
+    },
+    {
+      "epoch": 0.48601064555752693,
+      "grad_norm": 0.20108100771903992,
+      "learning_rate": 8e-05,
+      "loss": 1.4646,
+      "step": 3561
+    },
+    {
+      "epoch": 0.4861471270642828,
+      "grad_norm": 0.2038641721010208,
+      "learning_rate": 8e-05,
+      "loss": 1.5428,
+      "step": 3562
+    },
+    {
+      "epoch": 0.48628360857103864,
+      "grad_norm": 0.21385547518730164,
+      "learning_rate": 8e-05,
+      "loss": 1.4901,
+      "step": 3563
+    },
+    {
+      "epoch": 0.48642009007779446,
+      "grad_norm": 0.20411108434200287,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4865565715845503,
+      "grad_norm": 0.20390279591083527,
+      "learning_rate": 8e-05,
+      "loss": 1.5523,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4866930530913061,
+      "grad_norm": 0.20540545880794525,
+      "learning_rate": 8e-05,
+      "loss": 1.5132,
+      "step": 3566
+    },
+    {
+      "epoch": 0.48682953459806194,
+      "grad_norm": 0.20313748717308044,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 3567
+    },
+    {
+      "epoch": 0.4869660161048178,
+      "grad_norm": 0.2098429799079895,
+      "learning_rate": 8e-05,
+      "loss": 1.5084,
+      "step": 3568
+    },
+    {
+      "epoch": 0.48710249761157365,
+      "grad_norm": 0.20528361201286316,
+      "learning_rate": 8e-05,
+      "loss": 1.5273,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4872389791183295,
+      "grad_norm": 0.2076866179704666,
+      "learning_rate": 8e-05,
+      "loss": 1.551,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4873754606250853,
+      "grad_norm": 0.20142139494419098,
+      "learning_rate": 8e-05,
+      "loss": 1.5273,
+      "step": 3571
+    },
+    {
+      "epoch": 0.4875119421318411,
+      "grad_norm": 0.2126702517271042,
+      "learning_rate": 8e-05,
+      "loss": 1.5234,
+      "step": 3572
+    },
+    {
+      "epoch": 0.48764842363859695,
+      "grad_norm": 0.19976341724395752,
+      "learning_rate": 8e-05,
+      "loss": 1.4567,
+      "step": 3573
+    },
+    {
+      "epoch": 0.48778490514535283,
+      "grad_norm": 0.20824775099754333,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 3574
+    },
+    {
+      "epoch": 0.48792138665210866,
+      "grad_norm": 0.21170705556869507,
+      "learning_rate": 8e-05,
+      "loss": 1.4975,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4880578681588645,
+      "grad_norm": 0.20064543187618256,
+      "learning_rate": 8e-05,
+      "loss": 1.4591,
+      "step": 3576
+    },
+    {
+      "epoch": 0.4881943496656203,
+      "grad_norm": 0.2066013514995575,
+      "learning_rate": 8e-05,
+      "loss": 1.5006,
+      "step": 3577
+    },
+    {
+      "epoch": 0.48833083117237613,
+      "grad_norm": 0.20975522696971893,
+      "learning_rate": 8e-05,
+      "loss": 1.4598,
+      "step": 3578
+    },
+    {
+      "epoch": 0.48846731267913196,
+      "grad_norm": 0.20584852993488312,
+      "learning_rate": 8e-05,
+      "loss": 1.5,
+      "step": 3579
+    },
+    {
+      "epoch": 0.4886037941858878,
+      "grad_norm": 0.20562535524368286,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 3580
+    },
+    {
+      "epoch": 0.48874027569264367,
+      "grad_norm": 0.221099853515625,
+      "learning_rate": 8e-05,
+      "loss": 1.5759,
+      "step": 3581
+    },
+    {
+      "epoch": 0.4888767571993995,
+      "grad_norm": 0.2097145915031433,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 3582
+    },
+    {
+      "epoch": 0.4890132387061553,
+      "grad_norm": 0.2036169171333313,
+      "learning_rate": 8e-05,
+      "loss": 1.5385,
+      "step": 3583
+    },
+    {
+      "epoch": 0.48914972021291114,
+      "grad_norm": 0.21238572895526886,
+      "learning_rate": 8e-05,
+      "loss": 1.4617,
+      "step": 3584
+    },
+    {
+      "epoch": 0.48928620171966697,
+      "grad_norm": 0.2007947415113449,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4894226832264228,
+      "grad_norm": 0.21158722043037415,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 3586
+    },
+    {
+      "epoch": 0.4895591647331787,
+      "grad_norm": 0.21101102232933044,
+      "learning_rate": 8e-05,
+      "loss": 1.4622,
+      "step": 3587
+    },
+    {
+      "epoch": 0.4896956462399345,
+      "grad_norm": 0.21405421197414398,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4898321277466903,
+      "grad_norm": 0.21508747339248657,
+      "learning_rate": 8e-05,
+      "loss": 1.4814,
+      "step": 3589
+    },
+    {
+      "epoch": 0.48996860925344615,
+      "grad_norm": 0.216883584856987,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 3590
+    },
+    {
+      "epoch": 0.490105090760202,
+      "grad_norm": 0.20675033330917358,
+      "learning_rate": 8e-05,
+      "loss": 1.4823,
+      "step": 3591
+    },
+    {
+      "epoch": 0.4902415722669578,
+      "grad_norm": 0.21256797015666962,
+      "learning_rate": 8e-05,
+      "loss": 1.455,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4903780537737137,
+      "grad_norm": 0.21640059351921082,
+      "learning_rate": 8e-05,
+      "loss": 1.5673,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4905145352804695,
+      "grad_norm": 0.21707874536514282,
+      "learning_rate": 8e-05,
+      "loss": 1.5362,
+      "step": 3594
+    },
+    {
+      "epoch": 0.49065101678722534,
+      "grad_norm": 0.20597729086875916,
+      "learning_rate": 8e-05,
+      "loss": 1.4092,
+      "step": 3595
+    },
+    {
+      "epoch": 0.49078749829398116,
+      "grad_norm": 0.20640929043293,
+      "learning_rate": 8e-05,
+      "loss": 1.5204,
+      "step": 3596
+    },
+    {
+      "epoch": 0.490923979800737,
+      "grad_norm": 0.20196007192134857,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 3597
+    },
+    {
+      "epoch": 0.4910604613074928,
+      "grad_norm": 0.21097774803638458,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4911969428142487,
+      "grad_norm": 0.2047613561153412,
+      "learning_rate": 8e-05,
+      "loss": 1.5244,
+      "step": 3599
+    },
+    {
+      "epoch": 0.4913334243210045,
+      "grad_norm": 0.20047776401042938,
+      "learning_rate": 8e-05,
+      "loss": 1.495,
+      "step": 3600
+    },
+    {
+      "epoch": 0.49146990582776034,
+      "grad_norm": 0.20288142561912537,
+      "learning_rate": 8e-05,
+      "loss": 1.4614,
+      "step": 3601
+    },
+    {
+      "epoch": 0.49160638733451617,
+      "grad_norm": 0.20047034323215485,
+      "learning_rate": 8e-05,
+      "loss": 1.5124,
+      "step": 3602
+    },
+    {
+      "epoch": 0.491742868841272,
+      "grad_norm": 0.20609702169895172,
+      "learning_rate": 8e-05,
+      "loss": 1.5111,
+      "step": 3603
+    },
+    {
+      "epoch": 0.4918793503480278,
+      "grad_norm": 0.20599691569805145,
+      "learning_rate": 8e-05,
+      "loss": 1.4717,
+      "step": 3604
+    },
+    {
+      "epoch": 0.4920158318547837,
+      "grad_norm": 0.21311761438846588,
+      "learning_rate": 8e-05,
+      "loss": 1.5359,
+      "step": 3605
+    },
+    {
+      "epoch": 0.49215231336153953,
+      "grad_norm": 0.20836657285690308,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 3606
+    },
+    {
+      "epoch": 0.49228879486829535,
+      "grad_norm": 0.20738425850868225,
+      "learning_rate": 8e-05,
+      "loss": 1.5274,
+      "step": 3607
+    },
+    {
+      "epoch": 0.4924252763750512,
+      "grad_norm": 0.20239634811878204,
+      "learning_rate": 8e-05,
+      "loss": 1.4545,
+      "step": 3608
+    },
+    {
+      "epoch": 0.492561757881807,
+      "grad_norm": 0.20028546452522278,
+      "learning_rate": 8e-05,
+      "loss": 1.5029,
+      "step": 3609
+    },
+    {
+      "epoch": 0.49269823938856283,
+      "grad_norm": 0.20458798110485077,
+      "learning_rate": 8e-05,
+      "loss": 1.4169,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4928347208953187,
+      "grad_norm": 0.21394610404968262,
+      "learning_rate": 8e-05,
+      "loss": 1.5116,
+      "step": 3611
+    },
+    {
+      "epoch": 0.49297120240207454,
+      "grad_norm": 0.2124284952878952,
+      "learning_rate": 8e-05,
+      "loss": 1.5507,
+      "step": 3612
+    },
+    {
+      "epoch": 0.49310768390883036,
+      "grad_norm": 0.20730768144130707,
+      "learning_rate": 8e-05,
+      "loss": 1.5256,
+      "step": 3613
+    },
+    {
+      "epoch": 0.4932441654155862,
+      "grad_norm": 0.2099892944097519,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 3614
+    },
+    {
+      "epoch": 0.493380646922342,
+      "grad_norm": 0.21305988729000092,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 3615
+    },
+    {
+      "epoch": 0.49351712842909784,
+      "grad_norm": 0.21803003549575806,
+      "learning_rate": 8e-05,
+      "loss": 1.4666,
+      "step": 3616
+    },
+    {
+      "epoch": 0.49365360993585367,
+      "grad_norm": 0.2166893631219864,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 3617
+    },
+    {
+      "epoch": 0.49379009144260955,
+      "grad_norm": 0.20721480250358582,
+      "learning_rate": 8e-05,
+      "loss": 1.5001,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4939265729493654,
+      "grad_norm": 0.2038077861070633,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4940630544561212,
+      "grad_norm": 0.20818711817264557,
+      "learning_rate": 8e-05,
+      "loss": 1.5357,
+      "step": 3620
+    },
+    {
+      "epoch": 0.494199535962877,
+      "grad_norm": 0.20612770318984985,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 3621
+    },
+    {
+      "epoch": 0.49433601746963285,
+      "grad_norm": 0.20260211825370789,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 3622
+    },
+    {
+      "epoch": 0.4944724989763887,
+      "grad_norm": 0.20106421411037445,
+      "learning_rate": 8e-05,
+      "loss": 1.4803,
+      "step": 3623
+    },
+    {
+      "epoch": 0.49460898048314456,
+      "grad_norm": 0.2095767706632614,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 3624
+    },
+    {
+      "epoch": 0.4947454619899004,
+      "grad_norm": 0.20170405507087708,
+      "learning_rate": 8e-05,
+      "loss": 1.4637,
+      "step": 3625
+    },
+    {
+      "epoch": 0.4948819434966562,
+      "grad_norm": 0.2099498212337494,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 3626
+    },
+    {
+      "epoch": 0.49501842500341203,
+      "grad_norm": 0.20306894183158875,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 3627
+    },
+    {
+      "epoch": 0.49515490651016786,
+      "grad_norm": 0.21352609992027283,
+      "learning_rate": 8e-05,
+      "loss": 1.4947,
+      "step": 3628
+    },
+    {
+      "epoch": 0.4952913880169237,
+      "grad_norm": 0.20478996634483337,
+      "learning_rate": 8e-05,
+      "loss": 1.4511,
+      "step": 3629
+    },
+    {
+      "epoch": 0.49542786952367956,
+      "grad_norm": 0.21442031860351562,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4955643510304354,
+      "grad_norm": 0.21367311477661133,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 3631
+    },
+    {
+      "epoch": 0.4957008325371912,
+      "grad_norm": 0.20358863472938538,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 3632
+    },
+    {
+      "epoch": 0.49583731404394704,
+      "grad_norm": 0.21455514430999756,
+      "learning_rate": 8e-05,
+      "loss": 1.4692,
+      "step": 3633
+    },
+    {
+      "epoch": 0.49597379555070287,
+      "grad_norm": 0.20517876744270325,
+      "learning_rate": 8e-05,
+      "loss": 1.4596,
+      "step": 3634
+    },
+    {
+      "epoch": 0.4961102770574587,
+      "grad_norm": 0.20770393311977386,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4962467585642146,
+      "grad_norm": 0.21083983778953552,
+      "learning_rate": 8e-05,
+      "loss": 1.4478,
+      "step": 3636
+    },
+    {
+      "epoch": 0.4963832400709704,
+      "grad_norm": 0.20676255226135254,
+      "learning_rate": 8e-05,
+      "loss": 1.4813,
+      "step": 3637
+    },
+    {
+      "epoch": 0.4965197215777262,
+      "grad_norm": 0.20775020122528076,
+      "learning_rate": 8e-05,
+      "loss": 1.4894,
+      "step": 3638
+    },
+    {
+      "epoch": 0.49665620308448205,
+      "grad_norm": 0.20101529359817505,
+      "learning_rate": 8e-05,
+      "loss": 1.4659,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4967926845912379,
+      "grad_norm": 0.20888584852218628,
+      "learning_rate": 8e-05,
+      "loss": 1.4227,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4969291660979937,
+      "grad_norm": 0.20747052133083344,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 3641
+    },
+    {
+      "epoch": 0.4970656476047496,
+      "grad_norm": 0.19870388507843018,
+      "learning_rate": 8e-05,
+      "loss": 1.4293,
+      "step": 3642
+    },
+    {
+      "epoch": 0.4972021291115054,
+      "grad_norm": 0.2139759659767151,
+      "learning_rate": 8e-05,
+      "loss": 1.5168,
+      "step": 3643
+    },
+    {
+      "epoch": 0.49733861061826123,
+      "grad_norm": 0.20687726140022278,
+      "learning_rate": 8e-05,
+      "loss": 1.4944,
+      "step": 3644
+    },
+    {
+      "epoch": 0.49747509212501706,
+      "grad_norm": 0.20504942536354065,
+      "learning_rate": 8e-05,
+      "loss": 1.4031,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4976115736317729,
+      "grad_norm": 0.2088628113269806,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 3646
+    },
+    {
+      "epoch": 0.4977480551385287,
+      "grad_norm": 0.20516029000282288,
+      "learning_rate": 8e-05,
+      "loss": 1.4965,
+      "step": 3647
+    },
+    {
+      "epoch": 0.49788453664528454,
+      "grad_norm": 0.20718146860599518,
+      "learning_rate": 8e-05,
+      "loss": 1.498,
+      "step": 3648
+    },
+    {
+      "epoch": 0.4980210181520404,
+      "grad_norm": 0.20014266669750214,
+      "learning_rate": 8e-05,
+      "loss": 1.4571,
+      "step": 3649
+    },
+    {
+      "epoch": 0.49815749965879624,
+      "grad_norm": 0.20117169618606567,
+      "learning_rate": 8e-05,
+      "loss": 1.4921,
+      "step": 3650
+    },
+    {
+      "epoch": 0.49829398116555207,
+      "grad_norm": 0.21267274022102356,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4984304626723079,
+      "grad_norm": 0.20519742369651794,
+      "learning_rate": 8e-05,
+      "loss": 1.5216,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4985669441790637,
+      "grad_norm": 0.20292620360851288,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 3653
+    },
+    {
+      "epoch": 0.49870342568581955,
+      "grad_norm": 0.2203458547592163,
+      "learning_rate": 8e-05,
+      "loss": 1.5247,
+      "step": 3654
+    },
+    {
+      "epoch": 0.4988399071925754,
+      "grad_norm": 0.2080739289522171,
+      "learning_rate": 8e-05,
+      "loss": 1.5376,
+      "step": 3655
+    },
+    {
+      "epoch": 0.49897638869933125,
+      "grad_norm": 0.21375824511051178,
+      "learning_rate": 8e-05,
+      "loss": 1.5589,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4991128702060871,
+      "grad_norm": 0.20661212503910065,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 3657
+    },
+    {
+      "epoch": 0.4992493517128429,
+      "grad_norm": 0.1998106688261032,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3658
+    },
+    {
+      "epoch": 0.49938583321959873,
+      "grad_norm": 0.20285949110984802,
+      "learning_rate": 8e-05,
+      "loss": 1.5179,
+      "step": 3659
+    },
+    {
+      "epoch": 0.49952231472635455,
+      "grad_norm": 0.20221386849880219,
+      "learning_rate": 8e-05,
+      "loss": 1.5154,
+      "step": 3660
+    },
+    {
+      "epoch": 0.49965879623311044,
+      "grad_norm": 0.19674454629421234,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 3661
+    },
+    {
+      "epoch": 0.49979527773986626,
+      "grad_norm": 0.20864376425743103,
+      "learning_rate": 8e-05,
+      "loss": 1.5716,
+      "step": 3662
+    },
+    {
+      "epoch": 0.4999317592466221,
+      "grad_norm": 0.20189069211483002,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 3663
+    },
+    {
+      "epoch": 0.500068240753378,
+      "grad_norm": 0.20689326524734497,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5002047222601338,
+      "grad_norm": 0.21195192635059357,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5003412037668896,
+      "grad_norm": 0.20739254355430603,
+      "learning_rate": 8e-05,
+      "loss": 1.4884,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5004776852736454,
+      "grad_norm": 0.21032270789146423,
+      "learning_rate": 8e-05,
+      "loss": 1.525,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5006141667804013,
+      "grad_norm": 0.2124892622232437,
+      "learning_rate": 8e-05,
+      "loss": 1.5486,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5007506482871571,
+      "grad_norm": 0.21108348667621613,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5008871297939129,
+      "grad_norm": 0.2078608274459839,
+      "learning_rate": 8e-05,
+      "loss": 1.4826,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5010236113006687,
+      "grad_norm": 0.20696720480918884,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5011600928074246,
+      "grad_norm": 0.22207775712013245,
+      "learning_rate": 8e-05,
+      "loss": 1.4824,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5012965743141804,
+      "grad_norm": 0.20768852531909943,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5014330558209362,
+      "grad_norm": 0.20559285581111908,
+      "learning_rate": 8e-05,
+      "loss": 1.5003,
+      "step": 3674
+    },
+    {
+      "epoch": 0.501569537327692,
+      "grad_norm": 0.21189770102500916,
+      "learning_rate": 8e-05,
+      "loss": 1.5065,
+      "step": 3675
+    },
+    {
+      "epoch": 0.501706018834448,
+      "grad_norm": 0.21071873605251312,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 3676
+    },
+    {
+      "epoch": 0.5018425003412038,
+      "grad_norm": 0.19959363341331482,
+      "learning_rate": 8e-05,
+      "loss": 1.4835,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5019789818479596,
+      "grad_norm": 0.21146203577518463,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5021154633547155,
+      "grad_norm": 0.21805863082408905,
+      "learning_rate": 8e-05,
+      "loss": 1.4555,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5022519448614713,
+      "grad_norm": 0.21758250892162323,
+      "learning_rate": 8e-05,
+      "loss": 1.5638,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5023884263682271,
+      "grad_norm": 0.21342754364013672,
+      "learning_rate": 8e-05,
+      "loss": 1.4852,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5025249078749829,
+      "grad_norm": 0.20522929728031158,
+      "learning_rate": 8e-05,
+      "loss": 1.5125,
+      "step": 3682
+    },
+    {
+      "epoch": 0.5026613893817388,
+      "grad_norm": 0.19855450093746185,
+      "learning_rate": 8e-05,
+      "loss": 1.444,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5027978708884946,
+      "grad_norm": 0.2293214350938797,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5029343523952504,
+      "grad_norm": 0.21070286631584167,
+      "learning_rate": 8e-05,
+      "loss": 1.5231,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5030708339020062,
+      "grad_norm": 0.19877244532108307,
+      "learning_rate": 8e-05,
+      "loss": 1.462,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5032073154087621,
+      "grad_norm": 0.202780619263649,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5033437969155179,
+      "grad_norm": 0.21033748984336853,
+      "learning_rate": 8e-05,
+      "loss": 1.4226,
+      "step": 3688
+    },
+    {
+      "epoch": 0.5034802784222738,
+      "grad_norm": 0.20581790804862976,
+      "learning_rate": 8e-05,
+      "loss": 1.525,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5036167599290297,
+      "grad_norm": 0.21243390440940857,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5037532414357855,
+      "grad_norm": 0.2094869315624237,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5038897229425413,
+      "grad_norm": 0.20795494318008423,
+      "learning_rate": 8e-05,
+      "loss": 1.5501,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5040262044492971,
+      "grad_norm": 0.2130076289176941,
+      "learning_rate": 8e-05,
+      "loss": 1.4421,
+      "step": 3693
+    },
+    {
+      "epoch": 0.504162685956053,
+      "grad_norm": 0.21163424849510193,
+      "learning_rate": 8e-05,
+      "loss": 1.5637,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5042991674628088,
+      "grad_norm": 0.20242011547088623,
+      "learning_rate": 8e-05,
+      "loss": 1.507,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5044356489695646,
+      "grad_norm": 0.20719332993030548,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5045721304763204,
+      "grad_norm": 0.213754802942276,
+      "learning_rate": 8e-05,
+      "loss": 1.4698,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5047086119830763,
+      "grad_norm": 0.2040412873029709,
+      "learning_rate": 8e-05,
+      "loss": 1.4053,
+      "step": 3698
+    },
+    {
+      "epoch": 0.5048450934898321,
+      "grad_norm": 0.2151033878326416,
+      "learning_rate": 8e-05,
+      "loss": 1.516,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5049815749965879,
+      "grad_norm": 0.21601992845535278,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5051180565033438,
+      "grad_norm": 0.2160337269306183,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5052545380100997,
+      "grad_norm": 0.21680496633052826,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5053910195168555,
+      "grad_norm": 0.20526234805583954,
+      "learning_rate": 8e-05,
+      "loss": 1.4607,
+      "step": 3703
+    },
+    {
+      "epoch": 0.5055275010236113,
+      "grad_norm": 0.2081543207168579,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 3704
+    },
+    {
+      "epoch": 0.5056639825303672,
+      "grad_norm": 0.2106073647737503,
+      "learning_rate": 8e-05,
+      "loss": 1.5081,
+      "step": 3705
+    },
+    {
+      "epoch": 0.505800464037123,
+      "grad_norm": 0.21469458937644958,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 3706
+    },
+    {
+      "epoch": 0.5059369455438788,
+      "grad_norm": 0.20671242475509644,
+      "learning_rate": 8e-05,
+      "loss": 1.4682,
+      "step": 3707
+    },
+    {
+      "epoch": 0.5060734270506346,
+      "grad_norm": 0.20423923432826996,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5062099085573905,
+      "grad_norm": 0.21102315187454224,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5063463900641463,
+      "grad_norm": 0.2006913423538208,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5064828715709021,
+      "grad_norm": 0.2073921263217926,
+      "learning_rate": 8e-05,
+      "loss": 1.4356,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5066193530776579,
+      "grad_norm": 0.21121962368488312,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 3712
+    },
+    {
+      "epoch": 0.5067558345844139,
+      "grad_norm": 0.20361319184303284,
+      "learning_rate": 8e-05,
+      "loss": 1.5119,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5068923160911697,
+      "grad_norm": 0.20557788014411926,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5070287975979255,
+      "grad_norm": 0.20878492295742035,
+      "learning_rate": 8e-05,
+      "loss": 1.523,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5071652791046813,
+      "grad_norm": 0.20175319910049438,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 3716
+    },
+    {
+      "epoch": 0.5073017606114372,
+      "grad_norm": 0.2210397571325302,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 3717
+    },
+    {
+      "epoch": 0.507438242118193,
+      "grad_norm": 0.20550814270973206,
+      "learning_rate": 8e-05,
+      "loss": 1.4851,
+      "step": 3718
+    },
+    {
+      "epoch": 0.5075747236249488,
+      "grad_norm": 0.20879559218883514,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5077112051317046,
+      "grad_norm": 0.2052595466375351,
+      "learning_rate": 8e-05,
+      "loss": 1.4464,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5078476866384605,
+      "grad_norm": 0.21076737344264984,
+      "learning_rate": 8e-05,
+      "loss": 1.5255,
+      "step": 3721
+    },
+    {
+      "epoch": 0.5079841681452163,
+      "grad_norm": 0.20953722298145294,
+      "learning_rate": 8e-05,
+      "loss": 1.5138,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5081206496519721,
+      "grad_norm": 0.20284907519817352,
+      "learning_rate": 8e-05,
+      "loss": 1.46,
+      "step": 3723
+    },
+    {
+      "epoch": 0.508257131158728,
+      "grad_norm": 0.21095138788223267,
+      "learning_rate": 8e-05,
+      "loss": 1.4421,
+      "step": 3724
+    },
+    {
+      "epoch": 0.5083936126654838,
+      "grad_norm": 0.20719744265079498,
+      "learning_rate": 8e-05,
+      "loss": 1.5167,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5085300941722397,
+      "grad_norm": 0.21564018726348877,
+      "learning_rate": 8e-05,
+      "loss": 1.4899,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5086665756789955,
+      "grad_norm": 0.20956425368785858,
+      "learning_rate": 8e-05,
+      "loss": 1.4464,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5088030571857514,
+      "grad_norm": 0.21330702304840088,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 3728
+    },
+    {
+      "epoch": 0.5089395386925072,
+      "grad_norm": 0.2078198343515396,
+      "learning_rate": 8e-05,
+      "loss": 1.5246,
+      "step": 3729
+    },
+    {
+      "epoch": 0.509076020199263,
+      "grad_norm": 0.20376670360565186,
+      "learning_rate": 8e-05,
+      "loss": 1.5099,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5092125017060188,
+      "grad_norm": 0.206795334815979,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5093489832127747,
+      "grad_norm": 0.20798687636852264,
+      "learning_rate": 8e-05,
+      "loss": 1.5274,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5094854647195305,
+      "grad_norm": 0.21329432725906372,
+      "learning_rate": 8e-05,
+      "loss": 1.4971,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5096219462262863,
+      "grad_norm": 0.20452523231506348,
+      "learning_rate": 8e-05,
+      "loss": 1.5064,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5097584277330421,
+      "grad_norm": 0.2125357836484909,
+      "learning_rate": 8e-05,
+      "loss": 1.4141,
+      "step": 3735
+    },
+    {
+      "epoch": 0.509894909239798,
+      "grad_norm": 0.20367489755153656,
+      "learning_rate": 8e-05,
+      "loss": 1.4489,
+      "step": 3736
+    },
+    {
+      "epoch": 0.5100313907465538,
+      "grad_norm": 0.2023291438817978,
+      "learning_rate": 8e-05,
+      "loss": 1.4869,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5101678722533097,
+      "grad_norm": 0.20648032426834106,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 3738
+    },
+    {
+      "epoch": 0.5103043537600656,
+      "grad_norm": 0.2120596319437027,
+      "learning_rate": 8e-05,
+      "loss": 1.5264,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5104408352668214,
+      "grad_norm": 0.20421461760997772,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5105773167735772,
+      "grad_norm": 0.2109621912240982,
+      "learning_rate": 8e-05,
+      "loss": 1.4056,
+      "step": 3741
+    },
+    {
+      "epoch": 0.510713798280333,
+      "grad_norm": 0.20625890791416168,
+      "learning_rate": 8e-05,
+      "loss": 1.4449,
+      "step": 3742
+    },
+    {
+      "epoch": 0.5108502797870889,
+      "grad_norm": 0.21244002878665924,
+      "learning_rate": 8e-05,
+      "loss": 1.4772,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5109867612938447,
+      "grad_norm": 0.2075747400522232,
+      "learning_rate": 8e-05,
+      "loss": 1.4904,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5111232428006005,
+      "grad_norm": 0.22499270737171173,
+      "learning_rate": 8e-05,
+      "loss": 1.4971,
+      "step": 3745
+    },
+    {
+      "epoch": 0.5112597243073563,
+      "grad_norm": 0.20978441834449768,
+      "learning_rate": 8e-05,
+      "loss": 1.5155,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5113962058141122,
+      "grad_norm": 0.215524822473526,
+      "learning_rate": 8e-05,
+      "loss": 1.4872,
+      "step": 3747
+    },
+    {
+      "epoch": 0.511532687320868,
+      "grad_norm": 0.21619375050067902,
+      "learning_rate": 8e-05,
+      "loss": 1.4497,
+      "step": 3748
+    },
+    {
+      "epoch": 0.5116691688276238,
+      "grad_norm": 0.20606496930122375,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5118056503343797,
+      "grad_norm": 0.21071958541870117,
+      "learning_rate": 8e-05,
+      "loss": 1.483,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5119421318411356,
+      "grad_norm": 0.20513364672660828,
+      "learning_rate": 8e-05,
+      "loss": 1.3652,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5120786133478914,
+      "grad_norm": 0.2093500792980194,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5122150948546472,
+      "grad_norm": 0.20986369252204895,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 3753
+    },
+    {
+      "epoch": 0.512351576361403,
+      "grad_norm": 0.2090405821800232,
+      "learning_rate": 8e-05,
+      "loss": 1.5341,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5124880578681589,
+      "grad_norm": 0.20310235023498535,
+      "learning_rate": 8e-05,
+      "loss": 1.4269,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5126245393749147,
+      "grad_norm": 0.21172383427619934,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5127610208816705,
+      "grad_norm": 0.21122828125953674,
+      "learning_rate": 8e-05,
+      "loss": 1.4362,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5128975023884264,
+      "grad_norm": 0.2103128582239151,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5130339838951822,
+      "grad_norm": 0.21107755601406097,
+      "learning_rate": 8e-05,
+      "loss": 1.4483,
+      "step": 3759
+    },
+    {
+      "epoch": 0.513170465401938,
+      "grad_norm": 0.20862391591072083,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5133069469086938,
+      "grad_norm": 0.21597601473331451,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5134434284154497,
+      "grad_norm": 0.2235720455646515,
+      "learning_rate": 8e-05,
+      "loss": 1.5069,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5135799099222056,
+      "grad_norm": 0.2106025516986847,
+      "learning_rate": 8e-05,
+      "loss": 1.4861,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5137163914289614,
+      "grad_norm": 0.19796723127365112,
+      "learning_rate": 8e-05,
+      "loss": 1.3393,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5138528729357172,
+      "grad_norm": 0.22598546743392944,
+      "learning_rate": 8e-05,
+      "loss": 1.4736,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5139893544424731,
+      "grad_norm": 0.21574561297893524,
+      "learning_rate": 8e-05,
+      "loss": 1.4597,
+      "step": 3766
+    },
+    {
+      "epoch": 0.5141258359492289,
+      "grad_norm": 0.21665695309638977,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5142623174559847,
+      "grad_norm": 0.2258956879377365,
+      "learning_rate": 8e-05,
+      "loss": 1.5133,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5143987989627405,
+      "grad_norm": 0.20740656554698944,
+      "learning_rate": 8e-05,
+      "loss": 1.4984,
+      "step": 3769
+    },
+    {
+      "epoch": 0.5145352804694964,
+      "grad_norm": 0.2127566784620285,
+      "learning_rate": 8e-05,
+      "loss": 1.4984,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5146717619762522,
+      "grad_norm": 0.21286697685718536,
+      "learning_rate": 8e-05,
+      "loss": 1.4315,
+      "step": 3771
+    },
+    {
+      "epoch": 0.514808243483008,
+      "grad_norm": 0.20938505232334137,
+      "learning_rate": 8e-05,
+      "loss": 1.4674,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5149447249897638,
+      "grad_norm": 0.2103201299905777,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5150812064965197,
+      "grad_norm": 0.2165060192346573,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5152176880032756,
+      "grad_norm": 0.20270776748657227,
+      "learning_rate": 8e-05,
+      "loss": 1.4778,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5153541695100314,
+      "grad_norm": 0.2126288115978241,
+      "learning_rate": 8e-05,
+      "loss": 1.5109,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5154906510167873,
+      "grad_norm": 0.21846488118171692,
+      "learning_rate": 8e-05,
+      "loss": 1.5282,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5156271325235431,
+      "grad_norm": 0.20854903757572174,
+      "learning_rate": 8e-05,
+      "loss": 1.5025,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5157636140302989,
+      "grad_norm": 0.194810152053833,
+      "learning_rate": 8e-05,
+      "loss": 1.4112,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5159000955370547,
+      "grad_norm": 0.21077163517475128,
+      "learning_rate": 8e-05,
+      "loss": 1.5123,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5160365770438106,
+      "grad_norm": 0.20316071808338165,
+      "learning_rate": 8e-05,
+      "loss": 1.4156,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5161730585505664,
+      "grad_norm": 0.21339119970798492,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5163095400573222,
+      "grad_norm": 0.20145781338214874,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 3783
+    },
+    {
+      "epoch": 0.516446021564078,
+      "grad_norm": 0.22384998202323914,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5165825030708339,
+      "grad_norm": 0.2159966081380844,
+      "learning_rate": 8e-05,
+      "loss": 1.4336,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5167189845775897,
+      "grad_norm": 0.2209702730178833,
+      "learning_rate": 8e-05,
+      "loss": 1.5007,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5168554660843455,
+      "grad_norm": 0.23043642938137054,
+      "learning_rate": 8e-05,
+      "loss": 1.4736,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5169919475911015,
+      "grad_norm": 0.21323435008525848,
+      "learning_rate": 8e-05,
+      "loss": 1.4337,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5171284290978573,
+      "grad_norm": 0.21257047355175018,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5172649106046131,
+      "grad_norm": 0.2186005413532257,
+      "learning_rate": 8e-05,
+      "loss": 1.4627,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5174013921113689,
+      "grad_norm": 0.22197014093399048,
+      "learning_rate": 8e-05,
+      "loss": 1.5594,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5175378736181248,
+      "grad_norm": 0.20388200879096985,
+      "learning_rate": 8e-05,
+      "loss": 1.421,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5176743551248806,
+      "grad_norm": 0.216207817196846,
+      "learning_rate": 8e-05,
+      "loss": 1.499,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5178108366316364,
+      "grad_norm": 0.22408992052078247,
+      "learning_rate": 8e-05,
+      "loss": 1.4913,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5179473181383922,
+      "grad_norm": 0.2174125611782074,
+      "learning_rate": 8e-05,
+      "loss": 1.5571,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5180837996451481,
+      "grad_norm": 0.20621757209300995,
+      "learning_rate": 8e-05,
+      "loss": 1.48,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5182202811519039,
+      "grad_norm": 0.21526406705379486,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5183567626586597,
+      "grad_norm": 0.20705653727054596,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5184932441654155,
+      "grad_norm": 0.21425952017307281,
+      "learning_rate": 8e-05,
+      "loss": 1.5011,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5186297256721715,
+      "grad_norm": 0.20879510045051575,
+      "learning_rate": 8e-05,
+      "loss": 1.4366,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5187662071789273,
+      "grad_norm": 0.21131199598312378,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5189026886856831,
+      "grad_norm": 0.2092132568359375,
+      "learning_rate": 8e-05,
+      "loss": 1.4868,
+      "step": 3802
+    },
+    {
+      "epoch": 0.519039170192439,
+      "grad_norm": 0.2116786539554596,
+      "learning_rate": 8e-05,
+      "loss": 1.532,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5191756516991948,
+      "grad_norm": 0.212627574801445,
+      "learning_rate": 8e-05,
+      "loss": 1.5068,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5193121332059506,
+      "grad_norm": 0.20355840027332306,
+      "learning_rate": 8e-05,
+      "loss": 1.4063,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5194486147127064,
+      "grad_norm": 0.21222664415836334,
+      "learning_rate": 8e-05,
+      "loss": 1.4489,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5195850962194623,
+      "grad_norm": 0.21173831820487976,
+      "learning_rate": 8e-05,
+      "loss": 1.5613,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5197215777262181,
+      "grad_norm": 0.21133606135845184,
+      "learning_rate": 8e-05,
+      "loss": 1.5506,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5198580592329739,
+      "grad_norm": 0.2097448706626892,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5199945407397297,
+      "grad_norm": 0.2089434266090393,
+      "learning_rate": 8e-05,
+      "loss": 1.5649,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5201310222464856,
+      "grad_norm": 0.2001953423023224,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5202675037532415,
+      "grad_norm": 0.20784500241279602,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5204039852599973,
+      "grad_norm": 0.20602302253246307,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5205404667667531,
+      "grad_norm": 0.21282580494880676,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 3814
+    },
+    {
+      "epoch": 0.520676948273509,
+      "grad_norm": 0.20751439034938812,
+      "learning_rate": 8e-05,
+      "loss": 1.417,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5208134297802648,
+      "grad_norm": 0.20758989453315735,
+      "learning_rate": 8e-05,
+      "loss": 1.4878,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5209499112870206,
+      "grad_norm": 0.21621976792812347,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5210863927937764,
+      "grad_norm": 0.21620725095272064,
+      "learning_rate": 8e-05,
+      "loss": 1.459,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5212228743005323,
+      "grad_norm": 0.2067711502313614,
+      "learning_rate": 8e-05,
+      "loss": 1.4068,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5213593558072881,
+      "grad_norm": 0.21770837903022766,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5214958373140439,
+      "grad_norm": 0.21227087080478668,
+      "learning_rate": 8e-05,
+      "loss": 1.5182,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5216323188207997,
+      "grad_norm": 0.20959775149822235,
+      "learning_rate": 8e-05,
+      "loss": 1.5138,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5217688003275556,
+      "grad_norm": 0.2112349569797516,
+      "learning_rate": 8e-05,
+      "loss": 1.4204,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5219052818343114,
+      "grad_norm": 0.2108369767665863,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5220417633410673,
+      "grad_norm": 0.21882425248622894,
+      "learning_rate": 8e-05,
+      "loss": 1.4691,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5221782448478232,
+      "grad_norm": 0.21501420438289642,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 3826
+    },
+    {
+      "epoch": 0.522314726354579,
+      "grad_norm": 0.21300239861011505,
+      "learning_rate": 8e-05,
+      "loss": 1.4638,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5224512078613348,
+      "grad_norm": 0.22172772884368896,
+      "learning_rate": 8e-05,
+      "loss": 1.56,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5225876893680906,
+      "grad_norm": 0.22500614821910858,
+      "learning_rate": 8e-05,
+      "loss": 1.487,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5227241708748465,
+      "grad_norm": 0.21296440064907074,
+      "learning_rate": 8e-05,
+      "loss": 1.4943,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5228606523816023,
+      "grad_norm": 0.21398961544036865,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5229971338883581,
+      "grad_norm": 0.2164476215839386,
+      "learning_rate": 8e-05,
+      "loss": 1.4311,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5231336153951139,
+      "grad_norm": 0.2061043381690979,
+      "learning_rate": 8e-05,
+      "loss": 1.5311,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5232700969018698,
+      "grad_norm": 0.2058311253786087,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5234065784086256,
+      "grad_norm": 0.21108292043209076,
+      "learning_rate": 8e-05,
+      "loss": 1.4624,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5235430599153814,
+      "grad_norm": 0.21557392179965973,
+      "learning_rate": 8e-05,
+      "loss": 1.5074,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5236795414221374,
+      "grad_norm": 0.21924269199371338,
+      "learning_rate": 8e-05,
+      "loss": 1.4889,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5238160229288932,
+      "grad_norm": 0.21423862874507904,
+      "learning_rate": 8e-05,
+      "loss": 1.4989,
+      "step": 3838
+    },
+    {
+      "epoch": 0.523952504435649,
+      "grad_norm": 0.20681478083133698,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5240889859424048,
+      "grad_norm": 0.2093067318201065,
+      "learning_rate": 8e-05,
+      "loss": 1.4896,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5242254674491607,
+      "grad_norm": 0.20795153081417084,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5243619489559165,
+      "grad_norm": 0.20789562165737152,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5244984304626723,
+      "grad_norm": 0.2139589488506317,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5246349119694281,
+      "grad_norm": 0.2126191109418869,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 3844
+    },
+    {
+      "epoch": 0.524771393476184,
+      "grad_norm": 0.2106030136346817,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5249078749829398,
+      "grad_norm": 0.20340067148208618,
+      "learning_rate": 8e-05,
+      "loss": 1.4368,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5250443564896956,
+      "grad_norm": 0.20585723221302032,
+      "learning_rate": 8e-05,
+      "loss": 1.4946,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5251808379964514,
+      "grad_norm": 0.20665878057479858,
+      "learning_rate": 8e-05,
+      "loss": 1.3927,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5253173195032074,
+      "grad_norm": 0.20955102145671844,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5254538010099632,
+      "grad_norm": 0.21391011774539948,
+      "learning_rate": 8e-05,
+      "loss": 1.5452,
+      "step": 3850
+    },
+    {
+      "epoch": 0.525590282516719,
+      "grad_norm": 0.21142229437828064,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5257267640234748,
+      "grad_norm": 0.20864613354206085,
+      "learning_rate": 8e-05,
+      "loss": 1.5329,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5258632455302307,
+      "grad_norm": 0.20696291327476501,
+      "learning_rate": 8e-05,
+      "loss": 1.4782,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5259997270369865,
+      "grad_norm": 0.20553968846797943,
+      "learning_rate": 8e-05,
+      "loss": 1.4263,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5261362085437423,
+      "grad_norm": 0.20468705892562866,
+      "learning_rate": 8e-05,
+      "loss": 1.4806,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5262726900504981,
+      "grad_norm": 0.20709489285945892,
+      "learning_rate": 8e-05,
+      "loss": 1.5025,
+      "step": 3856
+    },
+    {
+      "epoch": 0.526409171557254,
+      "grad_norm": 0.20180247724056244,
+      "learning_rate": 8e-05,
+      "loss": 1.4691,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5265456530640098,
+      "grad_norm": 0.21098428964614868,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5266821345707656,
+      "grad_norm": 0.20508822798728943,
+      "learning_rate": 8e-05,
+      "loss": 1.462,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5268186160775215,
+      "grad_norm": 0.221853107213974,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5269550975842773,
+      "grad_norm": 0.2085074931383133,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5270915790910332,
+      "grad_norm": 0.210953027009964,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 3862
+    },
+    {
+      "epoch": 0.527228060597789,
+      "grad_norm": 0.20656022429466248,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5273645421045449,
+      "grad_norm": 0.22291792929172516,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5275010236113007,
+      "grad_norm": 0.20535455644130707,
+      "learning_rate": 8e-05,
+      "loss": 1.4194,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5276375051180565,
+      "grad_norm": 0.2144753783941269,
+      "learning_rate": 8e-05,
+      "loss": 1.4931,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5277739866248123,
+      "grad_norm": 0.21257278323173523,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5279104681315682,
+      "grad_norm": 0.2092253714799881,
+      "learning_rate": 8e-05,
+      "loss": 1.5194,
+      "step": 3868
+    },
+    {
+      "epoch": 0.528046949638324,
+      "grad_norm": 0.21771551668643951,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5281834311450798,
+      "grad_norm": 0.2243928760290146,
+      "learning_rate": 8e-05,
+      "loss": 1.5146,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5283199126518356,
+      "grad_norm": 0.21476124227046967,
+      "learning_rate": 8e-05,
+      "loss": 1.4115,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5284563941585915,
+      "grad_norm": 0.22355817258358002,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5285928756653473,
+      "grad_norm": 0.20510250329971313,
+      "learning_rate": 8e-05,
+      "loss": 1.4873,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5287293571721032,
+      "grad_norm": 0.21323156356811523,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5288658386788591,
+      "grad_norm": 0.22806447744369507,
+      "learning_rate": 8e-05,
+      "loss": 1.5237,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5290023201856149,
+      "grad_norm": 0.21466396749019623,
+      "learning_rate": 8e-05,
+      "loss": 1.4813,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5291388016923707,
+      "grad_norm": 0.21021884679794312,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5292752831991265,
+      "grad_norm": 0.217983677983284,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 0.21329325437545776,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5295482462126382,
+      "grad_norm": 0.20736046135425568,
+      "learning_rate": 8e-05,
+      "loss": 1.4447,
+      "step": 3880
+    },
+    {
+      "epoch": 0.529684727719394,
+      "grad_norm": 0.21129010617733002,
+      "learning_rate": 8e-05,
+      "loss": 1.5435,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5298212092261498,
+      "grad_norm": 0.21744251251220703,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5299576907329057,
+      "grad_norm": 0.19702544808387756,
+      "learning_rate": 8e-05,
+      "loss": 1.4878,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5300941722396615,
+      "grad_norm": 0.20643533766269684,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5302306537464173,
+      "grad_norm": 0.20823663473129272,
+      "learning_rate": 8e-05,
+      "loss": 1.4465,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5303671352531732,
+      "grad_norm": 0.21344545483589172,
+      "learning_rate": 8e-05,
+      "loss": 1.4446,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5305036167599291,
+      "grad_norm": 0.21572035551071167,
+      "learning_rate": 8e-05,
+      "loss": 1.4988,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5306400982666849,
+      "grad_norm": 0.21610423922538757,
+      "learning_rate": 8e-05,
+      "loss": 1.486,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5307765797734407,
+      "grad_norm": 0.2159118950366974,
+      "learning_rate": 8e-05,
+      "loss": 1.4806,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5309130612801966,
+      "grad_norm": 0.2053048312664032,
+      "learning_rate": 8e-05,
+      "loss": 1.4155,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5310495427869524,
+      "grad_norm": 0.2087618112564087,
+      "learning_rate": 8e-05,
+      "loss": 1.4348,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5311860242937082,
+      "grad_norm": 0.21650056540966034,
+      "learning_rate": 8e-05,
+      "loss": 1.488,
+      "step": 3892
+    },
+    {
+      "epoch": 0.531322505800464,
+      "grad_norm": 0.20754197239875793,
+      "learning_rate": 8e-05,
+      "loss": 1.4495,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5314589873072199,
+      "grad_norm": 0.20827847719192505,
+      "learning_rate": 8e-05,
+      "loss": 1.4734,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5315954688139757,
+      "grad_norm": 0.20556117594242096,
+      "learning_rate": 8e-05,
+      "loss": 1.4977,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5317319503207315,
+      "grad_norm": 0.22676751017570496,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5318684318274873,
+      "grad_norm": 0.21132470667362213,
+      "learning_rate": 8e-05,
+      "loss": 1.5216,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5320049133342432,
+      "grad_norm": 0.2211170196533203,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5321413948409991,
+      "grad_norm": 0.21029207110404968,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5322778763477549,
+      "grad_norm": 0.20929358899593353,
+      "learning_rate": 8e-05,
+      "loss": 1.4556,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5324143578545107,
+      "grad_norm": 0.20441946387290955,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5325508393612666,
+      "grad_norm": 0.21051496267318726,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5326873208680224,
+      "grad_norm": 0.20627525448799133,
+      "learning_rate": 8e-05,
+      "loss": 1.4385,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5328238023747782,
+      "grad_norm": 0.20343826711177826,
+      "learning_rate": 8e-05,
+      "loss": 1.4661,
+      "step": 3904
+    },
+    {
+      "epoch": 0.532960283881534,
+      "grad_norm": 0.21207121014595032,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5330967653882899,
+      "grad_norm": 0.20420005917549133,
+      "learning_rate": 8e-05,
+      "loss": 1.4499,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5332332468950457,
+      "grad_norm": 0.22073200345039368,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5333697284018015,
+      "grad_norm": 0.21736709773540497,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5335062099085573,
+      "grad_norm": 0.2219231128692627,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5336426914153132,
+      "grad_norm": 0.20779363811016083,
+      "learning_rate": 8e-05,
+      "loss": 1.5031,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5337791729220691,
+      "grad_norm": 0.20677430927753448,
+      "learning_rate": 8e-05,
+      "loss": 1.4398,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5339156544288249,
+      "grad_norm": 0.20664069056510925,
+      "learning_rate": 8e-05,
+      "loss": 1.4434,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5340521359355808,
+      "grad_norm": 0.20454922318458557,
+      "learning_rate": 8e-05,
+      "loss": 1.4163,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5341886174423366,
+      "grad_norm": 0.2075737863779068,
+      "learning_rate": 8e-05,
+      "loss": 1.4388,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5343250989490924,
+      "grad_norm": 0.22747676074504852,
+      "learning_rate": 8e-05,
+      "loss": 1.5373,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5344615804558482,
+      "grad_norm": 0.20168982446193695,
+      "learning_rate": 8e-05,
+      "loss": 1.4578,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5345980619626041,
+      "grad_norm": 0.2137005627155304,
+      "learning_rate": 8e-05,
+      "loss": 1.5625,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5347345434693599,
+      "grad_norm": 0.20854716002941132,
+      "learning_rate": 8e-05,
+      "loss": 1.5014,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5348710249761157,
+      "grad_norm": 0.21440574526786804,
+      "learning_rate": 8e-05,
+      "loss": 1.4709,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5350075064828715,
+      "grad_norm": 0.21601231396198273,
+      "learning_rate": 8e-05,
+      "loss": 1.4124,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5351439879896274,
+      "grad_norm": 0.21279220283031464,
+      "learning_rate": 8e-05,
+      "loss": 1.49,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5352804694963832,
+      "grad_norm": 0.21807019412517548,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5354169510031391,
+      "grad_norm": 0.21771401166915894,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 3923
+    },
+    {
+      "epoch": 0.535553432509895,
+      "grad_norm": 0.22003212571144104,
+      "learning_rate": 8e-05,
+      "loss": 1.5016,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5356899140166508,
+      "grad_norm": 0.209725484251976,
+      "learning_rate": 8e-05,
+      "loss": 1.385,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5358263955234066,
+      "grad_norm": 0.21322984993457794,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5359628770301624,
+      "grad_norm": 0.21876545250415802,
+      "learning_rate": 8e-05,
+      "loss": 1.4844,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5360993585369183,
+      "grad_norm": 0.21539653837680817,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5362358400436741,
+      "grad_norm": 0.21845944225788116,
+      "learning_rate": 8e-05,
+      "loss": 1.4573,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5363723215504299,
+      "grad_norm": 0.21582849323749542,
+      "learning_rate": 8e-05,
+      "loss": 1.4575,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5365088030571857,
+      "grad_norm": 0.21010752022266388,
+      "learning_rate": 8e-05,
+      "loss": 1.519,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5366452845639416,
+      "grad_norm": 0.20294950902462006,
+      "learning_rate": 8e-05,
+      "loss": 1.4673,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5367817660706974,
+      "grad_norm": 0.21116815507411957,
+      "learning_rate": 8e-05,
+      "loss": 1.4432,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5369182475774532,
+      "grad_norm": 0.2148934304714203,
+      "learning_rate": 8e-05,
+      "loss": 1.468,
+      "step": 3934
+    },
+    {
+      "epoch": 0.537054729084209,
+      "grad_norm": 0.21122683584690094,
+      "learning_rate": 8e-05,
+      "loss": 1.5065,
+      "step": 3935
+    },
+    {
+      "epoch": 0.537191210590965,
+      "grad_norm": 0.2137981504201889,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5373276920977208,
+      "grad_norm": 0.2208615243434906,
+      "learning_rate": 8e-05,
+      "loss": 1.5134,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5374641736044766,
+      "grad_norm": 0.21685846149921417,
+      "learning_rate": 8e-05,
+      "loss": 1.5534,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5376006551112325,
+      "grad_norm": 0.21427197754383087,
+      "learning_rate": 8e-05,
+      "loss": 1.5425,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5377371366179883,
+      "grad_norm": 0.2091689109802246,
+      "learning_rate": 8e-05,
+      "loss": 1.4375,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5378736181247441,
+      "grad_norm": 0.20809611678123474,
+      "learning_rate": 8e-05,
+      "loss": 1.4603,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5380100996314999,
+      "grad_norm": 0.20635966956615448,
+      "learning_rate": 8e-05,
+      "loss": 1.5201,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5381465811382558,
+      "grad_norm": 0.2120577096939087,
+      "learning_rate": 8e-05,
+      "loss": 1.4263,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5382830626450116,
+      "grad_norm": 0.20983166992664337,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5384195441517674,
+      "grad_norm": 0.22868888080120087,
+      "learning_rate": 8e-05,
+      "loss": 1.5333,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5385560256585232,
+      "grad_norm": 0.22498992085456848,
+      "learning_rate": 8e-05,
+      "loss": 1.432,
+      "step": 3946
+    },
+    {
+      "epoch": 0.538692507165279,
+      "grad_norm": 0.20812679827213287,
+      "learning_rate": 8e-05,
+      "loss": 1.4087,
+      "step": 3947
+    },
+    {
+      "epoch": 0.538828988672035,
+      "grad_norm": 0.2148916870355606,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5389654701787908,
+      "grad_norm": 0.22119343280792236,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5391019516855466,
+      "grad_norm": 0.20873288810253143,
+      "learning_rate": 8e-05,
+      "loss": 1.421,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5392384331923025,
+      "grad_norm": 0.20993028581142426,
+      "learning_rate": 8e-05,
+      "loss": 1.4836,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5393749146990583,
+      "grad_norm": 0.21094287931919098,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5395113962058141,
+      "grad_norm": 0.21700243651866913,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 3953
+    },
+    {
+      "epoch": 0.53964787771257,
+      "grad_norm": 0.21250613033771515,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5397843592193258,
+      "grad_norm": 0.23623088002204895,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5399208407260816,
+      "grad_norm": 0.21170058846473694,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 3956
+    },
+    {
+      "epoch": 0.5400573222328374,
+      "grad_norm": 0.21862830221652985,
+      "learning_rate": 8e-05,
+      "loss": 1.5265,
+      "step": 3957
+    },
+    {
+      "epoch": 0.5401938037395932,
+      "grad_norm": 0.20765745639801025,
+      "learning_rate": 8e-05,
+      "loss": 1.431,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5403302852463491,
+      "grad_norm": 0.21233604848384857,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5404667667531049,
+      "grad_norm": 0.208890900015831,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5406032482598608,
+      "grad_norm": 0.20955172181129456,
+      "learning_rate": 8e-05,
+      "loss": 1.4627,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5407397297666167,
+      "grad_norm": 0.22223545610904694,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5408762112733725,
+      "grad_norm": 0.20734697580337524,
+      "learning_rate": 8e-05,
+      "loss": 1.4233,
+      "step": 3963
+    },
+    {
+      "epoch": 0.5410126927801283,
+      "grad_norm": 0.21651016175746918,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 3964
+    },
+    {
+      "epoch": 0.5411491742868841,
+      "grad_norm": 0.21300671994686127,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 3965
+    },
+    {
+      "epoch": 0.54128565579364,
+      "grad_norm": 0.21661721169948578,
+      "learning_rate": 8e-05,
+      "loss": 1.4919,
+      "step": 3966
+    },
+    {
+      "epoch": 0.5414221373003958,
+      "grad_norm": 0.215090811252594,
+      "learning_rate": 8e-05,
+      "loss": 1.5335,
+      "step": 3967
+    },
+    {
+      "epoch": 0.5415586188071516,
+      "grad_norm": 0.2087790071964264,
+      "learning_rate": 8e-05,
+      "loss": 1.4603,
+      "step": 3968
+    },
+    {
+      "epoch": 0.5416951003139074,
+      "grad_norm": 0.20841914415359497,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 3969
+    },
+    {
+      "epoch": 0.5418315818206633,
+      "grad_norm": 0.2188490480184555,
+      "learning_rate": 8e-05,
+      "loss": 1.4941,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5419680633274191,
+      "grad_norm": 0.20759616792201996,
+      "learning_rate": 8e-05,
+      "loss": 1.4166,
+      "step": 3971
+    },
+    {
+      "epoch": 0.5421045448341749,
+      "grad_norm": 0.21219879388809204,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 3972
+    },
+    {
+      "epoch": 0.5422410263409309,
+      "grad_norm": 0.20806318521499634,
+      "learning_rate": 8e-05,
+      "loss": 1.4983,
+      "step": 3973
+    },
+    {
+      "epoch": 0.5423775078476867,
+      "grad_norm": 0.21870160102844238,
+      "learning_rate": 8e-05,
+      "loss": 1.4941,
+      "step": 3974
+    },
+    {
+      "epoch": 0.5425139893544425,
+      "grad_norm": 0.21698282659053802,
+      "learning_rate": 8e-05,
+      "loss": 1.5125,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5426504708611983,
+      "grad_norm": 0.21326006948947906,
+      "learning_rate": 8e-05,
+      "loss": 1.4657,
+      "step": 3976
+    },
+    {
+      "epoch": 0.5427869523679542,
+      "grad_norm": 0.2060515582561493,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 3977
+    },
+    {
+      "epoch": 0.54292343387471,
+      "grad_norm": 0.20540425181388855,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 3978
+    },
+    {
+      "epoch": 0.5430599153814658,
+      "grad_norm": 0.2081085592508316,
+      "learning_rate": 8e-05,
+      "loss": 1.4229,
+      "step": 3979
+    },
+    {
+      "epoch": 0.5431963968882216,
+      "grad_norm": 0.21043510735034943,
+      "learning_rate": 8e-05,
+      "loss": 1.468,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5433328783949775,
+      "grad_norm": 0.21225225925445557,
+      "learning_rate": 8e-05,
+      "loss": 1.5107,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5434693599017333,
+      "grad_norm": 0.20884309709072113,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 3982
+    },
+    {
+      "epoch": 0.5436058414084891,
+      "grad_norm": 0.2209782898426056,
+      "learning_rate": 8e-05,
+      "loss": 1.5313,
+      "step": 3983
+    },
+    {
+      "epoch": 0.5437423229152449,
+      "grad_norm": 0.20940878987312317,
+      "learning_rate": 8e-05,
+      "loss": 1.4542,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5438788044220009,
+      "grad_norm": 0.20551931858062744,
+      "learning_rate": 8e-05,
+      "loss": 1.5271,
+      "step": 3985
+    },
+    {
+      "epoch": 0.5440152859287567,
+      "grad_norm": 0.21722780168056488,
+      "learning_rate": 8e-05,
+      "loss": 1.5348,
+      "step": 3986
+    },
+    {
+      "epoch": 0.5441517674355125,
+      "grad_norm": 0.21256360411643982,
+      "learning_rate": 8e-05,
+      "loss": 1.5125,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5442882489422683,
+      "grad_norm": 0.21507519483566284,
+      "learning_rate": 8e-05,
+      "loss": 1.4884,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5444247304490242,
+      "grad_norm": 0.20599912106990814,
+      "learning_rate": 8e-05,
+      "loss": 1.4976,
+      "step": 3989
+    },
+    {
+      "epoch": 0.54456121195578,
+      "grad_norm": 0.20583081245422363,
+      "learning_rate": 8e-05,
+      "loss": 1.4143,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5446976934625358,
+      "grad_norm": 0.22008644044399261,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5448341749692917,
+      "grad_norm": 0.20400090515613556,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 3992
+    },
+    {
+      "epoch": 0.5449706564760475,
+      "grad_norm": 0.2150939404964447,
+      "learning_rate": 8e-05,
+      "loss": 1.5551,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5451071379828033,
+      "grad_norm": 0.21547016501426697,
+      "learning_rate": 8e-05,
+      "loss": 1.5237,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5452436194895591,
+      "grad_norm": 0.2080121487379074,
+      "learning_rate": 8e-05,
+      "loss": 1.4607,
+      "step": 3995
+    },
+    {
+      "epoch": 0.545380100996315,
+      "grad_norm": 0.2108195722103119,
+      "learning_rate": 8e-05,
+      "loss": 1.4577,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5455165825030708,
+      "grad_norm": 0.21768449246883392,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5456530640098267,
+      "grad_norm": 0.21431384980678558,
+      "learning_rate": 8e-05,
+      "loss": 1.5173,
+      "step": 3998
+    },
+    {
+      "epoch": 0.5457895455165825,
+      "grad_norm": 0.2129233330488205,
+      "learning_rate": 8e-05,
+      "loss": 1.463,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5459260270233384,
+      "grad_norm": 0.2187803089618683,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5460625085300942,
+      "grad_norm": 0.21778717637062073,
+      "learning_rate": 8e-05,
+      "loss": 1.5528,
+      "step": 4001
+    },
+    {
+      "epoch": 0.54619899003685,
+      "grad_norm": 0.2025071531534195,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5463354715436058,
+      "grad_norm": 0.21569125354290009,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 4003
+    },
+    {
+      "epoch": 0.5464719530503617,
+      "grad_norm": 0.21426476538181305,
+      "learning_rate": 8e-05,
+      "loss": 1.5413,
+      "step": 4004
+    },
+    {
+      "epoch": 0.5466084345571175,
+      "grad_norm": 0.21295900642871857,
+      "learning_rate": 8e-05,
+      "loss": 1.4776,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5467449160638733,
+      "grad_norm": 0.2086125612258911,
+      "learning_rate": 8e-05,
+      "loss": 1.5145,
+      "step": 4006
+    },
+    {
+      "epoch": 0.5468813975706291,
+      "grad_norm": 0.21766814589500427,
+      "learning_rate": 8e-05,
+      "loss": 1.5284,
+      "step": 4007
+    },
+    {
+      "epoch": 0.547017879077385,
+      "grad_norm": 0.21663233637809753,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5471543605841408,
+      "grad_norm": 0.21130946278572083,
+      "learning_rate": 8e-05,
+      "loss": 1.4747,
+      "step": 4009
+    },
+    {
+      "epoch": 0.5472908420908967,
+      "grad_norm": 0.20893415808677673,
+      "learning_rate": 8e-05,
+      "loss": 1.4409,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5474273235976526,
+      "grad_norm": 0.21529194712638855,
+      "learning_rate": 8e-05,
+      "loss": 1.4994,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5475638051044084,
+      "grad_norm": 0.2148546427488327,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 4012
+    },
+    {
+      "epoch": 0.5477002866111642,
+      "grad_norm": 0.21923306584358215,
+      "learning_rate": 8e-05,
+      "loss": 1.5096,
+      "step": 4013
+    },
+    {
+      "epoch": 0.54783676811792,
+      "grad_norm": 0.21592554450035095,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 4014
+    },
+    {
+      "epoch": 0.5479732496246759,
+      "grad_norm": 0.2120472490787506,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 4015
+    },
+    {
+      "epoch": 0.5481097311314317,
+      "grad_norm": 0.21756690740585327,
+      "learning_rate": 8e-05,
+      "loss": 1.4899,
+      "step": 4016
+    },
+    {
+      "epoch": 0.5482462126381875,
+      "grad_norm": 0.21122270822525024,
+      "learning_rate": 8e-05,
+      "loss": 1.5522,
+      "step": 4017
+    },
+    {
+      "epoch": 0.5483826941449433,
+      "grad_norm": 0.2096116691827774,
+      "learning_rate": 8e-05,
+      "loss": 1.5042,
+      "step": 4018
+    },
+    {
+      "epoch": 0.5485191756516992,
+      "grad_norm": 0.212167888879776,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 4019
+    },
+    {
+      "epoch": 0.548655657158455,
+      "grad_norm": 0.2202005535364151,
+      "learning_rate": 8e-05,
+      "loss": 1.5873,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5487921386652108,
+      "grad_norm": 0.20982125401496887,
+      "learning_rate": 8e-05,
+      "loss": 1.4891,
+      "step": 4021
+    },
+    {
+      "epoch": 0.5489286201719668,
+      "grad_norm": 0.21076913177967072,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 4022
+    },
+    {
+      "epoch": 0.5490651016787226,
+      "grad_norm": 0.21276473999023438,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 4023
+    },
+    {
+      "epoch": 0.5492015831854784,
+      "grad_norm": 0.2097943127155304,
+      "learning_rate": 8e-05,
+      "loss": 1.5022,
+      "step": 4024
+    },
+    {
+      "epoch": 0.5493380646922342,
+      "grad_norm": 0.2082330286502838,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 4025
+    },
+    {
+      "epoch": 0.54947454619899,
+      "grad_norm": 0.21301335096359253,
+      "learning_rate": 8e-05,
+      "loss": 1.5428,
+      "step": 4026
+    },
+    {
+      "epoch": 0.5496110277057459,
+      "grad_norm": 0.21643485128879547,
+      "learning_rate": 8e-05,
+      "loss": 1.4966,
+      "step": 4027
+    },
+    {
+      "epoch": 0.5497475092125017,
+      "grad_norm": 0.20908930897712708,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 4028
+    },
+    {
+      "epoch": 0.5498839907192575,
+      "grad_norm": 0.22110337018966675,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 4029
+    },
+    {
+      "epoch": 0.5500204722260134,
+      "grad_norm": 0.21730099618434906,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5501569537327692,
+      "grad_norm": 0.21115638315677643,
+      "learning_rate": 8e-05,
+      "loss": 1.4751,
+      "step": 4031
+    },
+    {
+      "epoch": 0.550293435239525,
+      "grad_norm": 0.21456487476825714,
+      "learning_rate": 8e-05,
+      "loss": 1.4253,
+      "step": 4032
+    },
+    {
+      "epoch": 0.5504299167462808,
+      "grad_norm": 0.21453270316123962,
+      "learning_rate": 8e-05,
+      "loss": 1.5473,
+      "step": 4033
+    },
+    {
+      "epoch": 0.5505663982530367,
+      "grad_norm": 0.20871831476688385,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 4034
+    },
+    {
+      "epoch": 0.5507028797597926,
+      "grad_norm": 0.21607492864131927,
+      "learning_rate": 8e-05,
+      "loss": 1.5027,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5508393612665484,
+      "grad_norm": 0.22087296843528748,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 4036
+    },
+    {
+      "epoch": 0.5509758427733042,
+      "grad_norm": 0.22039352357387543,
+      "learning_rate": 8e-05,
+      "loss": 1.5249,
+      "step": 4037
+    },
+    {
+      "epoch": 0.5511123242800601,
+      "grad_norm": 0.20952408015727997,
+      "learning_rate": 8e-05,
+      "loss": 1.5491,
+      "step": 4038
+    },
+    {
+      "epoch": 0.5512488057868159,
+      "grad_norm": 0.21672017872333527,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 4039
+    },
+    {
+      "epoch": 0.5513852872935717,
+      "grad_norm": 0.2058047205209732,
+      "learning_rate": 8e-05,
+      "loss": 1.402,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5515217688003275,
+      "grad_norm": 0.21104593575000763,
+      "learning_rate": 8e-05,
+      "loss": 1.4816,
+      "step": 4041
+    },
+    {
+      "epoch": 0.5516582503070834,
+      "grad_norm": 0.22632263600826263,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 4042
+    },
+    {
+      "epoch": 0.5517947318138392,
+      "grad_norm": 0.2167060524225235,
+      "learning_rate": 8e-05,
+      "loss": 1.5209,
+      "step": 4043
+    },
+    {
+      "epoch": 0.551931213320595,
+      "grad_norm": 0.2084483802318573,
+      "learning_rate": 8e-05,
+      "loss": 1.4777,
+      "step": 4044
+    },
+    {
+      "epoch": 0.5520676948273509,
+      "grad_norm": 0.20652832090854645,
+      "learning_rate": 8e-05,
+      "loss": 1.4419,
+      "step": 4045
+    },
+    {
+      "epoch": 0.5522041763341067,
+      "grad_norm": 0.20899629592895508,
+      "learning_rate": 8e-05,
+      "loss": 1.512,
+      "step": 4046
+    },
+    {
+      "epoch": 0.5523406578408626,
+      "grad_norm": 0.22118784487247467,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 4047
+    },
+    {
+      "epoch": 0.5524771393476184,
+      "grad_norm": 0.20944646000862122,
+      "learning_rate": 8e-05,
+      "loss": 1.4947,
+      "step": 4048
+    },
+    {
+      "epoch": 0.5526136208543743,
+      "grad_norm": 0.2094104290008545,
+      "learning_rate": 8e-05,
+      "loss": 1.4742,
+      "step": 4049
+    },
+    {
+      "epoch": 0.5527501023611301,
+      "grad_norm": 0.21074897050857544,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5528865838678859,
+      "grad_norm": 0.21322181820869446,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 4051
+    },
+    {
+      "epoch": 0.5530230653746417,
+      "grad_norm": 0.2102811634540558,
+      "learning_rate": 8e-05,
+      "loss": 1.4214,
+      "step": 4052
+    },
+    {
+      "epoch": 0.5531595468813976,
+      "grad_norm": 0.20323963463306427,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 4053
+    },
+    {
+      "epoch": 0.5532960283881534,
+      "grad_norm": 0.22281189262866974,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 4054
+    },
+    {
+      "epoch": 0.5534325098949092,
+      "grad_norm": 0.21449768543243408,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 4055
+    },
+    {
+      "epoch": 0.553568991401665,
+      "grad_norm": 0.22041170299053192,
+      "learning_rate": 8e-05,
+      "loss": 1.4704,
+      "step": 4056
+    },
+    {
+      "epoch": 0.5537054729084209,
+      "grad_norm": 0.22577407956123352,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 4057
+    },
+    {
+      "epoch": 0.5538419544151767,
+      "grad_norm": 0.20272588729858398,
+      "learning_rate": 8e-05,
+      "loss": 1.4564,
+      "step": 4058
+    },
+    {
+      "epoch": 0.5539784359219326,
+      "grad_norm": 0.2118610143661499,
+      "learning_rate": 8e-05,
+      "loss": 1.4368,
+      "step": 4059
+    },
+    {
+      "epoch": 0.5541149174286885,
+      "grad_norm": 0.213422030210495,
+      "learning_rate": 8e-05,
+      "loss": 1.4711,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5542513989354443,
+      "grad_norm": 0.2121119350194931,
+      "learning_rate": 8e-05,
+      "loss": 1.4682,
+      "step": 4061
+    },
+    {
+      "epoch": 0.5543878804422001,
+      "grad_norm": 0.21609561145305634,
+      "learning_rate": 8e-05,
+      "loss": 1.5048,
+      "step": 4062
+    },
+    {
+      "epoch": 0.5545243619489559,
+      "grad_norm": 0.21246649324893951,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 4063
+    },
+    {
+      "epoch": 0.5546608434557118,
+      "grad_norm": 0.21132156252861023,
+      "learning_rate": 8e-05,
+      "loss": 1.4764,
+      "step": 4064
+    },
+    {
+      "epoch": 0.5547973249624676,
+      "grad_norm": 0.21442018449306488,
+      "learning_rate": 8e-05,
+      "loss": 1.5307,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5549338064692234,
+      "grad_norm": 0.21475937962532043,
+      "learning_rate": 8e-05,
+      "loss": 1.47,
+      "step": 4066
+    },
+    {
+      "epoch": 0.5550702879759792,
+      "grad_norm": 0.20682257413864136,
+      "learning_rate": 8e-05,
+      "loss": 1.3839,
+      "step": 4067
+    },
+    {
+      "epoch": 0.5552067694827351,
+      "grad_norm": 0.21496626734733582,
+      "learning_rate": 8e-05,
+      "loss": 1.5106,
+      "step": 4068
+    },
+    {
+      "epoch": 0.5553432509894909,
+      "grad_norm": 0.22642920911312103,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 4069
+    },
+    {
+      "epoch": 0.5554797324962467,
+      "grad_norm": 0.21580594778060913,
+      "learning_rate": 8e-05,
+      "loss": 1.4999,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5556162140030025,
+      "grad_norm": 0.2172602415084839,
+      "learning_rate": 8e-05,
+      "loss": 1.4768,
+      "step": 4071
+    },
+    {
+      "epoch": 0.5557526955097585,
+      "grad_norm": 0.20682387053966522,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 4072
+    },
+    {
+      "epoch": 0.5558891770165143,
+      "grad_norm": 0.21785330772399902,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 4073
+    },
+    {
+      "epoch": 0.5560256585232701,
+      "grad_norm": 0.21788525581359863,
+      "learning_rate": 8e-05,
+      "loss": 1.4512,
+      "step": 4074
+    },
+    {
+      "epoch": 0.556162140030026,
+      "grad_norm": 0.23095788061618805,
+      "learning_rate": 8e-05,
+      "loss": 1.4586,
+      "step": 4075
+    },
+    {
+      "epoch": 0.5562986215367818,
+      "grad_norm": 0.2184412032365799,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 4076
+    },
+    {
+      "epoch": 0.5564351030435376,
+      "grad_norm": 0.21445901691913605,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 4077
+    },
+    {
+      "epoch": 0.5565715845502934,
+      "grad_norm": 0.22111648321151733,
+      "learning_rate": 8e-05,
+      "loss": 1.4721,
+      "step": 4078
+    },
+    {
+      "epoch": 0.5567080660570493,
+      "grad_norm": 0.22030308842658997,
+      "learning_rate": 8e-05,
+      "loss": 1.4827,
+      "step": 4079
+    },
+    {
+      "epoch": 0.5568445475638051,
+      "grad_norm": 0.21483373641967773,
+      "learning_rate": 8e-05,
+      "loss": 1.5023,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5569810290705609,
+      "grad_norm": 0.20193031430244446,
+      "learning_rate": 8e-05,
+      "loss": 1.4216,
+      "step": 4081
+    },
+    {
+      "epoch": 0.5571175105773167,
+      "grad_norm": 0.2037355899810791,
+      "learning_rate": 8e-05,
+      "loss": 1.4589,
+      "step": 4082
+    },
+    {
+      "epoch": 0.5572539920840726,
+      "grad_norm": 0.2091190218925476,
+      "learning_rate": 8e-05,
+      "loss": 1.4874,
+      "step": 4083
+    },
+    {
+      "epoch": 0.5573904735908285,
+      "grad_norm": 0.2225044220685959,
+      "learning_rate": 8e-05,
+      "loss": 1.4981,
+      "step": 4084
+    },
+    {
+      "epoch": 0.5575269550975843,
+      "grad_norm": 0.2126842588186264,
+      "learning_rate": 8e-05,
+      "loss": 1.3953,
+      "step": 4085
+    },
+    {
+      "epoch": 0.5576634366043401,
+      "grad_norm": 0.2057109922170639,
+      "learning_rate": 8e-05,
+      "loss": 1.3893,
+      "step": 4086
+    },
+    {
+      "epoch": 0.557799918111096,
+      "grad_norm": 0.21497681736946106,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 4087
+    },
+    {
+      "epoch": 0.5579363996178518,
+      "grad_norm": 0.2119443118572235,
+      "learning_rate": 8e-05,
+      "loss": 1.4575,
+      "step": 4088
+    },
+    {
+      "epoch": 0.5580728811246076,
+      "grad_norm": 0.2088376134634018,
+      "learning_rate": 8e-05,
+      "loss": 1.4457,
+      "step": 4089
+    },
+    {
+      "epoch": 0.5582093626313634,
+      "grad_norm": 0.2167643904685974,
+      "learning_rate": 8e-05,
+      "loss": 1.5152,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5583458441381193,
+      "grad_norm": 0.22091422975063324,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 4091
+    },
+    {
+      "epoch": 0.5584823256448751,
+      "grad_norm": 0.20456083118915558,
+      "learning_rate": 8e-05,
+      "loss": 1.3951,
+      "step": 4092
+    },
+    {
+      "epoch": 0.5586188071516309,
+      "grad_norm": 0.2048211395740509,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 4093
+    },
+    {
+      "epoch": 0.5587552886583868,
+      "grad_norm": 0.21812017261981964,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 4094
+    },
+    {
+      "epoch": 0.5588917701651426,
+      "grad_norm": 0.21348117291927338,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 4095
+    },
+    {
+      "epoch": 0.5590282516718985,
+      "grad_norm": 0.21337182819843292,
+      "learning_rate": 8e-05,
+      "loss": 1.5199,
+      "step": 4096
+    },
+    {
+      "epoch": 0.5591647331786543,
+      "grad_norm": 0.2136903554201126,
+      "learning_rate": 8e-05,
+      "loss": 1.4224,
+      "step": 4097
+    },
+    {
+      "epoch": 0.5593012146854102,
+      "grad_norm": 0.2061806470155716,
+      "learning_rate": 8e-05,
+      "loss": 1.443,
+      "step": 4098
+    },
+    {
+      "epoch": 0.559437696192166,
+      "grad_norm": 0.2124415636062622,
+      "learning_rate": 8e-05,
+      "loss": 1.4358,
+      "step": 4099
+    },
+    {
+      "epoch": 0.5595741776989218,
+      "grad_norm": 0.21656042337417603,
+      "learning_rate": 8e-05,
+      "loss": 1.4963,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5597106592056776,
+      "grad_norm": 0.21043917536735535,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 4101
+    },
+    {
+      "epoch": 0.5598471407124335,
+      "grad_norm": 0.21233618259429932,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 4102
+    },
+    {
+      "epoch": 0.5599836222191893,
+      "grad_norm": 0.2204185128211975,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 4103
+    },
+    {
+      "epoch": 0.5601201037259451,
+      "grad_norm": 0.20545360445976257,
+      "learning_rate": 8e-05,
+      "loss": 1.4258,
+      "step": 4104
+    },
+    {
+      "epoch": 0.5602565852327009,
+      "grad_norm": 0.21167825162410736,
+      "learning_rate": 8e-05,
+      "loss": 1.5446,
+      "step": 4105
+    },
+    {
+      "epoch": 0.5603930667394568,
+      "grad_norm": 0.2064371109008789,
+      "learning_rate": 8e-05,
+      "loss": 1.3925,
+      "step": 4106
+    },
+    {
+      "epoch": 0.5605295482462126,
+      "grad_norm": 0.22175613045692444,
+      "learning_rate": 8e-05,
+      "loss": 1.5344,
+      "step": 4107
+    },
+    {
+      "epoch": 0.5606660297529684,
+      "grad_norm": 0.21652434766292572,
+      "learning_rate": 8e-05,
+      "loss": 1.4515,
+      "step": 4108
+    },
+    {
+      "epoch": 0.5608025112597244,
+      "grad_norm": 0.21520711481571198,
+      "learning_rate": 8e-05,
+      "loss": 1.4471,
+      "step": 4109
+    },
+    {
+      "epoch": 0.5609389927664802,
+      "grad_norm": 0.20460045337677002,
+      "learning_rate": 8e-05,
+      "loss": 1.429,
+      "step": 4110
+    },
+    {
+      "epoch": 0.561075474273236,
+      "grad_norm": 0.2147601991891861,
+      "learning_rate": 8e-05,
+      "loss": 1.4549,
+      "step": 4111
+    },
+    {
+      "epoch": 0.5612119557799918,
+      "grad_norm": 0.21812841296195984,
+      "learning_rate": 8e-05,
+      "loss": 1.4382,
+      "step": 4112
+    },
+    {
+      "epoch": 0.5613484372867477,
+      "grad_norm": 0.22209259867668152,
+      "learning_rate": 8e-05,
+      "loss": 1.5323,
+      "step": 4113
+    },
+    {
+      "epoch": 0.5614849187935035,
+      "grad_norm": 0.20632174611091614,
+      "learning_rate": 8e-05,
+      "loss": 1.428,
+      "step": 4114
+    },
+    {
+      "epoch": 0.5616214003002593,
+      "grad_norm": 0.21677370369434357,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 4115
+    },
+    {
+      "epoch": 0.5617578818070151,
+      "grad_norm": 0.21731743216514587,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 4116
+    },
+    {
+      "epoch": 0.561894363313771,
+      "grad_norm": 0.2062983363866806,
+      "learning_rate": 8e-05,
+      "loss": 1.4425,
+      "step": 4117
+    },
+    {
+      "epoch": 0.5620308448205268,
+      "grad_norm": 0.21958425641059875,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 4118
+    },
+    {
+      "epoch": 0.5621673263272826,
+      "grad_norm": 0.2177102416753769,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 4119
+    },
+    {
+      "epoch": 0.5623038078340384,
+      "grad_norm": 0.21377559006214142,
+      "learning_rate": 8e-05,
+      "loss": 1.522,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5624402893407944,
+      "grad_norm": 0.20855441689491272,
+      "learning_rate": 8e-05,
+      "loss": 1.4521,
+      "step": 4121
+    },
+    {
+      "epoch": 0.5625767708475502,
+      "grad_norm": 0.2097448855638504,
+      "learning_rate": 8e-05,
+      "loss": 1.5261,
+      "step": 4122
+    },
+    {
+      "epoch": 0.562713252354306,
+      "grad_norm": 0.21126492321491241,
+      "learning_rate": 8e-05,
+      "loss": 1.5251,
+      "step": 4123
+    },
+    {
+      "epoch": 0.5628497338610619,
+      "grad_norm": 0.21582269668579102,
+      "learning_rate": 8e-05,
+      "loss": 1.4682,
+      "step": 4124
+    },
+    {
+      "epoch": 0.5629862153678177,
+      "grad_norm": 0.21713784337043762,
+      "learning_rate": 8e-05,
+      "loss": 1.4589,
+      "step": 4125
+    },
+    {
+      "epoch": 0.5631226968745735,
+      "grad_norm": 0.21080413460731506,
+      "learning_rate": 8e-05,
+      "loss": 1.5099,
+      "step": 4126
+    },
+    {
+      "epoch": 0.5632591783813293,
+      "grad_norm": 0.20678511261940002,
+      "learning_rate": 8e-05,
+      "loss": 1.4305,
+      "step": 4127
+    },
+    {
+      "epoch": 0.5633956598880852,
+      "grad_norm": 0.21512289345264435,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 4128
+    },
+    {
+      "epoch": 0.563532141394841,
+      "grad_norm": 0.21501654386520386,
+      "learning_rate": 8e-05,
+      "loss": 1.5102,
+      "step": 4129
+    },
+    {
+      "epoch": 0.5636686229015968,
+      "grad_norm": 0.21599584817886353,
+      "learning_rate": 8e-05,
+      "loss": 1.4619,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5638051044083526,
+      "grad_norm": 0.208168163895607,
+      "learning_rate": 8e-05,
+      "loss": 1.4876,
+      "step": 4131
+    },
+    {
+      "epoch": 0.5639415859151085,
+      "grad_norm": 0.2146674245595932,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 4132
+    },
+    {
+      "epoch": 0.5640780674218643,
+      "grad_norm": 0.21401694416999817,
+      "learning_rate": 8e-05,
+      "loss": 1.4308,
+      "step": 4133
+    },
+    {
+      "epoch": 0.5642145489286202,
+      "grad_norm": 0.2089165896177292,
+      "learning_rate": 8e-05,
+      "loss": 1.431,
+      "step": 4134
+    },
+    {
+      "epoch": 0.564351030435376,
+      "grad_norm": 0.20982258021831512,
+      "learning_rate": 8e-05,
+      "loss": 1.498,
+      "step": 4135
+    },
+    {
+      "epoch": 0.5644875119421319,
+      "grad_norm": 0.2157997488975525,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 4136
+    },
+    {
+      "epoch": 0.5646239934488877,
+      "grad_norm": 0.2020702064037323,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 4137
+    },
+    {
+      "epoch": 0.5647604749556435,
+      "grad_norm": 0.21577925980091095,
+      "learning_rate": 8e-05,
+      "loss": 1.509,
+      "step": 4138
+    },
+    {
+      "epoch": 0.5648969564623993,
+      "grad_norm": 0.21908605098724365,
+      "learning_rate": 8e-05,
+      "loss": 1.5158,
+      "step": 4139
+    },
+    {
+      "epoch": 0.5650334379691552,
+      "grad_norm": 0.21253089606761932,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 4140
+    },
+    {
+      "epoch": 0.565169919475911,
+      "grad_norm": 0.222563698887825,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 4141
+    },
+    {
+      "epoch": 0.5653064009826668,
+      "grad_norm": 0.2181646078824997,
+      "learning_rate": 8e-05,
+      "loss": 1.5006,
+      "step": 4142
+    },
+    {
+      "epoch": 0.5654428824894226,
+      "grad_norm": 0.20826943218708038,
+      "learning_rate": 8e-05,
+      "loss": 1.4852,
+      "step": 4143
+    },
+    {
+      "epoch": 0.5655793639961785,
+      "grad_norm": 0.2162502110004425,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 4144
+    },
+    {
+      "epoch": 0.5657158455029343,
+      "grad_norm": 0.20929653942584991,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 4145
+    },
+    {
+      "epoch": 0.5658523270096902,
+      "grad_norm": 0.21493981778621674,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 4146
+    },
+    {
+      "epoch": 0.5659888085164461,
+      "grad_norm": 0.21158720552921295,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 4147
+    },
+    {
+      "epoch": 0.5661252900232019,
+      "grad_norm": 0.20743803679943085,
+      "learning_rate": 8e-05,
+      "loss": 1.4894,
+      "step": 4148
+    },
+    {
+      "epoch": 0.5662617715299577,
+      "grad_norm": 0.22323064506053925,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 4149
+    },
+    {
+      "epoch": 0.5663982530367135,
+      "grad_norm": 0.2157958298921585,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5665347345434694,
+      "grad_norm": 0.2215644121170044,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 4151
+    },
+    {
+      "epoch": 0.5666712160502252,
+      "grad_norm": 0.21179339289665222,
+      "learning_rate": 8e-05,
+      "loss": 1.5174,
+      "step": 4152
+    },
+    {
+      "epoch": 0.566807697556981,
+      "grad_norm": 0.21576069295406342,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 4153
+    },
+    {
+      "epoch": 0.5669441790637368,
+      "grad_norm": 0.21412812173366547,
+      "learning_rate": 8e-05,
+      "loss": 1.5011,
+      "step": 4154
+    },
+    {
+      "epoch": 0.5670806605704927,
+      "grad_norm": 0.21336203813552856,
+      "learning_rate": 8e-05,
+      "loss": 1.5019,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5672171420772485,
+      "grad_norm": 0.20645149052143097,
+      "learning_rate": 8e-05,
+      "loss": 1.399,
+      "step": 4156
+    },
+    {
+      "epoch": 0.5673536235840043,
+      "grad_norm": 0.21714763343334198,
+      "learning_rate": 8e-05,
+      "loss": 1.4449,
+      "step": 4157
+    },
+    {
+      "epoch": 0.5674901050907603,
+      "grad_norm": 0.21368075907230377,
+      "learning_rate": 8e-05,
+      "loss": 1.4458,
+      "step": 4158
+    },
+    {
+      "epoch": 0.5676265865975161,
+      "grad_norm": 0.20892494916915894,
+      "learning_rate": 8e-05,
+      "loss": 1.4508,
+      "step": 4159
+    },
+    {
+      "epoch": 0.5677630681042719,
+      "grad_norm": 0.21361811459064484,
+      "learning_rate": 8e-05,
+      "loss": 1.4057,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5678995496110277,
+      "grad_norm": 0.2225876748561859,
+      "learning_rate": 8e-05,
+      "loss": 1.4793,
+      "step": 4161
+    },
+    {
+      "epoch": 0.5680360311177836,
+      "grad_norm": 0.21273940801620483,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 4162
+    },
+    {
+      "epoch": 0.5681725126245394,
+      "grad_norm": 0.21815603971481323,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 4163
+    },
+    {
+      "epoch": 0.5683089941312952,
+      "grad_norm": 0.2130299061536789,
+      "learning_rate": 8e-05,
+      "loss": 1.4188,
+      "step": 4164
+    },
+    {
+      "epoch": 0.568445475638051,
+      "grad_norm": 0.21651612222194672,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 4165
+    },
+    {
+      "epoch": 0.5685819571448069,
+      "grad_norm": 0.21552126109600067,
+      "learning_rate": 8e-05,
+      "loss": 1.4394,
+      "step": 4166
+    },
+    {
+      "epoch": 0.5687184386515627,
+      "grad_norm": 0.20757144689559937,
+      "learning_rate": 8e-05,
+      "loss": 1.4546,
+      "step": 4167
+    },
+    {
+      "epoch": 0.5688549201583185,
+      "grad_norm": 0.22573474049568176,
+      "learning_rate": 8e-05,
+      "loss": 1.5136,
+      "step": 4168
+    },
+    {
+      "epoch": 0.5689914016650743,
+      "grad_norm": 0.20971375703811646,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 4169
+    },
+    {
+      "epoch": 0.5691278831718302,
+      "grad_norm": 0.21533392369747162,
+      "learning_rate": 8e-05,
+      "loss": 1.4536,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5692643646785861,
+      "grad_norm": 0.21626019477844238,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 4171
+    },
+    {
+      "epoch": 0.5694008461853419,
+      "grad_norm": 0.20644056797027588,
+      "learning_rate": 8e-05,
+      "loss": 1.4657,
+      "step": 4172
+    },
+    {
+      "epoch": 0.5695373276920978,
+      "grad_norm": 0.21214927732944489,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 4173
+    },
+    {
+      "epoch": 0.5696738091988536,
+      "grad_norm": 0.20711253583431244,
+      "learning_rate": 8e-05,
+      "loss": 1.3842,
+      "step": 4174
+    },
+    {
+      "epoch": 0.5698102907056094,
+      "grad_norm": 0.22512763738632202,
+      "learning_rate": 8e-05,
+      "loss": 1.4976,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5699467722123652,
+      "grad_norm": 0.20673586428165436,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 4176
+    },
+    {
+      "epoch": 0.570083253719121,
+      "grad_norm": 0.21036086976528168,
+      "learning_rate": 8e-05,
+      "loss": 1.4698,
+      "step": 4177
+    },
+    {
+      "epoch": 0.5702197352258769,
+      "grad_norm": 0.20807988941669464,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 4178
+    },
+    {
+      "epoch": 0.5703562167326327,
+      "grad_norm": 0.20627886056900024,
+      "learning_rate": 8e-05,
+      "loss": 1.4361,
+      "step": 4179
+    },
+    {
+      "epoch": 0.5704926982393885,
+      "grad_norm": 0.2068205326795578,
+      "learning_rate": 8e-05,
+      "loss": 1.4131,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5706291797461444,
+      "grad_norm": 0.2120501846075058,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 4181
+    },
+    {
+      "epoch": 0.5707656612529002,
+      "grad_norm": 0.2079847753047943,
+      "learning_rate": 8e-05,
+      "loss": 1.5023,
+      "step": 4182
+    },
+    {
+      "epoch": 0.5709021427596561,
+      "grad_norm": 0.21569515764713287,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 4183
+    },
+    {
+      "epoch": 0.5710386242664119,
+      "grad_norm": 0.21743503212928772,
+      "learning_rate": 8e-05,
+      "loss": 1.4924,
+      "step": 4184
+    },
+    {
+      "epoch": 0.5711751057731678,
+      "grad_norm": 0.21706555783748627,
+      "learning_rate": 8e-05,
+      "loss": 1.4576,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5713115872799236,
+      "grad_norm": 0.20865269005298615,
+      "learning_rate": 8e-05,
+      "loss": 1.4638,
+      "step": 4186
+    },
+    {
+      "epoch": 0.5714480687866794,
+      "grad_norm": 0.21492353081703186,
+      "learning_rate": 8e-05,
+      "loss": 1.5075,
+      "step": 4187
+    },
+    {
+      "epoch": 0.5715845502934352,
+      "grad_norm": 0.21789975464344025,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 4188
+    },
+    {
+      "epoch": 0.5717210318001911,
+      "grad_norm": 0.21637997031211853,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 4189
+    },
+    {
+      "epoch": 0.5718575133069469,
+      "grad_norm": 0.20876067876815796,
+      "learning_rate": 8e-05,
+      "loss": 1.4276,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5719939948137027,
+      "grad_norm": 0.21783484518527985,
+      "learning_rate": 8e-05,
+      "loss": 1.4654,
+      "step": 4191
+    },
+    {
+      "epoch": 0.5721304763204585,
+      "grad_norm": 0.21375548839569092,
+      "learning_rate": 8e-05,
+      "loss": 1.4505,
+      "step": 4192
+    },
+    {
+      "epoch": 0.5722669578272144,
+      "grad_norm": 0.2086593508720398,
+      "learning_rate": 8e-05,
+      "loss": 1.3673,
+      "step": 4193
+    },
+    {
+      "epoch": 0.5724034393339702,
+      "grad_norm": 0.2099689394235611,
+      "learning_rate": 8e-05,
+      "loss": 1.4368,
+      "step": 4194
+    },
+    {
+      "epoch": 0.5725399208407261,
+      "grad_norm": 0.21592223644256592,
+      "learning_rate": 8e-05,
+      "loss": 1.4757,
+      "step": 4195
+    },
+    {
+      "epoch": 0.572676402347482,
+      "grad_norm": 0.21087807416915894,
+      "learning_rate": 8e-05,
+      "loss": 1.456,
+      "step": 4196
+    },
+    {
+      "epoch": 0.5728128838542378,
+      "grad_norm": 0.21318086981773376,
+      "learning_rate": 8e-05,
+      "loss": 1.5611,
+      "step": 4197
+    },
+    {
+      "epoch": 0.5729493653609936,
+      "grad_norm": 0.21953259408473969,
+      "learning_rate": 8e-05,
+      "loss": 1.4666,
+      "step": 4198
+    },
+    {
+      "epoch": 0.5730858468677494,
+      "grad_norm": 0.21577037870883942,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 4199
+    },
+    {
+      "epoch": 0.5732223283745053,
+      "grad_norm": 0.21646249294281006,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5733588098812611,
+      "grad_norm": 0.21039555966854095,
+      "learning_rate": 8e-05,
+      "loss": 1.4803,
+      "step": 4201
+    },
+    {
+      "epoch": 0.5734952913880169,
+      "grad_norm": 0.2150830179452896,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 4202
+    },
+    {
+      "epoch": 0.5736317728947727,
+      "grad_norm": 0.21341609954833984,
+      "learning_rate": 8e-05,
+      "loss": 1.4447,
+      "step": 4203
+    },
+    {
+      "epoch": 0.5737682544015286,
+      "grad_norm": 0.20764513313770294,
+      "learning_rate": 8e-05,
+      "loss": 1.4485,
+      "step": 4204
+    },
+    {
+      "epoch": 0.5739047359082844,
+      "grad_norm": 0.21798859536647797,
+      "learning_rate": 8e-05,
+      "loss": 1.542,
+      "step": 4205
+    },
+    {
+      "epoch": 0.5740412174150402,
+      "grad_norm": 0.21576344966888428,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 4206
+    },
+    {
+      "epoch": 0.574177698921796,
+      "grad_norm": 0.21215452253818512,
+      "learning_rate": 8e-05,
+      "loss": 1.4298,
+      "step": 4207
+    },
+    {
+      "epoch": 0.574314180428552,
+      "grad_norm": 0.2231324464082718,
+      "learning_rate": 8e-05,
+      "loss": 1.4565,
+      "step": 4208
+    },
+    {
+      "epoch": 0.5744506619353078,
+      "grad_norm": 0.22044414281845093,
+      "learning_rate": 8e-05,
+      "loss": 1.4826,
+      "step": 4209
+    },
+    {
+      "epoch": 0.5745871434420636,
+      "grad_norm": 0.22512578964233398,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 4210
+    },
+    {
+      "epoch": 0.5747236249488195,
+      "grad_norm": 0.21472027897834778,
+      "learning_rate": 8e-05,
+      "loss": 1.4495,
+      "step": 4211
+    },
+    {
+      "epoch": 0.5748601064555753,
+      "grad_norm": 0.2077253758907318,
+      "learning_rate": 8e-05,
+      "loss": 1.4634,
+      "step": 4212
+    },
+    {
+      "epoch": 0.5749965879623311,
+      "grad_norm": 0.2077658623456955,
+      "learning_rate": 8e-05,
+      "loss": 1.48,
+      "step": 4213
+    },
+    {
+      "epoch": 0.5751330694690869,
+      "grad_norm": 0.2147304117679596,
+      "learning_rate": 8e-05,
+      "loss": 1.5391,
+      "step": 4214
+    },
+    {
+      "epoch": 0.5752695509758428,
+      "grad_norm": 0.2135041058063507,
+      "learning_rate": 8e-05,
+      "loss": 1.4851,
+      "step": 4215
+    },
+    {
+      "epoch": 0.5754060324825986,
+      "grad_norm": 0.21538491547107697,
+      "learning_rate": 8e-05,
+      "loss": 1.4741,
+      "step": 4216
+    },
+    {
+      "epoch": 0.5755425139893544,
+      "grad_norm": 0.21669740974903107,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 4217
+    },
+    {
+      "epoch": 0.5756789954961102,
+      "grad_norm": 0.21382921934127808,
+      "learning_rate": 8e-05,
+      "loss": 1.5123,
+      "step": 4218
+    },
+    {
+      "epoch": 0.5758154770028661,
+      "grad_norm": 0.2091756910085678,
+      "learning_rate": 8e-05,
+      "loss": 1.4511,
+      "step": 4219
+    },
+    {
+      "epoch": 0.575951958509622,
+      "grad_norm": 0.22944357991218567,
+      "learning_rate": 8e-05,
+      "loss": 1.5292,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5760884400163778,
+      "grad_norm": 0.22602970898151398,
+      "learning_rate": 8e-05,
+      "loss": 1.5171,
+      "step": 4221
+    },
+    {
+      "epoch": 0.5762249215231336,
+      "grad_norm": 0.2166919857263565,
+      "learning_rate": 8e-05,
+      "loss": 1.4535,
+      "step": 4222
+    },
+    {
+      "epoch": 0.5763614030298895,
+      "grad_norm": 0.22732417285442352,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 4223
+    },
+    {
+      "epoch": 0.5764978845366453,
+      "grad_norm": 0.21656647324562073,
+      "learning_rate": 8e-05,
+      "loss": 1.4863,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5766343660434011,
+      "grad_norm": 0.21638262271881104,
+      "learning_rate": 8e-05,
+      "loss": 1.4304,
+      "step": 4225
+    },
+    {
+      "epoch": 0.576770847550157,
+      "grad_norm": 0.21608994901180267,
+      "learning_rate": 8e-05,
+      "loss": 1.4176,
+      "step": 4226
+    },
+    {
+      "epoch": 0.5769073290569128,
+      "grad_norm": 0.21043655276298523,
+      "learning_rate": 8e-05,
+      "loss": 1.4525,
+      "step": 4227
+    },
+    {
+      "epoch": 0.5770438105636686,
+      "grad_norm": 0.21401606500148773,
+      "learning_rate": 8e-05,
+      "loss": 1.4862,
+      "step": 4228
+    },
+    {
+      "epoch": 0.5771802920704244,
+      "grad_norm": 0.22227658331394196,
+      "learning_rate": 8e-05,
+      "loss": 1.468,
+      "step": 4229
+    },
+    {
+      "epoch": 0.5773167735771803,
+      "grad_norm": 0.22011655569076538,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5774532550839361,
+      "grad_norm": 0.213987797498703,
+      "learning_rate": 8e-05,
+      "loss": 1.4764,
+      "step": 4231
+    },
+    {
+      "epoch": 0.577589736590692,
+      "grad_norm": 0.2208341807126999,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 4232
+    },
+    {
+      "epoch": 0.5777262180974478,
+      "grad_norm": 0.2314114272594452,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 4233
+    },
+    {
+      "epoch": 0.5778626996042037,
+      "grad_norm": 0.2215908020734787,
+      "learning_rate": 8e-05,
+      "loss": 1.5022,
+      "step": 4234
+    },
+    {
+      "epoch": 0.5779991811109595,
+      "grad_norm": 0.21595498919487,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5781356626177153,
+      "grad_norm": 0.2144959270954132,
+      "learning_rate": 8e-05,
+      "loss": 1.5031,
+      "step": 4236
+    },
+    {
+      "epoch": 0.5782721441244711,
+      "grad_norm": 0.20927192270755768,
+      "learning_rate": 8e-05,
+      "loss": 1.4652,
+      "step": 4237
+    },
+    {
+      "epoch": 0.578408625631227,
+      "grad_norm": 0.22002463042736053,
+      "learning_rate": 8e-05,
+      "loss": 1.4242,
+      "step": 4238
+    },
+    {
+      "epoch": 0.5785451071379828,
+      "grad_norm": 0.21876122057437897,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5786815886447386,
+      "grad_norm": 0.22128397226333618,
+      "learning_rate": 8e-05,
+      "loss": 1.532,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5788180701514944,
+      "grad_norm": 0.20736324787139893,
+      "learning_rate": 8e-05,
+      "loss": 1.4651,
+      "step": 4241
+    },
+    {
+      "epoch": 0.5789545516582503,
+      "grad_norm": 0.20933184027671814,
+      "learning_rate": 8e-05,
+      "loss": 1.42,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5790910331650061,
+      "grad_norm": 0.21193736791610718,
+      "learning_rate": 8e-05,
+      "loss": 1.3944,
+      "step": 4243
+    },
+    {
+      "epoch": 0.5792275146717619,
+      "grad_norm": 0.21933895349502563,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 4244
+    },
+    {
+      "epoch": 0.5793639961785179,
+      "grad_norm": 0.21755674481391907,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5795004776852737,
+      "grad_norm": 0.21228887140750885,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 4246
+    },
+    {
+      "epoch": 0.5796369591920295,
+      "grad_norm": 0.2192274034023285,
+      "learning_rate": 8e-05,
+      "loss": 1.5161,
+      "step": 4247
+    },
+    {
+      "epoch": 0.5797734406987853,
+      "grad_norm": 0.21627812087535858,
+      "learning_rate": 8e-05,
+      "loss": 1.4516,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5799099222055412,
+      "grad_norm": 0.21966254711151123,
+      "learning_rate": 8e-05,
+      "loss": 1.4812,
+      "step": 4249
+    },
+    {
+      "epoch": 0.580046403712297,
+      "grad_norm": 0.21831735968589783,
+      "learning_rate": 8e-05,
+      "loss": 1.4219,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5801828852190528,
+      "grad_norm": 0.21938075125217438,
+      "learning_rate": 8e-05,
+      "loss": 1.5152,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5803193667258086,
+      "grad_norm": 0.21802319586277008,
+      "learning_rate": 8e-05,
+      "loss": 1.4701,
+      "step": 4252
+    },
+    {
+      "epoch": 0.5804558482325645,
+      "grad_norm": 0.21984423696994781,
+      "learning_rate": 8e-05,
+      "loss": 1.5009,
+      "step": 4253
+    },
+    {
+      "epoch": 0.5805923297393203,
+      "grad_norm": 0.23234982788562775,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 4254
+    },
+    {
+      "epoch": 0.5807288112460761,
+      "grad_norm": 0.22150513529777527,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5808652927528319,
+      "grad_norm": 0.22058182954788208,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 4256
+    },
+    {
+      "epoch": 0.5810017742595879,
+      "grad_norm": 0.21117989718914032,
+      "learning_rate": 8e-05,
+      "loss": 1.4165,
+      "step": 4257
+    },
+    {
+      "epoch": 0.5811382557663437,
+      "grad_norm": 0.2250738888978958,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 4258
+    },
+    {
+      "epoch": 0.5812747372730995,
+      "grad_norm": 0.2053462713956833,
+      "learning_rate": 8e-05,
+      "loss": 1.4029,
+      "step": 4259
+    },
+    {
+      "epoch": 0.5814112187798554,
+      "grad_norm": 0.22288133203983307,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5815477002866112,
+      "grad_norm": 0.21730506420135498,
+      "learning_rate": 8e-05,
+      "loss": 1.4327,
+      "step": 4261
+    },
+    {
+      "epoch": 0.581684181793367,
+      "grad_norm": 0.21201132237911224,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 4262
+    },
+    {
+      "epoch": 0.5818206633001228,
+      "grad_norm": 0.21976672112941742,
+      "learning_rate": 8e-05,
+      "loss": 1.4977,
+      "step": 4263
+    },
+    {
+      "epoch": 0.5819571448068787,
+      "grad_norm": 0.21687081456184387,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 4264
+    },
+    {
+      "epoch": 0.5820936263136345,
+      "grad_norm": 0.21505995094776154,
+      "learning_rate": 8e-05,
+      "loss": 1.4763,
+      "step": 4265
+    },
+    {
+      "epoch": 0.5822301078203903,
+      "grad_norm": 0.21913163363933563,
+      "learning_rate": 8e-05,
+      "loss": 1.5281,
+      "step": 4266
+    },
+    {
+      "epoch": 0.5823665893271461,
+      "grad_norm": 0.220303475856781,
+      "learning_rate": 8e-05,
+      "loss": 1.4776,
+      "step": 4267
+    },
+    {
+      "epoch": 0.582503070833902,
+      "grad_norm": 0.21774540841579437,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 4268
+    },
+    {
+      "epoch": 0.5826395523406579,
+      "grad_norm": 0.22189544141292572,
+      "learning_rate": 8e-05,
+      "loss": 1.5697,
+      "step": 4269
+    },
+    {
+      "epoch": 0.5827760338474137,
+      "grad_norm": 0.2093462198972702,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 4270
+    },
+    {
+      "epoch": 0.5829125153541695,
+      "grad_norm": 0.22963126003742218,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 4271
+    },
+    {
+      "epoch": 0.5830489968609254,
+      "grad_norm": 0.22692115604877472,
+      "learning_rate": 8e-05,
+      "loss": 1.532,
+      "step": 4272
+    },
+    {
+      "epoch": 0.5831854783676812,
+      "grad_norm": 0.21378524601459503,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4273
+    },
+    {
+      "epoch": 0.583321959874437,
+      "grad_norm": 0.21572093665599823,
+      "learning_rate": 8e-05,
+      "loss": 1.4874,
+      "step": 4274
+    },
+    {
+      "epoch": 0.5834584413811928,
+      "grad_norm": 0.21749195456504822,
+      "learning_rate": 8e-05,
+      "loss": 1.4915,
+      "step": 4275
+    },
+    {
+      "epoch": 0.5835949228879487,
+      "grad_norm": 0.21253842115402222,
+      "learning_rate": 8e-05,
+      "loss": 1.4768,
+      "step": 4276
+    },
+    {
+      "epoch": 0.5837314043947045,
+      "grad_norm": 0.2171137034893036,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 4277
+    },
+    {
+      "epoch": 0.5838678859014603,
+      "grad_norm": 0.20603013038635254,
+      "learning_rate": 8e-05,
+      "loss": 1.4383,
+      "step": 4278
+    },
+    {
+      "epoch": 0.5840043674082162,
+      "grad_norm": 0.20807403326034546,
+      "learning_rate": 8e-05,
+      "loss": 1.4516,
+      "step": 4279
+    },
+    {
+      "epoch": 0.584140848914972,
+      "grad_norm": 0.21349146962165833,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 4280
+    },
+    {
+      "epoch": 0.5842773304217278,
+      "grad_norm": 0.22212353348731995,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 4281
+    },
+    {
+      "epoch": 0.5844138119284837,
+      "grad_norm": 0.2233966886997223,
+      "learning_rate": 8e-05,
+      "loss": 1.5029,
+      "step": 4282
+    },
+    {
+      "epoch": 0.5845502934352396,
+      "grad_norm": 0.21821431815624237,
+      "learning_rate": 8e-05,
+      "loss": 1.4709,
+      "step": 4283
+    },
+    {
+      "epoch": 0.5846867749419954,
+      "grad_norm": 0.20830968022346497,
+      "learning_rate": 8e-05,
+      "loss": 1.4511,
+      "step": 4284
+    },
+    {
+      "epoch": 0.5848232564487512,
+      "grad_norm": 0.21082666516304016,
+      "learning_rate": 8e-05,
+      "loss": 1.4323,
+      "step": 4285
+    },
+    {
+      "epoch": 0.584959737955507,
+      "grad_norm": 0.21043847501277924,
+      "learning_rate": 8e-05,
+      "loss": 1.4373,
+      "step": 4286
+    },
+    {
+      "epoch": 0.5850962194622629,
+      "grad_norm": 0.2249893993139267,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 4287
+    },
+    {
+      "epoch": 0.5852327009690187,
+      "grad_norm": 0.21635067462921143,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 4288
+    },
+    {
+      "epoch": 0.5853691824757745,
+      "grad_norm": 0.21915218234062195,
+      "learning_rate": 8e-05,
+      "loss": 1.5016,
+      "step": 4289
+    },
+    {
+      "epoch": 0.5855056639825303,
+      "grad_norm": 0.2193540632724762,
+      "learning_rate": 8e-05,
+      "loss": 1.3985,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5856421454892862,
+      "grad_norm": 0.22085200250148773,
+      "learning_rate": 8e-05,
+      "loss": 1.4823,
+      "step": 4291
+    },
+    {
+      "epoch": 0.585778626996042,
+      "grad_norm": 0.22260814905166626,
+      "learning_rate": 8e-05,
+      "loss": 1.4219,
+      "step": 4292
+    },
+    {
+      "epoch": 0.5859151085027978,
+      "grad_norm": 0.22045950591564178,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 4293
+    },
+    {
+      "epoch": 0.5860515900095538,
+      "grad_norm": 0.21894249320030212,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 4294
+    },
+    {
+      "epoch": 0.5861880715163096,
+      "grad_norm": 0.21819111704826355,
+      "learning_rate": 8e-05,
+      "loss": 1.4882,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5863245530230654,
+      "grad_norm": 0.21281464397907257,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 4296
+    },
+    {
+      "epoch": 0.5864610345298212,
+      "grad_norm": 0.2163061797618866,
+      "learning_rate": 8e-05,
+      "loss": 1.469,
+      "step": 4297
+    },
+    {
+      "epoch": 0.5865975160365771,
+      "grad_norm": 0.21715545654296875,
+      "learning_rate": 8e-05,
+      "loss": 1.5247,
+      "step": 4298
+    },
+    {
+      "epoch": 0.5867339975433329,
+      "grad_norm": 0.2198237180709839,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 4299
+    },
+    {
+      "epoch": 0.5868704790500887,
+      "grad_norm": 0.21444237232208252,
+      "learning_rate": 8e-05,
+      "loss": 1.4838,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5870069605568445,
+      "grad_norm": 0.21362698078155518,
+      "learning_rate": 8e-05,
+      "loss": 1.4141,
+      "step": 4301
+    },
+    {
+      "epoch": 0.5871434420636004,
+      "grad_norm": 0.21045954525470734,
+      "learning_rate": 8e-05,
+      "loss": 1.4745,
+      "step": 4302
+    },
+    {
+      "epoch": 0.5872799235703562,
+      "grad_norm": 0.22438058257102966,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 4303
+    },
+    {
+      "epoch": 0.587416405077112,
+      "grad_norm": 0.22046953439712524,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4304
+    },
+    {
+      "epoch": 0.5875528865838678,
+      "grad_norm": 0.2153709977865219,
+      "learning_rate": 8e-05,
+      "loss": 1.4351,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5876893680906237,
+      "grad_norm": 0.22235837578773499,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 4306
+    },
+    {
+      "epoch": 0.5878258495973796,
+      "grad_norm": 0.20900604128837585,
+      "learning_rate": 8e-05,
+      "loss": 1.436,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5879623311041354,
+      "grad_norm": 0.21753929555416107,
+      "learning_rate": 8e-05,
+      "loss": 1.4049,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5880988126108913,
+      "grad_norm": 0.21831852197647095,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.22192572057247162,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5883717756244029,
+      "grad_norm": 0.21853183209896088,
+      "learning_rate": 8e-05,
+      "loss": 1.4609,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5885082571311587,
+      "grad_norm": 0.21943886578083038,
+      "learning_rate": 8e-05,
+      "loss": 1.4369,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5886447386379146,
+      "grad_norm": 0.21709951758384705,
+      "learning_rate": 8e-05,
+      "loss": 1.5777,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5887812201446704,
+      "grad_norm": 0.2150854915380478,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5889177016514262,
+      "grad_norm": 0.21481004357337952,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 4315
+    },
+    {
+      "epoch": 0.589054183158182,
+      "grad_norm": 0.20692256093025208,
+      "learning_rate": 8e-05,
+      "loss": 1.4759,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5891906646649379,
+      "grad_norm": 0.2155621349811554,
+      "learning_rate": 8e-05,
+      "loss": 1.4568,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5893271461716937,
+      "grad_norm": 0.22038252651691437,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 4318
+    },
+    {
+      "epoch": 0.5894636276784496,
+      "grad_norm": 0.2143627107143402,
+      "learning_rate": 8e-05,
+      "loss": 1.4941,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5896001091852054,
+      "grad_norm": 0.21517311036586761,
+      "learning_rate": 8e-05,
+      "loss": 1.4454,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5897365906919613,
+      "grad_norm": 0.212874174118042,
+      "learning_rate": 8e-05,
+      "loss": 1.4637,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5898730721987171,
+      "grad_norm": 0.21814869344234467,
+      "learning_rate": 8e-05,
+      "loss": 1.4488,
+      "step": 4322
+    },
+    {
+      "epoch": 0.5900095537054729,
+      "grad_norm": 0.21869415044784546,
+      "learning_rate": 8e-05,
+      "loss": 1.4721,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5901460352122287,
+      "grad_norm": 0.21262279152870178,
+      "learning_rate": 8e-05,
+      "loss": 1.3927,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5902825167189846,
+      "grad_norm": 0.20953911542892456,
+      "learning_rate": 8e-05,
+      "loss": 1.451,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5904189982257404,
+      "grad_norm": 0.21339575946331024,
+      "learning_rate": 8e-05,
+      "loss": 1.4435,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5905554797324962,
+      "grad_norm": 0.22168762981891632,
+      "learning_rate": 8e-05,
+      "loss": 1.5393,
+      "step": 4327
+    },
+    {
+      "epoch": 0.590691961239252,
+      "grad_norm": 0.22402767837047577,
+      "learning_rate": 8e-05,
+      "loss": 1.4545,
+      "step": 4328
+    },
+    {
+      "epoch": 0.5908284427460079,
+      "grad_norm": 0.21322095394134521,
+      "learning_rate": 8e-05,
+      "loss": 1.5347,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5909649242527637,
+      "grad_norm": 0.2209540754556656,
+      "learning_rate": 8e-05,
+      "loss": 1.5112,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5911014057595196,
+      "grad_norm": 0.2232252061367035,
+      "learning_rate": 8e-05,
+      "loss": 1.4771,
+      "step": 4331
+    },
+    {
+      "epoch": 0.5912378872662755,
+      "grad_norm": 0.21956667304039001,
+      "learning_rate": 8e-05,
+      "loss": 1.4988,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5913743687730313,
+      "grad_norm": 0.2251167744398117,
+      "learning_rate": 8e-05,
+      "loss": 1.4661,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5915108502797871,
+      "grad_norm": 0.20992282032966614,
+      "learning_rate": 8e-05,
+      "loss": 1.5166,
+      "step": 4334
+    },
+    {
+      "epoch": 0.5916473317865429,
+      "grad_norm": 0.20996713638305664,
+      "learning_rate": 8e-05,
+      "loss": 1.4228,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5917838132932988,
+      "grad_norm": 0.22183406352996826,
+      "learning_rate": 8e-05,
+      "loss": 1.4592,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5919202948000546,
+      "grad_norm": 0.21954160928726196,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 4337
+    },
+    {
+      "epoch": 0.5920567763068104,
+      "grad_norm": 0.2262692004442215,
+      "learning_rate": 8e-05,
+      "loss": 1.4789,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5921932578135662,
+      "grad_norm": 0.2285604327917099,
+      "learning_rate": 8e-05,
+      "loss": 1.5031,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5923297393203221,
+      "grad_norm": 0.22797690331935883,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5924662208270779,
+      "grad_norm": 0.2103288173675537,
+      "learning_rate": 8e-05,
+      "loss": 1.3893,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5926027023338337,
+      "grad_norm": 0.2192559391260147,
+      "learning_rate": 8e-05,
+      "loss": 1.4687,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5927391838405895,
+      "grad_norm": 0.21651481091976166,
+      "learning_rate": 8e-05,
+      "loss": 1.4397,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5928756653473455,
+      "grad_norm": 0.20992860198020935,
+      "learning_rate": 8e-05,
+      "loss": 1.4252,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5930121468541013,
+      "grad_norm": 0.22292640805244446,
+      "learning_rate": 8e-05,
+      "loss": 1.415,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5931486283608571,
+      "grad_norm": 0.2203783541917801,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 4346
+    },
+    {
+      "epoch": 0.593285109867613,
+      "grad_norm": 0.20854653418064117,
+      "learning_rate": 8e-05,
+      "loss": 1.4086,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5934215913743688,
+      "grad_norm": 0.22053131461143494,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5935580728811246,
+      "grad_norm": 0.21370857954025269,
+      "learning_rate": 8e-05,
+      "loss": 1.4635,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5936945543878804,
+      "grad_norm": 0.21468520164489746,
+      "learning_rate": 8e-05,
+      "loss": 1.5132,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5938310358946363,
+      "grad_norm": 0.20682238042354584,
+      "learning_rate": 8e-05,
+      "loss": 1.3926,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5939675174013921,
+      "grad_norm": 0.2136172503232956,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5941039989081479,
+      "grad_norm": 0.21396876871585846,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5942404804149037,
+      "grad_norm": 0.21121741831302643,
+      "learning_rate": 8e-05,
+      "loss": 1.4932,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5943769619216596,
+      "grad_norm": 0.22208641469478607,
+      "learning_rate": 8e-05,
+      "loss": 1.4828,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5945134434284155,
+      "grad_norm": 0.21309731900691986,
+      "learning_rate": 8e-05,
+      "loss": 1.4857,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5946499249351713,
+      "grad_norm": 0.21267130970954895,
+      "learning_rate": 8e-05,
+      "loss": 1.5325,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5947864064419272,
+      "grad_norm": 0.21058547496795654,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 4358
+    },
+    {
+      "epoch": 0.594922887948683,
+      "grad_norm": 0.22066126763820648,
+      "learning_rate": 8e-05,
+      "loss": 1.5336,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5950593694554388,
+      "grad_norm": 0.21684888005256653,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5951958509621946,
+      "grad_norm": 0.21596623957157135,
+      "learning_rate": 8e-05,
+      "loss": 1.3702,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5953323324689505,
+      "grad_norm": 0.21503545343875885,
+      "learning_rate": 8e-05,
+      "loss": 1.4561,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5954688139757063,
+      "grad_norm": 0.2201627790927887,
+      "learning_rate": 8e-05,
+      "loss": 1.4964,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5956052954824621,
+      "grad_norm": 0.21484950184822083,
+      "learning_rate": 8e-05,
+      "loss": 1.4513,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5957417769892179,
+      "grad_norm": 0.21879707276821136,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5958782584959738,
+      "grad_norm": 0.21750949323177338,
+      "learning_rate": 8e-05,
+      "loss": 1.4572,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5960147400027296,
+      "grad_norm": 0.220553919672966,
+      "learning_rate": 8e-05,
+      "loss": 1.4999,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5961512215094855,
+      "grad_norm": 0.21693886816501617,
+      "learning_rate": 8e-05,
+      "loss": 1.4629,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5962877030162413,
+      "grad_norm": 0.22107239067554474,
+      "learning_rate": 8e-05,
+      "loss": 1.4695,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5964241845229972,
+      "grad_norm": 0.21459034085273743,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 4370
+    },
+    {
+      "epoch": 0.596560666029753,
+      "grad_norm": 0.21887944638729095,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5966971475365088,
+      "grad_norm": 0.22586947679519653,
+      "learning_rate": 8e-05,
+      "loss": 1.4976,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5968336290432646,
+      "grad_norm": 0.21632890403270721,
+      "learning_rate": 8e-05,
+      "loss": 1.47,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5969701105500205,
+      "grad_norm": 0.22188736498355865,
+      "learning_rate": 8e-05,
+      "loss": 1.5414,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5971065920567763,
+      "grad_norm": 0.2110891044139862,
+      "learning_rate": 8e-05,
+      "loss": 1.4918,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5972430735635321,
+      "grad_norm": 0.21347954869270325,
+      "learning_rate": 8e-05,
+      "loss": 1.5443,
+      "step": 4376
+    },
+    {
+      "epoch": 0.597379555070288,
+      "grad_norm": 0.21682076156139374,
+      "learning_rate": 8e-05,
+      "loss": 1.522,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5975160365770438,
+      "grad_norm": 0.22219955921173096,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5976525180837996,
+      "grad_norm": 0.21607360243797302,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 4379
+    },
+    {
+      "epoch": 0.5977889995905554,
+      "grad_norm": 0.21729676425457,
+      "learning_rate": 8e-05,
+      "loss": 1.4853,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5979254810973114,
+      "grad_norm": 0.2149355709552765,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 4381
+    },
+    {
+      "epoch": 0.5980619626040672,
+      "grad_norm": 0.22193880379199982,
+      "learning_rate": 8e-05,
+      "loss": 1.4869,
+      "step": 4382
+    },
+    {
+      "epoch": 0.598198444110823,
+      "grad_norm": 0.2167275995016098,
+      "learning_rate": 8e-05,
+      "loss": 1.4121,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5983349256175788,
+      "grad_norm": 0.22075915336608887,
+      "learning_rate": 8e-05,
+      "loss": 1.5372,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5984714071243347,
+      "grad_norm": 0.22093677520751953,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5986078886310905,
+      "grad_norm": 0.2314719706773758,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5987443701378463,
+      "grad_norm": 0.21571598947048187,
+      "learning_rate": 8e-05,
+      "loss": 1.4838,
+      "step": 4387
+    },
+    {
+      "epoch": 0.5988808516446021,
+      "grad_norm": 0.21857315301895142,
+      "learning_rate": 8e-05,
+      "loss": 1.4812,
+      "step": 4388
+    },
+    {
+      "epoch": 0.599017333151358,
+      "grad_norm": 0.2169128954410553,
+      "learning_rate": 8e-05,
+      "loss": 1.4476,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5991538146581138,
+      "grad_norm": 0.22191467881202698,
+      "learning_rate": 8e-05,
+      "loss": 1.3826,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5992902961648696,
+      "grad_norm": 0.2280043363571167,
+      "learning_rate": 8e-05,
+      "loss": 1.5041,
+      "step": 4391
+    },
+    {
+      "epoch": 0.5994267776716254,
+      "grad_norm": 0.22332550585269928,
+      "learning_rate": 8e-05,
+      "loss": 1.4343,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5995632591783814,
+      "grad_norm": 0.2141445130109787,
+      "learning_rate": 8e-05,
+      "loss": 1.4765,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5996997406851372,
+      "grad_norm": 0.21590353548526764,
+      "learning_rate": 8e-05,
+      "loss": 1.498,
+      "step": 4394
+    },
+    {
+      "epoch": 0.599836222191893,
+      "grad_norm": 0.22238901257514954,
+      "learning_rate": 8e-05,
+      "loss": 1.5007,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5999727036986489,
+      "grad_norm": 0.21094655990600586,
+      "learning_rate": 8e-05,
+      "loss": 1.4596,
+      "step": 4396
+    },
+    {
+      "epoch": 0.6001091852054047,
+      "grad_norm": 0.22575178742408752,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4397
+    },
+    {
+      "epoch": 0.6002456667121605,
+      "grad_norm": 0.21845540404319763,
+      "learning_rate": 8e-05,
+      "loss": 1.4768,
+      "step": 4398
+    },
+    {
+      "epoch": 0.6003821482189163,
+      "grad_norm": 0.21899642050266266,
+      "learning_rate": 8e-05,
+      "loss": 1.4704,
+      "step": 4399
+    },
+    {
+      "epoch": 0.6005186297256722,
+      "grad_norm": 0.23280362784862518,
+      "learning_rate": 8e-05,
+      "loss": 1.479,
+      "step": 4400
+    },
+    {
+      "epoch": 0.600655111232428,
+      "grad_norm": 0.21607975661754608,
+      "learning_rate": 8e-05,
+      "loss": 1.4399,
+      "step": 4401
+    },
+    {
+      "epoch": 0.6007915927391838,
+      "grad_norm": 0.21461284160614014,
+      "learning_rate": 8e-05,
+      "loss": 1.4183,
+      "step": 4402
+    },
+    {
+      "epoch": 0.6009280742459396,
+      "grad_norm": 0.22280895709991455,
+      "learning_rate": 8e-05,
+      "loss": 1.4653,
+      "step": 4403
+    },
+    {
+      "epoch": 0.6010645557526955,
+      "grad_norm": 0.22206827998161316,
+      "learning_rate": 8e-05,
+      "loss": 1.4804,
+      "step": 4404
+    },
+    {
+      "epoch": 0.6012010372594514,
+      "grad_norm": 0.20972760021686554,
+      "learning_rate": 8e-05,
+      "loss": 1.4289,
+      "step": 4405
+    },
+    {
+      "epoch": 0.6013375187662072,
+      "grad_norm": 0.22166205942630768,
+      "learning_rate": 8e-05,
+      "loss": 1.4835,
+      "step": 4406
+    },
+    {
+      "epoch": 0.601474000272963,
+      "grad_norm": 0.2218756526708603,
+      "learning_rate": 8e-05,
+      "loss": 1.4418,
+      "step": 4407
+    },
+    {
+      "epoch": 0.6016104817797189,
+      "grad_norm": 0.20884943008422852,
+      "learning_rate": 8e-05,
+      "loss": 1.4343,
+      "step": 4408
+    },
+    {
+      "epoch": 0.6017469632864747,
+      "grad_norm": 0.21617570519447327,
+      "learning_rate": 8e-05,
+      "loss": 1.4472,
+      "step": 4409
+    },
+    {
+      "epoch": 0.6018834447932305,
+      "grad_norm": 0.2209385484457016,
+      "learning_rate": 8e-05,
+      "loss": 1.4672,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6020199262999864,
+      "grad_norm": 0.21622245013713837,
+      "learning_rate": 8e-05,
+      "loss": 1.4071,
+      "step": 4411
+    },
+    {
+      "epoch": 0.6021564078067422,
+      "grad_norm": 0.21497902274131775,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 4412
+    },
+    {
+      "epoch": 0.602292889313498,
+      "grad_norm": 0.2144337147474289,
+      "learning_rate": 8e-05,
+      "loss": 1.4656,
+      "step": 4413
+    },
+    {
+      "epoch": 0.6024293708202538,
+      "grad_norm": 0.21735015511512756,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 4414
+    },
+    {
+      "epoch": 0.6025658523270097,
+      "grad_norm": 0.2120286375284195,
+      "learning_rate": 8e-05,
+      "loss": 1.4712,
+      "step": 4415
+    },
+    {
+      "epoch": 0.6027023338337655,
+      "grad_norm": 0.222615048289299,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 4416
+    },
+    {
+      "epoch": 0.6028388153405213,
+      "grad_norm": 0.21790429949760437,
+      "learning_rate": 8e-05,
+      "loss": 1.4811,
+      "step": 4417
+    },
+    {
+      "epoch": 0.6029752968472772,
+      "grad_norm": 0.2232406884431839,
+      "learning_rate": 8e-05,
+      "loss": 1.4829,
+      "step": 4418
+    },
+    {
+      "epoch": 0.6031117783540331,
+      "grad_norm": 0.2130293995141983,
+      "learning_rate": 8e-05,
+      "loss": 1.4915,
+      "step": 4419
+    },
+    {
+      "epoch": 0.6032482598607889,
+      "grad_norm": 0.2127717137336731,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6033847413675447,
+      "grad_norm": 0.22728444635868073,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 4421
+    },
+    {
+      "epoch": 0.6035212228743005,
+      "grad_norm": 0.21519602835178375,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 4422
+    },
+    {
+      "epoch": 0.6036577043810564,
+      "grad_norm": 0.21472813189029694,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 4423
+    },
+    {
+      "epoch": 0.6037941858878122,
+      "grad_norm": 0.2117292732000351,
+      "learning_rate": 8e-05,
+      "loss": 1.4238,
+      "step": 4424
+    },
+    {
+      "epoch": 0.603930667394568,
+      "grad_norm": 0.21555130183696747,
+      "learning_rate": 8e-05,
+      "loss": 1.5036,
+      "step": 4425
+    },
+    {
+      "epoch": 0.6040671489013238,
+      "grad_norm": 0.20432916283607483,
+      "learning_rate": 8e-05,
+      "loss": 1.4271,
+      "step": 4426
+    },
+    {
+      "epoch": 0.6042036304080797,
+      "grad_norm": 0.22728495299816132,
+      "learning_rate": 8e-05,
+      "loss": 1.5311,
+      "step": 4427
+    },
+    {
+      "epoch": 0.6043401119148355,
+      "grad_norm": 0.22419095039367676,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 4428
+    },
+    {
+      "epoch": 0.6044765934215913,
+      "grad_norm": 0.21420539915561676,
+      "learning_rate": 8e-05,
+      "loss": 1.4006,
+      "step": 4429
+    },
+    {
+      "epoch": 0.6046130749283473,
+      "grad_norm": 0.21803054213523865,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6047495564351031,
+      "grad_norm": 0.22154615819454193,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 4431
+    },
+    {
+      "epoch": 0.6048860379418589,
+      "grad_norm": 0.213043212890625,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 4432
+    },
+    {
+      "epoch": 0.6050225194486147,
+      "grad_norm": 0.22341564297676086,
+      "learning_rate": 8e-05,
+      "loss": 1.4967,
+      "step": 4433
+    },
+    {
+      "epoch": 0.6051590009553706,
+      "grad_norm": 0.21855106949806213,
+      "learning_rate": 8e-05,
+      "loss": 1.4474,
+      "step": 4434
+    },
+    {
+      "epoch": 0.6052954824621264,
+      "grad_norm": 0.2241329550743103,
+      "learning_rate": 8e-05,
+      "loss": 1.4737,
+      "step": 4435
+    },
+    {
+      "epoch": 0.6054319639688822,
+      "grad_norm": 0.21529099345207214,
+      "learning_rate": 8e-05,
+      "loss": 1.4422,
+      "step": 4436
+    },
+    {
+      "epoch": 0.605568445475638,
+      "grad_norm": 0.22286510467529297,
+      "learning_rate": 8e-05,
+      "loss": 1.4814,
+      "step": 4437
+    },
+    {
+      "epoch": 0.6057049269823939,
+      "grad_norm": 0.23753386735916138,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 4438
+    },
+    {
+      "epoch": 0.6058414084891497,
+      "grad_norm": 0.22039948403835297,
+      "learning_rate": 8e-05,
+      "loss": 1.4174,
+      "step": 4439
+    },
+    {
+      "epoch": 0.6059778899959055,
+      "grad_norm": 0.22198425233364105,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6061143715026613,
+      "grad_norm": 0.22369617223739624,
+      "learning_rate": 8e-05,
+      "loss": 1.4947,
+      "step": 4441
+    },
+    {
+      "epoch": 0.6062508530094173,
+      "grad_norm": 0.22572718560695648,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 4442
+    },
+    {
+      "epoch": 0.6063873345161731,
+      "grad_norm": 0.2198096364736557,
+      "learning_rate": 8e-05,
+      "loss": 1.4973,
+      "step": 4443
+    },
+    {
+      "epoch": 0.6065238160229289,
+      "grad_norm": 0.22440186142921448,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 4444
+    },
+    {
+      "epoch": 0.6066602975296848,
+      "grad_norm": 0.21684330701828003,
+      "learning_rate": 8e-05,
+      "loss": 1.447,
+      "step": 4445
+    },
+    {
+      "epoch": 0.6067967790364406,
+      "grad_norm": 0.21544249355793,
+      "learning_rate": 8e-05,
+      "loss": 1.3919,
+      "step": 4446
+    },
+    {
+      "epoch": 0.6069332605431964,
+      "grad_norm": 0.22006885707378387,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 4447
+    },
+    {
+      "epoch": 0.6070697420499522,
+      "grad_norm": 0.2181139439344406,
+      "learning_rate": 8e-05,
+      "loss": 1.4701,
+      "step": 4448
+    },
+    {
+      "epoch": 0.6072062235567081,
+      "grad_norm": 0.2256515622138977,
+      "learning_rate": 8e-05,
+      "loss": 1.4683,
+      "step": 4449
+    },
+    {
+      "epoch": 0.6073427050634639,
+      "grad_norm": 0.2163325846195221,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6074791865702197,
+      "grad_norm": 0.22249732911586761,
+      "learning_rate": 8e-05,
+      "loss": 1.488,
+      "step": 4451
+    },
+    {
+      "epoch": 0.6076156680769755,
+      "grad_norm": 0.21261267364025116,
+      "learning_rate": 8e-05,
+      "loss": 1.3938,
+      "step": 4452
+    },
+    {
+      "epoch": 0.6077521495837314,
+      "grad_norm": 0.22215509414672852,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 4453
+    },
+    {
+      "epoch": 0.6078886310904872,
+      "grad_norm": 0.21968472003936768,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 4454
+    },
+    {
+      "epoch": 0.6080251125972431,
+      "grad_norm": 0.22067271173000336,
+      "learning_rate": 8e-05,
+      "loss": 1.4487,
+      "step": 4455
+    },
+    {
+      "epoch": 0.608161594103999,
+      "grad_norm": 0.21898362040519714,
+      "learning_rate": 8e-05,
+      "loss": 1.4703,
+      "step": 4456
+    },
+    {
+      "epoch": 0.6082980756107548,
+      "grad_norm": 0.22021907567977905,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 4457
+    },
+    {
+      "epoch": 0.6084345571175106,
+      "grad_norm": 0.22044025361537933,
+      "learning_rate": 8e-05,
+      "loss": 1.3641,
+      "step": 4458
+    },
+    {
+      "epoch": 0.6085710386242664,
+      "grad_norm": 0.22148315608501434,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 4459
+    },
+    {
+      "epoch": 0.6087075201310223,
+      "grad_norm": 0.2195834070444107,
+      "learning_rate": 8e-05,
+      "loss": 1.5054,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6088440016377781,
+      "grad_norm": 0.2089032232761383,
+      "learning_rate": 8e-05,
+      "loss": 1.3413,
+      "step": 4461
+    },
+    {
+      "epoch": 0.6089804831445339,
+      "grad_norm": 0.21604180335998535,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 4462
+    },
+    {
+      "epoch": 0.6091169646512897,
+      "grad_norm": 0.21705101430416107,
+      "learning_rate": 8e-05,
+      "loss": 1.4492,
+      "step": 4463
+    },
+    {
+      "epoch": 0.6092534461580456,
+      "grad_norm": 0.22070030868053436,
+      "learning_rate": 8e-05,
+      "loss": 1.4829,
+      "step": 4464
+    },
+    {
+      "epoch": 0.6093899276648014,
+      "grad_norm": 0.2307535707950592,
+      "learning_rate": 8e-05,
+      "loss": 1.58,
+      "step": 4465
+    },
+    {
+      "epoch": 0.6095264091715572,
+      "grad_norm": 0.21350371837615967,
+      "learning_rate": 8e-05,
+      "loss": 1.4244,
+      "step": 4466
+    },
+    {
+      "epoch": 0.6096628906783131,
+      "grad_norm": 0.21948514878749847,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 4467
+    },
+    {
+      "epoch": 0.609799372185069,
+      "grad_norm": 0.2194800227880478,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 4468
+    },
+    {
+      "epoch": 0.6099358536918248,
+      "grad_norm": 0.22375673055648804,
+      "learning_rate": 8e-05,
+      "loss": 1.5222,
+      "step": 4469
+    },
+    {
+      "epoch": 0.6100723351985806,
+      "grad_norm": 0.22076460719108582,
+      "learning_rate": 8e-05,
+      "loss": 1.571,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6102088167053364,
+      "grad_norm": 0.21777185797691345,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4471
+    },
+    {
+      "epoch": 0.6103452982120923,
+      "grad_norm": 0.22119440138339996,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 4472
+    },
+    {
+      "epoch": 0.6104817797188481,
+      "grad_norm": 0.22467420995235443,
+      "learning_rate": 8e-05,
+      "loss": 1.5233,
+      "step": 4473
+    },
+    {
+      "epoch": 0.6106182612256039,
+      "grad_norm": 0.22084012627601624,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 4474
+    },
+    {
+      "epoch": 0.6107547427323597,
+      "grad_norm": 0.21889568865299225,
+      "learning_rate": 8e-05,
+      "loss": 1.444,
+      "step": 4475
+    },
+    {
+      "epoch": 0.6108912242391156,
+      "grad_norm": 0.20903992652893066,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 4476
+    },
+    {
+      "epoch": 0.6110277057458714,
+      "grad_norm": 0.2155047059059143,
+      "learning_rate": 8e-05,
+      "loss": 1.503,
+      "step": 4477
+    },
+    {
+      "epoch": 0.6111641872526272,
+      "grad_norm": 0.2194477915763855,
+      "learning_rate": 8e-05,
+      "loss": 1.4638,
+      "step": 4478
+    },
+    {
+      "epoch": 0.6113006687593832,
+      "grad_norm": 0.2223316878080368,
+      "learning_rate": 8e-05,
+      "loss": 1.4389,
+      "step": 4479
+    },
+    {
+      "epoch": 0.611437150266139,
+      "grad_norm": 0.2256895750761032,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6115736317728948,
+      "grad_norm": 0.21991045773029327,
+      "learning_rate": 8e-05,
+      "loss": 1.4957,
+      "step": 4481
+    },
+    {
+      "epoch": 0.6117101132796506,
+      "grad_norm": 0.24591347575187683,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 4482
+    },
+    {
+      "epoch": 0.6118465947864065,
+      "grad_norm": 0.21895599365234375,
+      "learning_rate": 8e-05,
+      "loss": 1.5097,
+      "step": 4483
+    },
+    {
+      "epoch": 0.6119830762931623,
+      "grad_norm": 0.21773697435855865,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 4484
+    },
+    {
+      "epoch": 0.6121195577999181,
+      "grad_norm": 0.21382389962673187,
+      "learning_rate": 8e-05,
+      "loss": 1.5024,
+      "step": 4485
+    },
+    {
+      "epoch": 0.6122560393066739,
+      "grad_norm": 0.21890582144260406,
+      "learning_rate": 8e-05,
+      "loss": 1.4283,
+      "step": 4486
+    },
+    {
+      "epoch": 0.6123925208134298,
+      "grad_norm": 0.21945111453533173,
+      "learning_rate": 8e-05,
+      "loss": 1.5511,
+      "step": 4487
+    },
+    {
+      "epoch": 0.6125290023201856,
+      "grad_norm": 0.2107498198747635,
+      "learning_rate": 8e-05,
+      "loss": 1.4393,
+      "step": 4488
+    },
+    {
+      "epoch": 0.6126654838269414,
+      "grad_norm": 0.21651676297187805,
+      "learning_rate": 8e-05,
+      "loss": 1.4496,
+      "step": 4489
+    },
+    {
+      "epoch": 0.6128019653336972,
+      "grad_norm": 0.21643996238708496,
+      "learning_rate": 8e-05,
+      "loss": 1.382,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6129384468404531,
+      "grad_norm": 0.23711667954921722,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 4491
+    },
+    {
+      "epoch": 0.613074928347209,
+      "grad_norm": 0.22493483126163483,
+      "learning_rate": 8e-05,
+      "loss": 1.5201,
+      "step": 4492
+    },
+    {
+      "epoch": 0.6132114098539648,
+      "grad_norm": 0.22030393779277802,
+      "learning_rate": 8e-05,
+      "loss": 1.4746,
+      "step": 4493
+    },
+    {
+      "epoch": 0.6133478913607207,
+      "grad_norm": 0.218339741230011,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 4494
+    },
+    {
+      "epoch": 0.6134843728674765,
+      "grad_norm": 0.21164961159229279,
+      "learning_rate": 8e-05,
+      "loss": 1.4615,
+      "step": 4495
+    },
+    {
+      "epoch": 0.6136208543742323,
+      "grad_norm": 0.21988964080810547,
+      "learning_rate": 8e-05,
+      "loss": 1.4829,
+      "step": 4496
+    },
+    {
+      "epoch": 0.6137573358809881,
+      "grad_norm": 0.21515649557113647,
+      "learning_rate": 8e-05,
+      "loss": 1.4605,
+      "step": 4497
+    },
+    {
+      "epoch": 0.613893817387744,
+      "grad_norm": 0.22704173624515533,
+      "learning_rate": 8e-05,
+      "loss": 1.539,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6140302988944998,
+      "grad_norm": 0.21372933685779572,
+      "learning_rate": 8e-05,
+      "loss": 1.4904,
+      "step": 4499
+    },
+    {
+      "epoch": 0.6141667804012556,
+      "grad_norm": 0.22876450419425964,
+      "learning_rate": 8e-05,
+      "loss": 1.5148,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6143032619080114,
+      "grad_norm": 0.21961663663387299,
+      "learning_rate": 8e-05,
+      "loss": 1.4927,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6144397434147673,
+      "grad_norm": 0.21080389618873596,
+      "learning_rate": 8e-05,
+      "loss": 1.4283,
+      "step": 4502
+    },
+    {
+      "epoch": 0.6145762249215231,
+      "grad_norm": 0.2210008054971695,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 4503
+    },
+    {
+      "epoch": 0.614712706428279,
+      "grad_norm": 0.21898935735225677,
+      "learning_rate": 8e-05,
+      "loss": 1.4297,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6148491879350348,
+      "grad_norm": 0.22737912833690643,
+      "learning_rate": 8e-05,
+      "loss": 1.5438,
+      "step": 4505
+    },
+    {
+      "epoch": 0.6149856694417907,
+      "grad_norm": 0.22331592440605164,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 4506
+    },
+    {
+      "epoch": 0.6151221509485465,
+      "grad_norm": 0.21980531513690948,
+      "learning_rate": 8e-05,
+      "loss": 1.4633,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6152586324553023,
+      "grad_norm": 0.22071319818496704,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 4508
+    },
+    {
+      "epoch": 0.6153951139620581,
+      "grad_norm": 0.21640212833881378,
+      "learning_rate": 8e-05,
+      "loss": 1.4671,
+      "step": 4509
+    },
+    {
+      "epoch": 0.615531595468814,
+      "grad_norm": 0.23226206004619598,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6156680769755698,
+      "grad_norm": 0.21327665448188782,
+      "learning_rate": 8e-05,
+      "loss": 1.3976,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6158045584823256,
+      "grad_norm": 0.22597189247608185,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6159410399890815,
+      "grad_norm": 0.21529574692249298,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6160775214958373,
+      "grad_norm": 0.2108212113380432,
+      "learning_rate": 8e-05,
+      "loss": 1.4375,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6162140030025931,
+      "grad_norm": 0.23344916105270386,
+      "learning_rate": 8e-05,
+      "loss": 1.5552,
+      "step": 4515
+    },
+    {
+      "epoch": 0.6163504845093489,
+      "grad_norm": 0.221527561545372,
+      "learning_rate": 8e-05,
+      "loss": 1.496,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6164869660161049,
+      "grad_norm": 0.2471136599779129,
+      "learning_rate": 8e-05,
+      "loss": 1.5501,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6166234475228607,
+      "grad_norm": 0.22684770822525024,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6167599290296165,
+      "grad_norm": 0.2155984789133072,
+      "learning_rate": 8e-05,
+      "loss": 1.4549,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6168964105363723,
+      "grad_norm": 0.20664231479167938,
+      "learning_rate": 8e-05,
+      "loss": 1.4339,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6170328920431282,
+      "grad_norm": 0.21298521757125854,
+      "learning_rate": 8e-05,
+      "loss": 1.3996,
+      "step": 4521
+    },
+    {
+      "epoch": 0.617169373549884,
+      "grad_norm": 0.21821504831314087,
+      "learning_rate": 8e-05,
+      "loss": 1.5086,
+      "step": 4522
+    },
+    {
+      "epoch": 0.6173058550566398,
+      "grad_norm": 0.2191344052553177,
+      "learning_rate": 8e-05,
+      "loss": 1.443,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6174423365633956,
+      "grad_norm": 0.2219267189502716,
+      "learning_rate": 8e-05,
+      "loss": 1.4185,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6175788180701515,
+      "grad_norm": 0.23031525313854218,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 4525
+    },
+    {
+      "epoch": 0.6177152995769073,
+      "grad_norm": 0.2170715481042862,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 4526
+    },
+    {
+      "epoch": 0.6178517810836631,
+      "grad_norm": 0.22577206790447235,
+      "learning_rate": 8e-05,
+      "loss": 1.5006,
+      "step": 4527
+    },
+    {
+      "epoch": 0.617988262590419,
+      "grad_norm": 0.21724019944667816,
+      "learning_rate": 8e-05,
+      "loss": 1.4165,
+      "step": 4528
+    },
+    {
+      "epoch": 0.6181247440971749,
+      "grad_norm": 0.2155725508928299,
+      "learning_rate": 8e-05,
+      "loss": 1.4624,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6182612256039307,
+      "grad_norm": 0.2135869413614273,
+      "learning_rate": 8e-05,
+      "loss": 1.5414,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6183977071106865,
+      "grad_norm": 0.22273477911949158,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6185341886174424,
+      "grad_norm": 0.22739556431770325,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6186706701241982,
+      "grad_norm": 0.2174486368894577,
+      "learning_rate": 8e-05,
+      "loss": 1.3822,
+      "step": 4533
+    },
+    {
+      "epoch": 0.618807151630954,
+      "grad_norm": 0.21750394999980927,
+      "learning_rate": 8e-05,
+      "loss": 1.4679,
+      "step": 4534
+    },
+    {
+      "epoch": 0.6189436331377098,
+      "grad_norm": 0.21488697826862335,
+      "learning_rate": 8e-05,
+      "loss": 1.3563,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6190801146444657,
+      "grad_norm": 0.2237270623445511,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6192165961512215,
+      "grad_norm": 0.22856594622135162,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6193530776579773,
+      "grad_norm": 0.211374893784523,
+      "learning_rate": 8e-05,
+      "loss": 1.3905,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6194895591647331,
+      "grad_norm": 0.21623264253139496,
+      "learning_rate": 8e-05,
+      "loss": 1.5292,
+      "step": 4539
+    },
+    {
+      "epoch": 0.619626040671489,
+      "grad_norm": 0.22048892080783844,
+      "learning_rate": 8e-05,
+      "loss": 1.4539,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6197625221782449,
+      "grad_norm": 0.2201317548751831,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6198990036850007,
+      "grad_norm": 0.22852268815040588,
+      "learning_rate": 8e-05,
+      "loss": 1.5251,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6200354851917566,
+      "grad_norm": 0.2265300750732422,
+      "learning_rate": 8e-05,
+      "loss": 1.4819,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6201719666985124,
+      "grad_norm": 0.22713707387447357,
+      "learning_rate": 8e-05,
+      "loss": 1.5013,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6203084482052682,
+      "grad_norm": 0.21535217761993408,
+      "learning_rate": 8e-05,
+      "loss": 1.4675,
+      "step": 4545
+    },
+    {
+      "epoch": 0.620444929712024,
+      "grad_norm": 0.21663063764572144,
+      "learning_rate": 8e-05,
+      "loss": 1.4429,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6205814112187799,
+      "grad_norm": 0.22075310349464417,
+      "learning_rate": 8e-05,
+      "loss": 1.482,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6207178927255357,
+      "grad_norm": 0.2360958307981491,
+      "learning_rate": 8e-05,
+      "loss": 1.4974,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6208543742322915,
+      "grad_norm": 0.22052042186260223,
+      "learning_rate": 8e-05,
+      "loss": 1.5452,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6209908557390473,
+      "grad_norm": 0.2116163820028305,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6211273372458032,
+      "grad_norm": 0.2235938012599945,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 4551
+    },
+    {
+      "epoch": 0.621263818752559,
+      "grad_norm": 0.21472059190273285,
+      "learning_rate": 8e-05,
+      "loss": 1.4508,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6214003002593148,
+      "grad_norm": 0.2278815060853958,
+      "learning_rate": 8e-05,
+      "loss": 1.5019,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6215367817660707,
+      "grad_norm": 0.20541225373744965,
+      "learning_rate": 8e-05,
+      "loss": 1.4077,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6216732632728266,
+      "grad_norm": 0.22905808687210083,
+      "learning_rate": 8e-05,
+      "loss": 1.4679,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6218097447795824,
+      "grad_norm": 0.2160603553056717,
+      "learning_rate": 8e-05,
+      "loss": 1.469,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6219462262863382,
+      "grad_norm": 0.22439983487129211,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 4557
+    },
+    {
+      "epoch": 0.622082707793094,
+      "grad_norm": 0.22758635878562927,
+      "learning_rate": 8e-05,
+      "loss": 1.4272,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6222191892998499,
+      "grad_norm": 0.2183096706867218,
+      "learning_rate": 8e-05,
+      "loss": 1.406,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6223556708066057,
+      "grad_norm": 0.22571112215518951,
+      "learning_rate": 8e-05,
+      "loss": 1.471,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6224921523133615,
+      "grad_norm": 0.23306390643119812,
+      "learning_rate": 8e-05,
+      "loss": 1.5578,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6226286338201173,
+      "grad_norm": 0.22306998074054718,
+      "learning_rate": 8e-05,
+      "loss": 1.4923,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6227651153268732,
+      "grad_norm": 0.2155541181564331,
+      "learning_rate": 8e-05,
+      "loss": 1.5457,
+      "step": 4563
+    },
+    {
+      "epoch": 0.622901596833629,
+      "grad_norm": 0.22404338419437408,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6230380783403848,
+      "grad_norm": 0.21983791887760162,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6231745598471408,
+      "grad_norm": 0.22456729412078857,
+      "learning_rate": 8e-05,
+      "loss": 1.4743,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6233110413538966,
+      "grad_norm": 0.23818491399288177,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6234475228606524,
+      "grad_norm": 0.22465217113494873,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6235840043674082,
+      "grad_norm": 0.22717180848121643,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6237204858741641,
+      "grad_norm": 0.21714508533477783,
+      "learning_rate": 8e-05,
+      "loss": 1.4591,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6238569673809199,
+      "grad_norm": 0.22139602899551392,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 4571
+    },
+    {
+      "epoch": 0.6239934488876757,
+      "grad_norm": 0.2278294414281845,
+      "learning_rate": 8e-05,
+      "loss": 1.5099,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6241299303944315,
+      "grad_norm": 0.21526354551315308,
+      "learning_rate": 8e-05,
+      "loss": 1.4982,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6242664119011874,
+      "grad_norm": 0.22482097148895264,
+      "learning_rate": 8e-05,
+      "loss": 1.5189,
+      "step": 4574
+    },
+    {
+      "epoch": 0.6244028934079432,
+      "grad_norm": 0.22274430096149445,
+      "learning_rate": 8e-05,
+      "loss": 1.3734,
+      "step": 4575
+    },
+    {
+      "epoch": 0.624539374914699,
+      "grad_norm": 0.22055159509181976,
+      "learning_rate": 8e-05,
+      "loss": 1.3901,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6246758564214548,
+      "grad_norm": 0.223702535033226,
+      "learning_rate": 8e-05,
+      "loss": 1.4506,
+      "step": 4577
+    },
+    {
+      "epoch": 0.6248123379282108,
+      "grad_norm": 0.22523395717144012,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6249488194349666,
+      "grad_norm": 0.2230580747127533,
+      "learning_rate": 8e-05,
+      "loss": 1.428,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6250853009417224,
+      "grad_norm": 0.21592172980308533,
+      "learning_rate": 8e-05,
+      "loss": 1.4482,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6252217824484783,
+      "grad_norm": 0.21322357654571533,
+      "learning_rate": 8e-05,
+      "loss": 1.4407,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6253582639552341,
+      "grad_norm": 0.2294072061777115,
+      "learning_rate": 8e-05,
+      "loss": 1.4537,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6254947454619899,
+      "grad_norm": 0.2150132954120636,
+      "learning_rate": 8e-05,
+      "loss": 1.4684,
+      "step": 4583
+    },
+    {
+      "epoch": 0.6256312269687457,
+      "grad_norm": 0.21548451483249664,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6257677084755016,
+      "grad_norm": 0.2210894227027893,
+      "learning_rate": 8e-05,
+      "loss": 1.4739,
+      "step": 4585
+    },
+    {
+      "epoch": 0.6259041899822574,
+      "grad_norm": 0.23256449401378632,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6260406714890132,
+      "grad_norm": 0.2129555195569992,
+      "learning_rate": 8e-05,
+      "loss": 1.4096,
+      "step": 4587
+    },
+    {
+      "epoch": 0.626177152995769,
+      "grad_norm": 0.22758153080940247,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6263136345025249,
+      "grad_norm": 0.21438907086849213,
+      "learning_rate": 8e-05,
+      "loss": 1.4614,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6264501160092807,
+      "grad_norm": 0.2170381247997284,
+      "learning_rate": 8e-05,
+      "loss": 1.3728,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6265865975160366,
+      "grad_norm": 0.2122417390346527,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 4591
+    },
+    {
+      "epoch": 0.6267230790227925,
+      "grad_norm": 0.2286798506975174,
+      "learning_rate": 8e-05,
+      "loss": 1.4809,
+      "step": 4592
+    },
+    {
+      "epoch": 0.6268595605295483,
+      "grad_norm": 0.22717009484767914,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6269960420363041,
+      "grad_norm": 0.22560113668441772,
+      "learning_rate": 8e-05,
+      "loss": 1.4603,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6271325235430599,
+      "grad_norm": 0.22476686537265778,
+      "learning_rate": 8e-05,
+      "loss": 1.4726,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6272690050498158,
+      "grad_norm": 0.21955826878547668,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 4596
+    },
+    {
+      "epoch": 0.6274054865565716,
+      "grad_norm": 0.22009466588497162,
+      "learning_rate": 8e-05,
+      "loss": 1.4766,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6275419680633274,
+      "grad_norm": 0.22455346584320068,
+      "learning_rate": 8e-05,
+      "loss": 1.4874,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6276784495700832,
+      "grad_norm": 0.219157412648201,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 4599
+    },
+    {
+      "epoch": 0.6278149310768391,
+      "grad_norm": 0.2223764955997467,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6279514125835949,
+      "grad_norm": 0.2246936410665512,
+      "learning_rate": 8e-05,
+      "loss": 1.454,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6280878940903507,
+      "grad_norm": 0.21355292201042175,
+      "learning_rate": 8e-05,
+      "loss": 1.4093,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6282243755971066,
+      "grad_norm": 0.2190273404121399,
+      "learning_rate": 8e-05,
+      "loss": 1.4644,
+      "step": 4603
+    },
+    {
+      "epoch": 0.6283608571038625,
+      "grad_norm": 0.2212258130311966,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6284973386106183,
+      "grad_norm": 0.22286319732666016,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6286338201173741,
+      "grad_norm": 0.22692331671714783,
+      "learning_rate": 8e-05,
+      "loss": 1.3897,
+      "step": 4606
+    },
+    {
+      "epoch": 0.62877030162413,
+      "grad_norm": 0.21866215765476227,
+      "learning_rate": 8e-05,
+      "loss": 1.4308,
+      "step": 4607
+    },
+    {
+      "epoch": 0.6289067831308858,
+      "grad_norm": 0.2197587639093399,
+      "learning_rate": 8e-05,
+      "loss": 1.5291,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6290432646376416,
+      "grad_norm": 0.22517380118370056,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 4609
+    },
+    {
+      "epoch": 0.6291797461443974,
+      "grad_norm": 0.21190588176250458,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6293162276511532,
+      "grad_norm": 0.22436542809009552,
+      "learning_rate": 8e-05,
+      "loss": 1.445,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6294527091579091,
+      "grad_norm": 0.22565437853336334,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 4612
+    },
+    {
+      "epoch": 0.6295891906646649,
+      "grad_norm": 0.21263591945171356,
+      "learning_rate": 8e-05,
+      "loss": 1.4412,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6297256721714207,
+      "grad_norm": 0.2213335782289505,
+      "learning_rate": 8e-05,
+      "loss": 1.4525,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6298621536781767,
+      "grad_norm": 0.22224274277687073,
+      "learning_rate": 8e-05,
+      "loss": 1.5425,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6299986351849325,
+      "grad_norm": 0.22343961894512177,
+      "learning_rate": 8e-05,
+      "loss": 1.5104,
+      "step": 4616
+    },
+    {
+      "epoch": 0.6301351166916883,
+      "grad_norm": 0.22053085267543793,
+      "learning_rate": 8e-05,
+      "loss": 1.4877,
+      "step": 4617
+    },
+    {
+      "epoch": 0.6302715981984441,
+      "grad_norm": 0.2179500311613083,
+      "learning_rate": 8e-05,
+      "loss": 1.4672,
+      "step": 4618
+    },
+    {
+      "epoch": 0.6304080797052,
+      "grad_norm": 0.22118046879768372,
+      "learning_rate": 8e-05,
+      "loss": 1.5334,
+      "step": 4619
+    },
+    {
+      "epoch": 0.6305445612119558,
+      "grad_norm": 0.2125585526227951,
+      "learning_rate": 8e-05,
+      "loss": 1.4509,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6306810427187116,
+      "grad_norm": 0.22199584543704987,
+      "learning_rate": 8e-05,
+      "loss": 1.4671,
+      "step": 4621
+    },
+    {
+      "epoch": 0.6308175242254674,
+      "grad_norm": 0.22178277373313904,
+      "learning_rate": 8e-05,
+      "loss": 1.4565,
+      "step": 4622
+    },
+    {
+      "epoch": 0.6309540057322233,
+      "grad_norm": 0.22410519421100616,
+      "learning_rate": 8e-05,
+      "loss": 1.3981,
+      "step": 4623
+    },
+    {
+      "epoch": 0.6310904872389791,
+      "grad_norm": 0.2181062549352646,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 4624
+    },
+    {
+      "epoch": 0.6312269687457349,
+      "grad_norm": 0.2156004160642624,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 4625
+    },
+    {
+      "epoch": 0.6313634502524907,
+      "grad_norm": 0.22380991280078888,
+      "learning_rate": 8e-05,
+      "loss": 1.4649,
+      "step": 4626
+    },
+    {
+      "epoch": 0.6314999317592466,
+      "grad_norm": 0.21817751228809357,
+      "learning_rate": 8e-05,
+      "loss": 1.4697,
+      "step": 4627
+    },
+    {
+      "epoch": 0.6316364132660025,
+      "grad_norm": 0.21054239571094513,
+      "learning_rate": 8e-05,
+      "loss": 1.4228,
+      "step": 4628
+    },
+    {
+      "epoch": 0.6317728947727583,
+      "grad_norm": 0.2192118763923645,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 4629
+    },
+    {
+      "epoch": 0.6319093762795142,
+      "grad_norm": 0.2229197472333908,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 4630
+    },
+    {
+      "epoch": 0.63204585778627,
+      "grad_norm": 0.2127682864665985,
+      "learning_rate": 8e-05,
+      "loss": 1.4435,
+      "step": 4631
+    },
+    {
+      "epoch": 0.6321823392930258,
+      "grad_norm": 0.22397445142269135,
+      "learning_rate": 8e-05,
+      "loss": 1.441,
+      "step": 4632
+    },
+    {
+      "epoch": 0.6323188207997816,
+      "grad_norm": 0.20974363386631012,
+      "learning_rate": 8e-05,
+      "loss": 1.4048,
+      "step": 4633
+    },
+    {
+      "epoch": 0.6324553023065375,
+      "grad_norm": 0.22104448080062866,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 4634
+    },
+    {
+      "epoch": 0.6325917838132933,
+      "grad_norm": 0.21387861669063568,
+      "learning_rate": 8e-05,
+      "loss": 1.4287,
+      "step": 4635
+    },
+    {
+      "epoch": 0.6327282653200491,
+      "grad_norm": 0.21601299941539764,
+      "learning_rate": 8e-05,
+      "loss": 1.4591,
+      "step": 4636
+    },
+    {
+      "epoch": 0.6328647468268049,
+      "grad_norm": 0.2263786792755127,
+      "learning_rate": 8e-05,
+      "loss": 1.4821,
+      "step": 4637
+    },
+    {
+      "epoch": 0.6330012283335608,
+      "grad_norm": 0.22038711607456207,
+      "learning_rate": 8e-05,
+      "loss": 1.4461,
+      "step": 4638
+    },
+    {
+      "epoch": 0.6331377098403166,
+      "grad_norm": 0.22404839098453522,
+      "learning_rate": 8e-05,
+      "loss": 1.4498,
+      "step": 4639
+    },
+    {
+      "epoch": 0.6332741913470725,
+      "grad_norm": 0.21835957467556,
+      "learning_rate": 8e-05,
+      "loss": 1.4563,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6334106728538283,
+      "grad_norm": 0.21802298724651337,
+      "learning_rate": 8e-05,
+      "loss": 1.4689,
+      "step": 4641
+    },
+    {
+      "epoch": 0.6335471543605842,
+      "grad_norm": 0.22251929342746735,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 4642
+    },
+    {
+      "epoch": 0.63368363586734,
+      "grad_norm": 0.218058243393898,
+      "learning_rate": 8e-05,
+      "loss": 1.4409,
+      "step": 4643
+    },
+    {
+      "epoch": 0.6338201173740958,
+      "grad_norm": 0.21673572063446045,
+      "learning_rate": 8e-05,
+      "loss": 1.3803,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6339565988808517,
+      "grad_norm": 0.21810711920261383,
+      "learning_rate": 8e-05,
+      "loss": 1.453,
+      "step": 4645
+    },
+    {
+      "epoch": 0.6340930803876075,
+      "grad_norm": 0.22305771708488464,
+      "learning_rate": 8e-05,
+      "loss": 1.4995,
+      "step": 4646
+    },
+    {
+      "epoch": 0.6342295618943633,
+      "grad_norm": 0.22756484150886536,
+      "learning_rate": 8e-05,
+      "loss": 1.5225,
+      "step": 4647
+    },
+    {
+      "epoch": 0.6343660434011191,
+      "grad_norm": 0.2284999042749405,
+      "learning_rate": 8e-05,
+      "loss": 1.436,
+      "step": 4648
+    },
+    {
+      "epoch": 0.634502524907875,
+      "grad_norm": 0.2279583066701889,
+      "learning_rate": 8e-05,
+      "loss": 1.4646,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6346390064146308,
+      "grad_norm": 0.22005701065063477,
+      "learning_rate": 8e-05,
+      "loss": 1.3923,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6347754879213866,
+      "grad_norm": 0.20984189212322235,
+      "learning_rate": 8e-05,
+      "loss": 1.3929,
+      "step": 4651
+    },
+    {
+      "epoch": 0.6349119694281425,
+      "grad_norm": 0.22153371572494507,
+      "learning_rate": 8e-05,
+      "loss": 1.4896,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6350484509348984,
+      "grad_norm": 0.22544576227664948,
+      "learning_rate": 8e-05,
+      "loss": 1.4896,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6351849324416542,
+      "grad_norm": 0.22344543039798737,
+      "learning_rate": 8e-05,
+      "loss": 1.5057,
+      "step": 4654
+    },
+    {
+      "epoch": 0.63532141394841,
+      "grad_norm": 0.22263257205486298,
+      "learning_rate": 8e-05,
+      "loss": 1.4561,
+      "step": 4655
+    },
+    {
+      "epoch": 0.6354578954551658,
+      "grad_norm": 0.22542722523212433,
+      "learning_rate": 8e-05,
+      "loss": 1.4506,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6355943769619217,
+      "grad_norm": 0.22193047404289246,
+      "learning_rate": 8e-05,
+      "loss": 1.473,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6357308584686775,
+      "grad_norm": 0.22100095450878143,
+      "learning_rate": 8e-05,
+      "loss": 1.4293,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6358673399754333,
+      "grad_norm": 0.21931973099708557,
+      "learning_rate": 8e-05,
+      "loss": 1.4418,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6360038214821891,
+      "grad_norm": 0.2258632481098175,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 4660
+    },
+    {
+      "epoch": 0.636140302988945,
+      "grad_norm": 0.22061054408550262,
+      "learning_rate": 8e-05,
+      "loss": 1.4211,
+      "step": 4661
+    },
+    {
+      "epoch": 0.6362767844957008,
+      "grad_norm": 0.21946576237678528,
+      "learning_rate": 8e-05,
+      "loss": 1.4316,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6364132660024566,
+      "grad_norm": 0.23044532537460327,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6365497475092124,
+      "grad_norm": 0.23592929542064667,
+      "learning_rate": 8e-05,
+      "loss": 1.4532,
+      "step": 4664
+    },
+    {
+      "epoch": 0.6366862290159684,
+      "grad_norm": 0.21670588850975037,
+      "learning_rate": 8e-05,
+      "loss": 1.4584,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6368227105227242,
+      "grad_norm": 0.22482140362262726,
+      "learning_rate": 8e-05,
+      "loss": 1.519,
+      "step": 4666
+    },
+    {
+      "epoch": 0.63695919202948,
+      "grad_norm": 0.22188323736190796,
+      "learning_rate": 8e-05,
+      "loss": 1.5304,
+      "step": 4667
+    },
+    {
+      "epoch": 0.6370956735362359,
+      "grad_norm": 0.2268882840871811,
+      "learning_rate": 8e-05,
+      "loss": 1.4572,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6372321550429917,
+      "grad_norm": 0.21680812537670135,
+      "learning_rate": 8e-05,
+      "loss": 1.4597,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6373686365497475,
+      "grad_norm": 0.23040521144866943,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6375051180565033,
+      "grad_norm": 0.22414684295654297,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6376415995632592,
+      "grad_norm": 0.2309904545545578,
+      "learning_rate": 8e-05,
+      "loss": 1.4142,
+      "step": 4672
+    },
+    {
+      "epoch": 0.637778081070015,
+      "grad_norm": 0.23031876981258392,
+      "learning_rate": 8e-05,
+      "loss": 1.5281,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6379145625767708,
+      "grad_norm": 0.22511087357997894,
+      "learning_rate": 8e-05,
+      "loss": 1.466,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6380510440835266,
+      "grad_norm": 0.21712131798267365,
+      "learning_rate": 8e-05,
+      "loss": 1.4357,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6381875255902825,
+      "grad_norm": 0.2201714664697647,
+      "learning_rate": 8e-05,
+      "loss": 1.4442,
+      "step": 4676
+    },
+    {
+      "epoch": 0.6383240070970384,
+      "grad_norm": 0.21700957417488098,
+      "learning_rate": 8e-05,
+      "loss": 1.4427,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6384604886037942,
+      "grad_norm": 0.2210516780614853,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 4678
+    },
+    {
+      "epoch": 0.63859697011055,
+      "grad_norm": 0.21742641925811768,
+      "learning_rate": 8e-05,
+      "loss": 1.3438,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6387334516173059,
+      "grad_norm": 0.23449508845806122,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6388699331240617,
+      "grad_norm": 0.22773809731006622,
+      "learning_rate": 8e-05,
+      "loss": 1.4779,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6390064146308175,
+      "grad_norm": 0.2365875095129013,
+      "learning_rate": 8e-05,
+      "loss": 1.4438,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6391428961375734,
+      "grad_norm": 0.21995630860328674,
+      "learning_rate": 8e-05,
+      "loss": 1.4766,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6392793776443292,
+      "grad_norm": 0.2281271070241928,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 4684
+    },
+    {
+      "epoch": 0.639415859151085,
+      "grad_norm": 0.22183838486671448,
+      "learning_rate": 8e-05,
+      "loss": 1.4539,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6395523406578408,
+      "grad_norm": 0.21777133643627167,
+      "learning_rate": 8e-05,
+      "loss": 1.4944,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6396888221645967,
+      "grad_norm": 0.222051739692688,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 4687
+    },
+    {
+      "epoch": 0.6398253036713525,
+      "grad_norm": 0.2328973412513733,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6399617851781083,
+      "grad_norm": 0.22004377841949463,
+      "learning_rate": 8e-05,
+      "loss": 1.5062,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6400982666848642,
+      "grad_norm": 0.23322167992591858,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6402347481916201,
+      "grad_norm": 0.22014138102531433,
+      "learning_rate": 8e-05,
+      "loss": 1.5008,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6403712296983759,
+      "grad_norm": 0.22135215997695923,
+      "learning_rate": 8e-05,
+      "loss": 1.4516,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6405077112051317,
+      "grad_norm": 0.22722910344600677,
+      "learning_rate": 8e-05,
+      "loss": 1.4994,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6406441927118876,
+      "grad_norm": 0.2225387543439865,
+      "learning_rate": 8e-05,
+      "loss": 1.487,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6407806742186434,
+      "grad_norm": 0.2370699942111969,
+      "learning_rate": 8e-05,
+      "loss": 1.5507,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6409171557253992,
+      "grad_norm": 0.23134349286556244,
+      "learning_rate": 8e-05,
+      "loss": 1.4559,
+      "step": 4696
+    },
+    {
+      "epoch": 0.641053637232155,
+      "grad_norm": 0.21570952236652374,
+      "learning_rate": 8e-05,
+      "loss": 1.4273,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6411901187389109,
+      "grad_norm": 0.2156054973602295,
+      "learning_rate": 8e-05,
+      "loss": 1.4612,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6413266002456667,
+      "grad_norm": 0.23077009618282318,
+      "learning_rate": 8e-05,
+      "loss": 1.5153,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6414630817524225,
+      "grad_norm": 0.2185460478067398,
+      "learning_rate": 8e-05,
+      "loss": 1.4978,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6415995632591783,
+      "grad_norm": 0.22641827166080475,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6417360447659343,
+      "grad_norm": 0.21381402015686035,
+      "learning_rate": 8e-05,
+      "loss": 1.438,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6418725262726901,
+      "grad_norm": 0.21371695399284363,
+      "learning_rate": 8e-05,
+      "loss": 1.4763,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6420090077794459,
+      "grad_norm": 0.21719419956207275,
+      "learning_rate": 8e-05,
+      "loss": 1.4404,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6421454892862017,
+      "grad_norm": 0.22064515948295593,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6422819707929576,
+      "grad_norm": 0.2249942421913147,
+      "learning_rate": 8e-05,
+      "loss": 1.4689,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6424184522997134,
+      "grad_norm": 0.22294658422470093,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6425549338064692,
+      "grad_norm": 0.21578864753246307,
+      "learning_rate": 8e-05,
+      "loss": 1.4742,
+      "step": 4708
+    },
+    {
+      "epoch": 0.642691415313225,
+      "grad_norm": 0.22582624852657318,
+      "learning_rate": 8e-05,
+      "loss": 1.4423,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6428278968199809,
+      "grad_norm": 0.22269436717033386,
+      "learning_rate": 8e-05,
+      "loss": 1.5331,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6429643783267367,
+      "grad_norm": 0.21985948085784912,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6431008598334925,
+      "grad_norm": 0.21801301836967468,
+      "learning_rate": 8e-05,
+      "loss": 1.4797,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6432373413402483,
+      "grad_norm": 0.22284044325351715,
+      "learning_rate": 8e-05,
+      "loss": 1.4297,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6433738228470043,
+      "grad_norm": 0.21237695217132568,
+      "learning_rate": 8e-05,
+      "loss": 1.4998,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6435103043537601,
+      "grad_norm": 0.22127491235733032,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6436467858605159,
+      "grad_norm": 0.2248886674642563,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6437832673672718,
+      "grad_norm": 0.2306990623474121,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6439197488740276,
+      "grad_norm": 0.2188940793275833,
+      "learning_rate": 8e-05,
+      "loss": 1.4716,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6440562303807834,
+      "grad_norm": 0.22470921277999878,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6441927118875392,
+      "grad_norm": 0.21132726967334747,
+      "learning_rate": 8e-05,
+      "loss": 1.4207,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6443291933942951,
+      "grad_norm": 0.222573921084404,
+      "learning_rate": 8e-05,
+      "loss": 1.4544,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6444656749010509,
+      "grad_norm": 0.22469067573547363,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6446021564078067,
+      "grad_norm": 0.21866431832313538,
+      "learning_rate": 8e-05,
+      "loss": 1.3667,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6447386379145625,
+      "grad_norm": 0.22790862619876862,
+      "learning_rate": 8e-05,
+      "loss": 1.3955,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6448751194213184,
+      "grad_norm": 0.22355790436267853,
+      "learning_rate": 8e-05,
+      "loss": 1.4696,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6450116009280742,
+      "grad_norm": 0.22280742228031158,
+      "learning_rate": 8e-05,
+      "loss": 1.467,
+      "step": 4726
+    },
+    {
+      "epoch": 0.6451480824348301,
+      "grad_norm": 0.22323842346668243,
+      "learning_rate": 8e-05,
+      "loss": 1.4484,
+      "step": 4727
+    },
+    {
+      "epoch": 0.645284563941586,
+      "grad_norm": 0.22016021609306335,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6454210454483418,
+      "grad_norm": 0.21709200739860535,
+      "learning_rate": 8e-05,
+      "loss": 1.462,
+      "step": 4729
+    },
+    {
+      "epoch": 0.6455575269550976,
+      "grad_norm": 0.22512587904930115,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6456940084618534,
+      "grad_norm": 0.22144664824008942,
+      "learning_rate": 8e-05,
+      "loss": 1.4605,
+      "step": 4731
+    },
+    {
+      "epoch": 0.6458304899686093,
+      "grad_norm": 0.22873267531394958,
+      "learning_rate": 8e-05,
+      "loss": 1.5313,
+      "step": 4732
+    },
+    {
+      "epoch": 0.6459669714753651,
+      "grad_norm": 0.227690652012825,
+      "learning_rate": 8e-05,
+      "loss": 1.478,
+      "step": 4733
+    },
+    {
+      "epoch": 0.6461034529821209,
+      "grad_norm": 0.22937248647212982,
+      "learning_rate": 8e-05,
+      "loss": 1.4923,
+      "step": 4734
+    },
+    {
+      "epoch": 0.6462399344888767,
+      "grad_norm": 0.22020775079727173,
+      "learning_rate": 8e-05,
+      "loss": 1.4465,
+      "step": 4735
+    },
+    {
+      "epoch": 0.6463764159956326,
+      "grad_norm": 0.2242947220802307,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 4736
+    },
+    {
+      "epoch": 0.6465128975023884,
+      "grad_norm": 0.2241961508989334,
+      "learning_rate": 8e-05,
+      "loss": 1.4654,
+      "step": 4737
+    },
+    {
+      "epoch": 0.6466493790091442,
+      "grad_norm": 0.21888315677642822,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 4738
+    },
+    {
+      "epoch": 0.6467858605159001,
+      "grad_norm": 0.2144775241613388,
+      "learning_rate": 8e-05,
+      "loss": 1.4598,
+      "step": 4739
+    },
+    {
+      "epoch": 0.646922342022656,
+      "grad_norm": 0.22623242437839508,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 0.21717987954616547,
+      "learning_rate": 8e-05,
+      "loss": 1.4793,
+      "step": 4741
+    },
+    {
+      "epoch": 0.6471953050361676,
+      "grad_norm": 0.24406373500823975,
+      "learning_rate": 8e-05,
+      "loss": 1.526,
+      "step": 4742
+    },
+    {
+      "epoch": 0.6473317865429234,
+      "grad_norm": 0.22075706720352173,
+      "learning_rate": 8e-05,
+      "loss": 1.4362,
+      "step": 4743
+    },
+    {
+      "epoch": 0.6474682680496793,
+      "grad_norm": 0.22959192097187042,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 4744
+    },
+    {
+      "epoch": 0.6476047495564351,
+      "grad_norm": 0.22353392839431763,
+      "learning_rate": 8e-05,
+      "loss": 1.4983,
+      "step": 4745
+    },
+    {
+      "epoch": 0.6477412310631909,
+      "grad_norm": 0.23261985182762146,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 4746
+    },
+    {
+      "epoch": 0.6478777125699468,
+      "grad_norm": 0.24176758527755737,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 4747
+    },
+    {
+      "epoch": 0.6480141940767026,
+      "grad_norm": 0.22109319269657135,
+      "learning_rate": 8e-05,
+      "loss": 1.479,
+      "step": 4748
+    },
+    {
+      "epoch": 0.6481506755834584,
+      "grad_norm": 0.23933464288711548,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 4749
+    },
+    {
+      "epoch": 0.6482871570902142,
+      "grad_norm": 0.2160293161869049,
+      "learning_rate": 8e-05,
+      "loss": 1.4483,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6484236385969702,
+      "grad_norm": 0.22686642408370972,
+      "learning_rate": 8e-05,
+      "loss": 1.4975,
+      "step": 4751
+    },
+    {
+      "epoch": 0.648560120103726,
+      "grad_norm": 0.21621020138263702,
+      "learning_rate": 8e-05,
+      "loss": 1.4265,
+      "step": 4752
+    },
+    {
+      "epoch": 0.6486966016104818,
+      "grad_norm": 0.23167496919631958,
+      "learning_rate": 8e-05,
+      "loss": 1.5256,
+      "step": 4753
+    },
+    {
+      "epoch": 0.6488330831172376,
+      "grad_norm": 0.22529476881027222,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 4754
+    },
+    {
+      "epoch": 0.6489695646239935,
+      "grad_norm": 0.2205408811569214,
+      "learning_rate": 8e-05,
+      "loss": 1.4231,
+      "step": 4755
+    },
+    {
+      "epoch": 0.6491060461307493,
+      "grad_norm": 0.22217302024364471,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 4756
+    },
+    {
+      "epoch": 0.6492425276375051,
+      "grad_norm": 0.22462791204452515,
+      "learning_rate": 8e-05,
+      "loss": 1.4787,
+      "step": 4757
+    },
+    {
+      "epoch": 0.6493790091442609,
+      "grad_norm": 0.23155753314495087,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 4758
+    },
+    {
+      "epoch": 0.6495154906510168,
+      "grad_norm": 0.22752583026885986,
+      "learning_rate": 8e-05,
+      "loss": 1.5061,
+      "step": 4759
+    },
+    {
+      "epoch": 0.6496519721577726,
+      "grad_norm": 0.2195422500371933,
+      "learning_rate": 8e-05,
+      "loss": 1.4383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6497884536645284,
+      "grad_norm": 0.21294036507606506,
+      "learning_rate": 8e-05,
+      "loss": 1.4774,
+      "step": 4761
+    },
+    {
+      "epoch": 0.6499249351712842,
+      "grad_norm": 0.21663227677345276,
+      "learning_rate": 8e-05,
+      "loss": 1.453,
+      "step": 4762
+    },
+    {
+      "epoch": 0.6500614166780401,
+      "grad_norm": 0.2227342128753662,
+      "learning_rate": 8e-05,
+      "loss": 1.4878,
+      "step": 4763
+    },
+    {
+      "epoch": 0.650197898184796,
+      "grad_norm": 0.23289898037910461,
+      "learning_rate": 8e-05,
+      "loss": 1.5197,
+      "step": 4764
+    },
+    {
+      "epoch": 0.6503343796915518,
+      "grad_norm": 0.22494392096996307,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 4765
+    },
+    {
+      "epoch": 0.6504708611983077,
+      "grad_norm": 0.23340477049350739,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 4766
+    },
+    {
+      "epoch": 0.6506073427050635,
+      "grad_norm": 0.21949252486228943,
+      "learning_rate": 8e-05,
+      "loss": 1.5001,
+      "step": 4767
+    },
+    {
+      "epoch": 0.6507438242118193,
+      "grad_norm": 0.2134321630001068,
+      "learning_rate": 8e-05,
+      "loss": 1.3509,
+      "step": 4768
+    },
+    {
+      "epoch": 0.6508803057185751,
+      "grad_norm": 0.2202940136194229,
+      "learning_rate": 8e-05,
+      "loss": 1.481,
+      "step": 4769
+    },
+    {
+      "epoch": 0.651016787225331,
+      "grad_norm": 0.22476086020469666,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6511532687320868,
+      "grad_norm": 0.22764909267425537,
+      "learning_rate": 8e-05,
+      "loss": 1.5429,
+      "step": 4771
+    },
+    {
+      "epoch": 0.6512897502388426,
+      "grad_norm": 0.22695961594581604,
+      "learning_rate": 8e-05,
+      "loss": 1.4858,
+      "step": 4772
+    },
+    {
+      "epoch": 0.6514262317455984,
+      "grad_norm": 0.22650068998336792,
+      "learning_rate": 8e-05,
+      "loss": 1.4072,
+      "step": 4773
+    },
+    {
+      "epoch": 0.6515627132523543,
+      "grad_norm": 0.21934522688388824,
+      "learning_rate": 8e-05,
+      "loss": 1.4634,
+      "step": 4774
+    },
+    {
+      "epoch": 0.6516991947591101,
+      "grad_norm": 0.21516390144824982,
+      "learning_rate": 8e-05,
+      "loss": 1.3712,
+      "step": 4775
+    },
+    {
+      "epoch": 0.651835676265866,
+      "grad_norm": 0.22440971434116364,
+      "learning_rate": 8e-05,
+      "loss": 1.4333,
+      "step": 4776
+    },
+    {
+      "epoch": 0.6519721577726219,
+      "grad_norm": 0.23376627266407013,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 4777
+    },
+    {
+      "epoch": 0.6521086392793777,
+      "grad_norm": 0.22055548429489136,
+      "learning_rate": 8e-05,
+      "loss": 1.4462,
+      "step": 4778
+    },
+    {
+      "epoch": 0.6522451207861335,
+      "grad_norm": 0.2215365320444107,
+      "learning_rate": 8e-05,
+      "loss": 1.4301,
+      "step": 4779
+    },
+    {
+      "epoch": 0.6523816022928893,
+      "grad_norm": 0.21756118535995483,
+      "learning_rate": 8e-05,
+      "loss": 1.4187,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6525180837996452,
+      "grad_norm": 0.2270948588848114,
+      "learning_rate": 8e-05,
+      "loss": 1.5665,
+      "step": 4781
+    },
+    {
+      "epoch": 0.652654565306401,
+      "grad_norm": 0.23348069190979004,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 4782
+    },
+    {
+      "epoch": 0.6527910468131568,
+      "grad_norm": 0.2224881798028946,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 4783
+    },
+    {
+      "epoch": 0.6529275283199126,
+      "grad_norm": 0.22335462272167206,
+      "learning_rate": 8e-05,
+      "loss": 1.4998,
+      "step": 4784
+    },
+    {
+      "epoch": 0.6530640098266685,
+      "grad_norm": 0.22532108426094055,
+      "learning_rate": 8e-05,
+      "loss": 1.4779,
+      "step": 4785
+    },
+    {
+      "epoch": 0.6532004913334243,
+      "grad_norm": 0.21094249188899994,
+      "learning_rate": 8e-05,
+      "loss": 1.4098,
+      "step": 4786
+    },
+    {
+      "epoch": 0.6533369728401801,
+      "grad_norm": 0.22443215548992157,
+      "learning_rate": 8e-05,
+      "loss": 1.4953,
+      "step": 4787
+    },
+    {
+      "epoch": 0.653473454346936,
+      "grad_norm": 0.21294505894184113,
+      "learning_rate": 8e-05,
+      "loss": 1.4089,
+      "step": 4788
+    },
+    {
+      "epoch": 0.6536099358536919,
+      "grad_norm": 0.22449451684951782,
+      "learning_rate": 8e-05,
+      "loss": 1.5627,
+      "step": 4789
+    },
+    {
+      "epoch": 0.6537464173604477,
+      "grad_norm": 0.22543059289455414,
+      "learning_rate": 8e-05,
+      "loss": 1.4943,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6538828988672035,
+      "grad_norm": 0.225270077586174,
+      "learning_rate": 8e-05,
+      "loss": 1.3823,
+      "step": 4791
+    },
+    {
+      "epoch": 0.6540193803739593,
+      "grad_norm": 0.223805770277977,
+      "learning_rate": 8e-05,
+      "loss": 1.4371,
+      "step": 4792
+    },
+    {
+      "epoch": 0.6541558618807152,
+      "grad_norm": 0.22868068516254425,
+      "learning_rate": 8e-05,
+      "loss": 1.5025,
+      "step": 4793
+    },
+    {
+      "epoch": 0.654292343387471,
+      "grad_norm": 0.21598564088344574,
+      "learning_rate": 8e-05,
+      "loss": 1.4061,
+      "step": 4794
+    },
+    {
+      "epoch": 0.6544288248942268,
+      "grad_norm": 0.21698512136936188,
+      "learning_rate": 8e-05,
+      "loss": 1.4803,
+      "step": 4795
+    },
+    {
+      "epoch": 0.6545653064009826,
+      "grad_norm": 0.216933935880661,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 4796
+    },
+    {
+      "epoch": 0.6547017879077385,
+      "grad_norm": 0.21854951977729797,
+      "learning_rate": 8e-05,
+      "loss": 1.4745,
+      "step": 4797
+    },
+    {
+      "epoch": 0.6548382694144943,
+      "grad_norm": 0.21450889110565186,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 4798
+    },
+    {
+      "epoch": 0.6549747509212501,
+      "grad_norm": 0.22637712955474854,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 4799
+    },
+    {
+      "epoch": 0.655111232428006,
+      "grad_norm": 0.22508609294891357,
+      "learning_rate": 8e-05,
+      "loss": 1.4222,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6552477139347619,
+      "grad_norm": 0.22846727073192596,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 4801
+    },
+    {
+      "epoch": 0.6553841954415177,
+      "grad_norm": 0.22929659485816956,
+      "learning_rate": 8e-05,
+      "loss": 1.4331,
+      "step": 4802
+    },
+    {
+      "epoch": 0.6555206769482735,
+      "grad_norm": 0.22304792702198029,
+      "learning_rate": 8e-05,
+      "loss": 1.4908,
+      "step": 4803
+    },
+    {
+      "epoch": 0.6556571584550294,
+      "grad_norm": 0.22466646134853363,
+      "learning_rate": 8e-05,
+      "loss": 1.4804,
+      "step": 4804
+    },
+    {
+      "epoch": 0.6557936399617852,
+      "grad_norm": 0.226219043135643,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 4805
+    },
+    {
+      "epoch": 0.655930121468541,
+      "grad_norm": 0.23656295239925385,
+      "learning_rate": 8e-05,
+      "loss": 1.5041,
+      "step": 4806
+    },
+    {
+      "epoch": 0.6560666029752968,
+      "grad_norm": 0.2190861701965332,
+      "learning_rate": 8e-05,
+      "loss": 1.4372,
+      "step": 4807
+    },
+    {
+      "epoch": 0.6562030844820527,
+      "grad_norm": 0.22630518674850464,
+      "learning_rate": 8e-05,
+      "loss": 1.4726,
+      "step": 4808
+    },
+    {
+      "epoch": 0.6563395659888085,
+      "grad_norm": 0.21447741985321045,
+      "learning_rate": 8e-05,
+      "loss": 1.4116,
+      "step": 4809
+    },
+    {
+      "epoch": 0.6564760474955643,
+      "grad_norm": 0.22013404965400696,
+      "learning_rate": 8e-05,
+      "loss": 1.445,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6566125290023201,
+      "grad_norm": 0.2186201810836792,
+      "learning_rate": 8e-05,
+      "loss": 1.4337,
+      "step": 4811
+    },
+    {
+      "epoch": 0.656749010509076,
+      "grad_norm": 0.23519742488861084,
+      "learning_rate": 8e-05,
+      "loss": 1.5156,
+      "step": 4812
+    },
+    {
+      "epoch": 0.6568854920158319,
+      "grad_norm": 0.21780787408351898,
+      "learning_rate": 8e-05,
+      "loss": 1.4833,
+      "step": 4813
+    },
+    {
+      "epoch": 0.6570219735225877,
+      "grad_norm": 0.2263040691614151,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 4814
+    },
+    {
+      "epoch": 0.6571584550293436,
+      "grad_norm": 0.2249605506658554,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 4815
+    },
+    {
+      "epoch": 0.6572949365360994,
+      "grad_norm": 0.22105370461940765,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 4816
+    },
+    {
+      "epoch": 0.6574314180428552,
+      "grad_norm": 0.2265872061252594,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 4817
+    },
+    {
+      "epoch": 0.657567899549611,
+      "grad_norm": 0.22143472731113434,
+      "learning_rate": 8e-05,
+      "loss": 1.4474,
+      "step": 4818
+    },
+    {
+      "epoch": 0.6577043810563669,
+      "grad_norm": 0.21892483532428741,
+      "learning_rate": 8e-05,
+      "loss": 1.4338,
+      "step": 4819
+    },
+    {
+      "epoch": 0.6578408625631227,
+      "grad_norm": 0.22374549508094788,
+      "learning_rate": 8e-05,
+      "loss": 1.4352,
+      "step": 4820
+    },
+    {
+      "epoch": 0.6579773440698785,
+      "grad_norm": 0.22548454999923706,
+      "learning_rate": 8e-05,
+      "loss": 1.4356,
+      "step": 4821
+    },
+    {
+      "epoch": 0.6581138255766343,
+      "grad_norm": 0.22463259100914001,
+      "learning_rate": 8e-05,
+      "loss": 1.4787,
+      "step": 4822
+    },
+    {
+      "epoch": 0.6582503070833902,
+      "grad_norm": 0.22149738669395447,
+      "learning_rate": 8e-05,
+      "loss": 1.462,
+      "step": 4823
+    },
+    {
+      "epoch": 0.658386788590146,
+      "grad_norm": 0.21920955181121826,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 4824
+    },
+    {
+      "epoch": 0.6585232700969019,
+      "grad_norm": 0.22516049444675446,
+      "learning_rate": 8e-05,
+      "loss": 1.4442,
+      "step": 4825
+    },
+    {
+      "epoch": 0.6586597516036578,
+      "grad_norm": 0.21563005447387695,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4826
+    },
+    {
+      "epoch": 0.6587962331104136,
+      "grad_norm": 0.22993803024291992,
+      "learning_rate": 8e-05,
+      "loss": 1.3885,
+      "step": 4827
+    },
+    {
+      "epoch": 0.6589327146171694,
+      "grad_norm": 0.24025046825408936,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 4828
+    },
+    {
+      "epoch": 0.6590691961239252,
+      "grad_norm": 0.22974734008312225,
+      "learning_rate": 8e-05,
+      "loss": 1.4857,
+      "step": 4829
+    },
+    {
+      "epoch": 0.659205677630681,
+      "grad_norm": 0.2239166498184204,
+      "learning_rate": 8e-05,
+      "loss": 1.4518,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6593421591374369,
+      "grad_norm": 0.22037450969219208,
+      "learning_rate": 8e-05,
+      "loss": 1.3848,
+      "step": 4831
+    },
+    {
+      "epoch": 0.6594786406441927,
+      "grad_norm": 0.22467266023159027,
+      "learning_rate": 8e-05,
+      "loss": 1.4436,
+      "step": 4832
+    },
+    {
+      "epoch": 0.6596151221509485,
+      "grad_norm": 0.22956858575344086,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 4833
+    },
+    {
+      "epoch": 0.6597516036577044,
+      "grad_norm": 0.23247022926807404,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 4834
+    },
+    {
+      "epoch": 0.6598880851644602,
+      "grad_norm": 0.2317829430103302,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 4835
+    },
+    {
+      "epoch": 0.660024566671216,
+      "grad_norm": 0.2160494178533554,
+      "learning_rate": 8e-05,
+      "loss": 1.3702,
+      "step": 4836
+    },
+    {
+      "epoch": 0.6601610481779718,
+      "grad_norm": 0.22596830129623413,
+      "learning_rate": 8e-05,
+      "loss": 1.4501,
+      "step": 4837
+    },
+    {
+      "epoch": 0.6602975296847278,
+      "grad_norm": 0.22802773118019104,
+      "learning_rate": 8e-05,
+      "loss": 1.4462,
+      "step": 4838
+    },
+    {
+      "epoch": 0.6604340111914836,
+      "grad_norm": 0.21835008263587952,
+      "learning_rate": 8e-05,
+      "loss": 1.4373,
+      "step": 4839
+    },
+    {
+      "epoch": 0.6605704926982394,
+      "grad_norm": 0.22111277282238007,
+      "learning_rate": 8e-05,
+      "loss": 1.4654,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6607069742049952,
+      "grad_norm": 0.22188541293144226,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 4841
+    },
+    {
+      "epoch": 0.6608434557117511,
+      "grad_norm": 0.22060491144657135,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 4842
+    },
+    {
+      "epoch": 0.6609799372185069,
+      "grad_norm": 0.22839945554733276,
+      "learning_rate": 8e-05,
+      "loss": 1.4606,
+      "step": 4843
+    },
+    {
+      "epoch": 0.6611164187252627,
+      "grad_norm": 0.22106820344924927,
+      "learning_rate": 8e-05,
+      "loss": 1.4646,
+      "step": 4844
+    },
+    {
+      "epoch": 0.6612529002320185,
+      "grad_norm": 0.22889405488967896,
+      "learning_rate": 8e-05,
+      "loss": 1.5057,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6613893817387744,
+      "grad_norm": 0.21413250267505646,
+      "learning_rate": 8e-05,
+      "loss": 1.4267,
+      "step": 4846
+    },
+    {
+      "epoch": 0.6615258632455302,
+      "grad_norm": 0.2106185257434845,
+      "learning_rate": 8e-05,
+      "loss": 1.3816,
+      "step": 4847
+    },
+    {
+      "epoch": 0.661662344752286,
+      "grad_norm": 0.22740787267684937,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 4848
+    },
+    {
+      "epoch": 0.6617988262590419,
+      "grad_norm": 0.22585178911685944,
+      "learning_rate": 8e-05,
+      "loss": 1.468,
+      "step": 4849
+    },
+    {
+      "epoch": 0.6619353077657978,
+      "grad_norm": 0.22686122357845306,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6620717892725536,
+      "grad_norm": 0.21957051753997803,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 4851
+    },
+    {
+      "epoch": 0.6622082707793094,
+      "grad_norm": 0.23281526565551758,
+      "learning_rate": 8e-05,
+      "loss": 1.5103,
+      "step": 4852
+    },
+    {
+      "epoch": 0.6623447522860653,
+      "grad_norm": 0.2223740816116333,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 4853
+    },
+    {
+      "epoch": 0.6624812337928211,
+      "grad_norm": 0.22596101462841034,
+      "learning_rate": 8e-05,
+      "loss": 1.476,
+      "step": 4854
+    },
+    {
+      "epoch": 0.6626177152995769,
+      "grad_norm": 0.22248028218746185,
+      "learning_rate": 8e-05,
+      "loss": 1.4358,
+      "step": 4855
+    },
+    {
+      "epoch": 0.6627541968063327,
+      "grad_norm": 0.22292497754096985,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 4856
+    },
+    {
+      "epoch": 0.6628906783130886,
+      "grad_norm": 0.2175852358341217,
+      "learning_rate": 8e-05,
+      "loss": 1.4742,
+      "step": 4857
+    },
+    {
+      "epoch": 0.6630271598198444,
+      "grad_norm": 0.22351761162281036,
+      "learning_rate": 8e-05,
+      "loss": 1.4057,
+      "step": 4858
+    },
+    {
+      "epoch": 0.6631636413266002,
+      "grad_norm": 0.22377492487430573,
+      "learning_rate": 8e-05,
+      "loss": 1.4806,
+      "step": 4859
+    },
+    {
+      "epoch": 0.663300122833356,
+      "grad_norm": 0.22567082941532135,
+      "learning_rate": 8e-05,
+      "loss": 1.4576,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6634366043401119,
+      "grad_norm": 0.2307164967060089,
+      "learning_rate": 8e-05,
+      "loss": 1.5273,
+      "step": 4861
+    },
+    {
+      "epoch": 0.6635730858468677,
+      "grad_norm": 0.2237037569284439,
+      "learning_rate": 8e-05,
+      "loss": 1.439,
+      "step": 4862
+    },
+    {
+      "epoch": 0.6637095673536236,
+      "grad_norm": 0.22361455857753754,
+      "learning_rate": 8e-05,
+      "loss": 1.4173,
+      "step": 4863
+    },
+    {
+      "epoch": 0.6638460488603795,
+      "grad_norm": 0.22322291135787964,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 4864
+    },
+    {
+      "epoch": 0.6639825303671353,
+      "grad_norm": 0.2231309711933136,
+      "learning_rate": 8e-05,
+      "loss": 1.4427,
+      "step": 4865
+    },
+    {
+      "epoch": 0.6641190118738911,
+      "grad_norm": 0.22826983034610748,
+      "learning_rate": 8e-05,
+      "loss": 1.4418,
+      "step": 4866
+    },
+    {
+      "epoch": 0.6642554933806469,
+      "grad_norm": 0.22191420197486877,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 4867
+    },
+    {
+      "epoch": 0.6643919748874028,
+      "grad_norm": 0.22914348542690277,
+      "learning_rate": 8e-05,
+      "loss": 1.5048,
+      "step": 4868
+    },
+    {
+      "epoch": 0.6645284563941586,
+      "grad_norm": 0.22400781512260437,
+      "learning_rate": 8e-05,
+      "loss": 1.4784,
+      "step": 4869
+    },
+    {
+      "epoch": 0.6646649379009144,
+      "grad_norm": 0.22684012353420258,
+      "learning_rate": 8e-05,
+      "loss": 1.4736,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6648014194076702,
+      "grad_norm": 0.2189878672361374,
+      "learning_rate": 8e-05,
+      "loss": 1.4344,
+      "step": 4871
+    },
+    {
+      "epoch": 0.6649379009144261,
+      "grad_norm": 0.21996714174747467,
+      "learning_rate": 8e-05,
+      "loss": 1.3952,
+      "step": 4872
+    },
+    {
+      "epoch": 0.6650743824211819,
+      "grad_norm": 0.22384747862815857,
+      "learning_rate": 8e-05,
+      "loss": 1.4758,
+      "step": 4873
+    },
+    {
+      "epoch": 0.6652108639279377,
+      "grad_norm": 0.22900566458702087,
+      "learning_rate": 8e-05,
+      "loss": 1.4313,
+      "step": 4874
+    },
+    {
+      "epoch": 0.6653473454346936,
+      "grad_norm": 0.2260541319847107,
+      "learning_rate": 8e-05,
+      "loss": 1.4314,
+      "step": 4875
+    },
+    {
+      "epoch": 0.6654838269414495,
+      "grad_norm": 0.22457273304462433,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 4876
+    },
+    {
+      "epoch": 0.6656203084482053,
+      "grad_norm": 0.2243596464395523,
+      "learning_rate": 8e-05,
+      "loss": 1.4496,
+      "step": 4877
+    },
+    {
+      "epoch": 0.6657567899549611,
+      "grad_norm": 0.22972197830677032,
+      "learning_rate": 8e-05,
+      "loss": 1.4549,
+      "step": 4878
+    },
+    {
+      "epoch": 0.665893271461717,
+      "grad_norm": 0.22231552004814148,
+      "learning_rate": 8e-05,
+      "loss": 1.495,
+      "step": 4879
+    },
+    {
+      "epoch": 0.6660297529684728,
+      "grad_norm": 0.2262486070394516,
+      "learning_rate": 8e-05,
+      "loss": 1.4297,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6661662344752286,
+      "grad_norm": 0.22087006270885468,
+      "learning_rate": 8e-05,
+      "loss": 1.454,
+      "step": 4881
+    },
+    {
+      "epoch": 0.6663027159819844,
+      "grad_norm": 0.2140192836523056,
+      "learning_rate": 8e-05,
+      "loss": 1.4475,
+      "step": 4882
+    },
+    {
+      "epoch": 0.6664391974887403,
+      "grad_norm": 0.22473061084747314,
+      "learning_rate": 8e-05,
+      "loss": 1.4418,
+      "step": 4883
+    },
+    {
+      "epoch": 0.6665756789954961,
+      "grad_norm": 0.22641408443450928,
+      "learning_rate": 8e-05,
+      "loss": 1.3948,
+      "step": 4884
+    },
+    {
+      "epoch": 0.6667121605022519,
+      "grad_norm": 0.21617218852043152,
+      "learning_rate": 8e-05,
+      "loss": 1.4503,
+      "step": 4885
+    },
+    {
+      "epoch": 0.6668486420090077,
+      "grad_norm": 0.2311301976442337,
+      "learning_rate": 8e-05,
+      "loss": 1.4791,
+      "step": 4886
+    },
+    {
+      "epoch": 0.6669851235157637,
+      "grad_norm": 0.2306140512228012,
+      "learning_rate": 8e-05,
+      "loss": 1.4462,
+      "step": 4887
+    },
+    {
+      "epoch": 0.6671216050225195,
+      "grad_norm": 0.2413230985403061,
+      "learning_rate": 8e-05,
+      "loss": 1.4906,
+      "step": 4888
+    },
+    {
+      "epoch": 0.6672580865292753,
+      "grad_norm": 0.2357136607170105,
+      "learning_rate": 8e-05,
+      "loss": 1.4267,
+      "step": 4889
+    },
+    {
+      "epoch": 0.6673945680360311,
+      "grad_norm": 0.23352450132369995,
+      "learning_rate": 8e-05,
+      "loss": 1.4838,
+      "step": 4890
+    },
+    {
+      "epoch": 0.667531049542787,
+      "grad_norm": 0.22657707333564758,
+      "learning_rate": 8e-05,
+      "loss": 1.4608,
+      "step": 4891
+    },
+    {
+      "epoch": 0.6676675310495428,
+      "grad_norm": 0.22659124433994293,
+      "learning_rate": 8e-05,
+      "loss": 1.452,
+      "step": 4892
+    },
+    {
+      "epoch": 0.6678040125562986,
+      "grad_norm": 0.2184356302022934,
+      "learning_rate": 8e-05,
+      "loss": 1.4008,
+      "step": 4893
+    },
+    {
+      "epoch": 0.6679404940630544,
+      "grad_norm": 0.22633616626262665,
+      "learning_rate": 8e-05,
+      "loss": 1.45,
+      "step": 4894
+    },
+    {
+      "epoch": 0.6680769755698103,
+      "grad_norm": 0.2222093939781189,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 4895
+    },
+    {
+      "epoch": 0.6682134570765661,
+      "grad_norm": 0.22351354360580444,
+      "learning_rate": 8e-05,
+      "loss": 1.429,
+      "step": 4896
+    },
+    {
+      "epoch": 0.6683499385833219,
+      "grad_norm": 0.21014724671840668,
+      "learning_rate": 8e-05,
+      "loss": 1.3845,
+      "step": 4897
+    },
+    {
+      "epoch": 0.6684864200900777,
+      "grad_norm": 0.2480313777923584,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 4898
+    },
+    {
+      "epoch": 0.6686229015968336,
+      "grad_norm": 0.2141083925962448,
+      "learning_rate": 8e-05,
+      "loss": 1.3595,
+      "step": 4899
+    },
+    {
+      "epoch": 0.6687593831035895,
+      "grad_norm": 0.22754399478435516,
+      "learning_rate": 8e-05,
+      "loss": 1.3802,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6688958646103453,
+      "grad_norm": 0.2352703958749771,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 4901
+    },
+    {
+      "epoch": 0.6690323461171012,
+      "grad_norm": 0.22965887188911438,
+      "learning_rate": 8e-05,
+      "loss": 1.5009,
+      "step": 4902
+    },
+    {
+      "epoch": 0.669168827623857,
+      "grad_norm": 0.22749994695186615,
+      "learning_rate": 8e-05,
+      "loss": 1.5276,
+      "step": 4903
+    },
+    {
+      "epoch": 0.6693053091306128,
+      "grad_norm": 0.22773905098438263,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 4904
+    },
+    {
+      "epoch": 0.6694417906373686,
+      "grad_norm": 0.22422143816947937,
+      "learning_rate": 8e-05,
+      "loss": 1.4213,
+      "step": 4905
+    },
+    {
+      "epoch": 0.6695782721441245,
+      "grad_norm": 0.22762185335159302,
+      "learning_rate": 8e-05,
+      "loss": 1.4212,
+      "step": 4906
+    },
+    {
+      "epoch": 0.6697147536508803,
+      "grad_norm": 0.22892232239246368,
+      "learning_rate": 8e-05,
+      "loss": 1.4378,
+      "step": 4907
+    },
+    {
+      "epoch": 0.6698512351576361,
+      "grad_norm": 0.22859223186969757,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 4908
+    },
+    {
+      "epoch": 0.6699877166643919,
+      "grad_norm": 0.2299458235502243,
+      "learning_rate": 8e-05,
+      "loss": 1.4645,
+      "step": 4909
+    },
+    {
+      "epoch": 0.6701241981711478,
+      "grad_norm": 0.22010061144828796,
+      "learning_rate": 8e-05,
+      "loss": 1.4523,
+      "step": 4910
+    },
+    {
+      "epoch": 0.6702606796779036,
+      "grad_norm": 0.22795507311820984,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 4911
+    },
+    {
+      "epoch": 0.6703971611846595,
+      "grad_norm": 0.22864168882369995,
+      "learning_rate": 8e-05,
+      "loss": 1.4218,
+      "step": 4912
+    },
+    {
+      "epoch": 0.6705336426914154,
+      "grad_norm": 0.2272016555070877,
+      "learning_rate": 8e-05,
+      "loss": 1.4777,
+      "step": 4913
+    },
+    {
+      "epoch": 0.6706701241981712,
+      "grad_norm": 0.21866293251514435,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 4914
+    },
+    {
+      "epoch": 0.670806605704927,
+      "grad_norm": 0.22426877915859222,
+      "learning_rate": 8e-05,
+      "loss": 1.5097,
+      "step": 4915
+    },
+    {
+      "epoch": 0.6709430872116828,
+      "grad_norm": 0.22679133713245392,
+      "learning_rate": 8e-05,
+      "loss": 1.4773,
+      "step": 4916
+    },
+    {
+      "epoch": 0.6710795687184387,
+      "grad_norm": 0.22114460170269012,
+      "learning_rate": 8e-05,
+      "loss": 1.4161,
+      "step": 4917
+    },
+    {
+      "epoch": 0.6712160502251945,
+      "grad_norm": 0.23109637200832367,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 4918
+    },
+    {
+      "epoch": 0.6713525317319503,
+      "grad_norm": 0.22195497155189514,
+      "learning_rate": 8e-05,
+      "loss": 1.5173,
+      "step": 4919
+    },
+    {
+      "epoch": 0.6714890132387061,
+      "grad_norm": 0.23254257440567017,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 4920
+    },
+    {
+      "epoch": 0.671625494745462,
+      "grad_norm": 0.21887019276618958,
+      "learning_rate": 8e-05,
+      "loss": 1.424,
+      "step": 4921
+    },
+    {
+      "epoch": 0.6717619762522178,
+      "grad_norm": 0.23122921586036682,
+      "learning_rate": 8e-05,
+      "loss": 1.4609,
+      "step": 4922
+    },
+    {
+      "epoch": 0.6718984577589736,
+      "grad_norm": 0.2219562530517578,
+      "learning_rate": 8e-05,
+      "loss": 1.4389,
+      "step": 4923
+    },
+    {
+      "epoch": 0.6720349392657295,
+      "grad_norm": 0.22108280658721924,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 4924
+    },
+    {
+      "epoch": 0.6721714207724854,
+      "grad_norm": 0.22000271081924438,
+      "learning_rate": 8e-05,
+      "loss": 1.4092,
+      "step": 4925
+    },
+    {
+      "epoch": 0.6723079022792412,
+      "grad_norm": 0.2324644774198532,
+      "learning_rate": 8e-05,
+      "loss": 1.4544,
+      "step": 4926
+    },
+    {
+      "epoch": 0.672444383785997,
+      "grad_norm": 0.22577349841594696,
+      "learning_rate": 8e-05,
+      "loss": 1.4262,
+      "step": 4927
+    },
+    {
+      "epoch": 0.6725808652927528,
+      "grad_norm": 0.23456013202667236,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 4928
+    },
+    {
+      "epoch": 0.6727173467995087,
+      "grad_norm": 0.21577303111553192,
+      "learning_rate": 8e-05,
+      "loss": 1.3793,
+      "step": 4929
+    },
+    {
+      "epoch": 0.6728538283062645,
+      "grad_norm": 0.2322041094303131,
+      "learning_rate": 8e-05,
+      "loss": 1.4349,
+      "step": 4930
+    },
+    {
+      "epoch": 0.6729903098130203,
+      "grad_norm": 0.23538881540298462,
+      "learning_rate": 8e-05,
+      "loss": 1.4448,
+      "step": 4931
+    },
+    {
+      "epoch": 0.6731267913197762,
+      "grad_norm": 0.22795306146144867,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 4932
+    },
+    {
+      "epoch": 0.673263272826532,
+      "grad_norm": 0.23400461673736572,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 4933
+    },
+    {
+      "epoch": 0.6733997543332878,
+      "grad_norm": 0.2218569964170456,
+      "learning_rate": 8e-05,
+      "loss": 1.478,
+      "step": 4934
+    },
+    {
+      "epoch": 0.6735362358400436,
+      "grad_norm": 0.22215279936790466,
+      "learning_rate": 8e-05,
+      "loss": 1.4377,
+      "step": 4935
+    },
+    {
+      "epoch": 0.6736727173467995,
+      "grad_norm": 0.21808898448944092,
+      "learning_rate": 8e-05,
+      "loss": 1.4161,
+      "step": 4936
+    },
+    {
+      "epoch": 0.6738091988535554,
+      "grad_norm": 0.22512227296829224,
+      "learning_rate": 8e-05,
+      "loss": 1.4086,
+      "step": 4937
+    },
+    {
+      "epoch": 0.6739456803603112,
+      "grad_norm": 0.2279331088066101,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 4938
+    },
+    {
+      "epoch": 0.674082161867067,
+      "grad_norm": 0.22706177830696106,
+      "learning_rate": 8e-05,
+      "loss": 1.4835,
+      "step": 4939
+    },
+    {
+      "epoch": 0.6742186433738229,
+      "grad_norm": 0.22859534621238708,
+      "learning_rate": 8e-05,
+      "loss": 1.5402,
+      "step": 4940
+    },
+    {
+      "epoch": 0.6743551248805787,
+      "grad_norm": 0.23102515935897827,
+      "learning_rate": 8e-05,
+      "loss": 1.4611,
+      "step": 4941
+    },
+    {
+      "epoch": 0.6744916063873345,
+      "grad_norm": 0.22415749728679657,
+      "learning_rate": 8e-05,
+      "loss": 1.47,
+      "step": 4942
+    },
+    {
+      "epoch": 0.6746280878940903,
+      "grad_norm": 0.22143179178237915,
+      "learning_rate": 8e-05,
+      "loss": 1.4665,
+      "step": 4943
+    },
+    {
+      "epoch": 0.6747645694008462,
+      "grad_norm": 0.22009111940860748,
+      "learning_rate": 8e-05,
+      "loss": 1.4634,
+      "step": 4944
+    },
+    {
+      "epoch": 0.674901050907602,
+      "grad_norm": 0.2190183848142624,
+      "learning_rate": 8e-05,
+      "loss": 1.4684,
+      "step": 4945
+    },
+    {
+      "epoch": 0.6750375324143578,
+      "grad_norm": 0.2439507395029068,
+      "learning_rate": 8e-05,
+      "loss": 1.5442,
+      "step": 4946
+    },
+    {
+      "epoch": 0.6751740139211136,
+      "grad_norm": 0.21957656741142273,
+      "learning_rate": 8e-05,
+      "loss": 1.4431,
+      "step": 4947
+    },
+    {
+      "epoch": 0.6753104954278695,
+      "grad_norm": 0.23306095600128174,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 4948
+    },
+    {
+      "epoch": 0.6754469769346254,
+      "grad_norm": 0.23035183548927307,
+      "learning_rate": 8e-05,
+      "loss": 1.4345,
+      "step": 4949
+    },
+    {
+      "epoch": 0.6755834584413812,
+      "grad_norm": 0.2369430810213089,
+      "learning_rate": 8e-05,
+      "loss": 1.5298,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6757199399481371,
+      "grad_norm": 0.2293514609336853,
+      "learning_rate": 8e-05,
+      "loss": 1.478,
+      "step": 4951
+    },
+    {
+      "epoch": 0.6758564214548929,
+      "grad_norm": 0.2301241159439087,
+      "learning_rate": 8e-05,
+      "loss": 1.4466,
+      "step": 4952
+    },
+    {
+      "epoch": 0.6759929029616487,
+      "grad_norm": 0.2373674362897873,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 4953
+    },
+    {
+      "epoch": 0.6761293844684045,
+      "grad_norm": 0.2303207814693451,
+      "learning_rate": 8e-05,
+      "loss": 1.4528,
+      "step": 4954
+    },
+    {
+      "epoch": 0.6762658659751604,
+      "grad_norm": 0.22433367371559143,
+      "learning_rate": 8e-05,
+      "loss": 1.4345,
+      "step": 4955
+    },
+    {
+      "epoch": 0.6764023474819162,
+      "grad_norm": 0.23255306482315063,
+      "learning_rate": 8e-05,
+      "loss": 1.4989,
+      "step": 4956
+    },
+    {
+      "epoch": 0.676538828988672,
+      "grad_norm": 0.22652313113212585,
+      "learning_rate": 8e-05,
+      "loss": 1.4862,
+      "step": 4957
+    },
+    {
+      "epoch": 0.6766753104954278,
+      "grad_norm": 0.23774616420269012,
+      "learning_rate": 8e-05,
+      "loss": 1.5009,
+      "step": 4958
+    },
+    {
+      "epoch": 0.6768117920021837,
+      "grad_norm": 0.23445482552051544,
+      "learning_rate": 8e-05,
+      "loss": 1.5441,
+      "step": 4959
+    },
+    {
+      "epoch": 0.6769482735089395,
+      "grad_norm": 0.22669482231140137,
+      "learning_rate": 8e-05,
+      "loss": 1.499,
+      "step": 4960
+    },
+    {
+      "epoch": 0.6770847550156954,
+      "grad_norm": 0.2172665148973465,
+      "learning_rate": 8e-05,
+      "loss": 1.4124,
+      "step": 4961
+    },
+    {
+      "epoch": 0.6772212365224513,
+      "grad_norm": 0.21616071462631226,
+      "learning_rate": 8e-05,
+      "loss": 1.4306,
+      "step": 4962
+    },
+    {
+      "epoch": 0.6773577180292071,
+      "grad_norm": 0.2244592010974884,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 4963
+    },
+    {
+      "epoch": 0.6774941995359629,
+      "grad_norm": 0.22665441036224365,
+      "learning_rate": 8e-05,
+      "loss": 1.4988,
+      "step": 4964
+    },
+    {
+      "epoch": 0.6776306810427187,
+      "grad_norm": 0.2233804613351822,
+      "learning_rate": 8e-05,
+      "loss": 1.4369,
+      "step": 4965
+    },
+    {
+      "epoch": 0.6777671625494746,
+      "grad_norm": 0.2426506131887436,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 4966
+    },
+    {
+      "epoch": 0.6779036440562304,
+      "grad_norm": 0.23367524147033691,
+      "learning_rate": 8e-05,
+      "loss": 1.4667,
+      "step": 4967
+    },
+    {
+      "epoch": 0.6780401255629862,
+      "grad_norm": 0.218913272023201,
+      "learning_rate": 8e-05,
+      "loss": 1.3863,
+      "step": 4968
+    },
+    {
+      "epoch": 0.678176607069742,
+      "grad_norm": 0.2259284257888794,
+      "learning_rate": 8e-05,
+      "loss": 1.4675,
+      "step": 4969
+    },
+    {
+      "epoch": 0.6783130885764979,
+      "grad_norm": 0.22640331089496613,
+      "learning_rate": 8e-05,
+      "loss": 1.4751,
+      "step": 4970
+    },
+    {
+      "epoch": 0.6784495700832537,
+      "grad_norm": 0.2255599945783615,
+      "learning_rate": 8e-05,
+      "loss": 1.4622,
+      "step": 4971
+    },
+    {
+      "epoch": 0.6785860515900095,
+      "grad_norm": 0.2304847240447998,
+      "learning_rate": 8e-05,
+      "loss": 1.4923,
+      "step": 4972
+    },
+    {
+      "epoch": 0.6787225330967653,
+      "grad_norm": 0.2298392355442047,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 4973
+    },
+    {
+      "epoch": 0.6788590146035213,
+      "grad_norm": 0.2209482342004776,
+      "learning_rate": 8e-05,
+      "loss": 1.4619,
+      "step": 4974
+    },
+    {
+      "epoch": 0.6789954961102771,
+      "grad_norm": 0.22957704961299896,
+      "learning_rate": 8e-05,
+      "loss": 1.5004,
+      "step": 4975
+    },
+    {
+      "epoch": 0.6791319776170329,
+      "grad_norm": 0.22390051186084747,
+      "learning_rate": 8e-05,
+      "loss": 1.4065,
+      "step": 4976
+    },
+    {
+      "epoch": 0.6792684591237887,
+      "grad_norm": 0.22878144681453705,
+      "learning_rate": 8e-05,
+      "loss": 1.4588,
+      "step": 4977
+    },
+    {
+      "epoch": 0.6794049406305446,
+      "grad_norm": 0.22576715052127838,
+      "learning_rate": 8e-05,
+      "loss": 1.3586,
+      "step": 4978
+    },
+    {
+      "epoch": 0.6795414221373004,
+      "grad_norm": 0.23070695996284485,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 4979
+    },
+    {
+      "epoch": 0.6796779036440562,
+      "grad_norm": 0.23301833868026733,
+      "learning_rate": 8e-05,
+      "loss": 1.4213,
+      "step": 4980
+    },
+    {
+      "epoch": 0.679814385150812,
+      "grad_norm": 0.22595970332622528,
+      "learning_rate": 8e-05,
+      "loss": 1.3941,
+      "step": 4981
+    },
+    {
+      "epoch": 0.6799508666575679,
+      "grad_norm": 0.2268526405096054,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 4982
+    },
+    {
+      "epoch": 0.6800873481643237,
+      "grad_norm": 0.22498205304145813,
+      "learning_rate": 8e-05,
+      "loss": 1.4033,
+      "step": 4983
+    },
+    {
+      "epoch": 0.6802238296710795,
+      "grad_norm": 0.22633042931556702,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 4984
+    },
+    {
+      "epoch": 0.6803603111778354,
+      "grad_norm": 0.2367359846830368,
+      "learning_rate": 8e-05,
+      "loss": 1.4974,
+      "step": 4985
+    },
+    {
+      "epoch": 0.6804967926845913,
+      "grad_norm": 0.22913554310798645,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 4986
+    },
+    {
+      "epoch": 0.6806332741913471,
+      "grad_norm": 0.22574195265769958,
+      "learning_rate": 8e-05,
+      "loss": 1.4723,
+      "step": 4987
+    },
+    {
+      "epoch": 0.6807697556981029,
+      "grad_norm": 0.2198648601770401,
+      "learning_rate": 8e-05,
+      "loss": 1.3365,
+      "step": 4988
+    },
+    {
+      "epoch": 0.6809062372048588,
+      "grad_norm": 0.22332395613193512,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 4989
+    },
+    {
+      "epoch": 0.6810427187116146,
+      "grad_norm": 0.2273492068052292,
+      "learning_rate": 8e-05,
+      "loss": 1.441,
+      "step": 4990
+    },
+    {
+      "epoch": 0.6811792002183704,
+      "grad_norm": 0.22265857458114624,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 4991
+    },
+    {
+      "epoch": 0.6813156817251262,
+      "grad_norm": 0.23119039833545685,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 4992
+    },
+    {
+      "epoch": 0.6814521632318821,
+      "grad_norm": 0.23217029869556427,
+      "learning_rate": 8e-05,
+      "loss": 1.4716,
+      "step": 4993
+    },
+    {
+      "epoch": 0.6815886447386379,
+      "grad_norm": 0.22838056087493896,
+      "learning_rate": 8e-05,
+      "loss": 1.4904,
+      "step": 4994
+    },
+    {
+      "epoch": 0.6817251262453937,
+      "grad_norm": 0.22141489386558533,
+      "learning_rate": 8e-05,
+      "loss": 1.3976,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6818616077521495,
+      "grad_norm": 0.23647750914096832,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 4996
+    },
+    {
+      "epoch": 0.6819980892589054,
+      "grad_norm": 0.22490087151527405,
+      "learning_rate": 8e-05,
+      "loss": 1.4621,
+      "step": 4997
+    },
+    {
+      "epoch": 0.6821345707656613,
+      "grad_norm": 0.21920499205589294,
+      "learning_rate": 8e-05,
+      "loss": 1.3793,
+      "step": 4998
+    },
+    {
+      "epoch": 0.6822710522724171,
+      "grad_norm": 0.23083849251270294,
+      "learning_rate": 8e-05,
+      "loss": 1.5515,
+      "step": 4999
+    },
+    {
+      "epoch": 0.682407533779173,
+      "grad_norm": 0.2255854308605194,
+      "learning_rate": 8e-05,
+      "loss": 1.4975,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6825440152859288,
+      "grad_norm": 0.22592756152153015,
+      "learning_rate": 8e-05,
+      "loss": 1.4833,
+      "step": 5001
+    },
+    {
+      "epoch": 0.6826804967926846,
+      "grad_norm": 0.22824306786060333,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 5002
+    },
+    {
+      "epoch": 0.6828169782994404,
+      "grad_norm": 0.2275524139404297,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 5003
+    },
+    {
+      "epoch": 0.6829534598061963,
+      "grad_norm": 0.22983388602733612,
+      "learning_rate": 8e-05,
+      "loss": 1.5549,
+      "step": 5004
+    },
+    {
+      "epoch": 0.6830899413129521,
+      "grad_norm": 0.2244010865688324,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6832264228197079,
+      "grad_norm": 0.2313825786113739,
+      "learning_rate": 8e-05,
+      "loss": 1.4259,
+      "step": 5006
+    },
+    {
+      "epoch": 0.6833629043264637,
+      "grad_norm": 0.22144214808940887,
+      "learning_rate": 8e-05,
+      "loss": 1.3826,
+      "step": 5007
+    },
+    {
+      "epoch": 0.6834993858332196,
+      "grad_norm": 0.23368731141090393,
+      "learning_rate": 8e-05,
+      "loss": 1.4919,
+      "step": 5008
+    },
+    {
+      "epoch": 0.6836358673399754,
+      "grad_norm": 0.2234659492969513,
+      "learning_rate": 8e-05,
+      "loss": 1.4538,
+      "step": 5009
+    },
+    {
+      "epoch": 0.6837723488467312,
+      "grad_norm": 0.2257426232099533,
+      "learning_rate": 8e-05,
+      "loss": 1.4588,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6839088303534872,
+      "grad_norm": 0.2614365220069885,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 5011
+    },
+    {
+      "epoch": 0.684045311860243,
+      "grad_norm": 0.23928286135196686,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 5012
+    },
+    {
+      "epoch": 0.6841817933669988,
+      "grad_norm": 0.23042115569114685,
+      "learning_rate": 8e-05,
+      "loss": 1.4579,
+      "step": 5013
+    },
+    {
+      "epoch": 0.6843182748737546,
+      "grad_norm": 0.23792703449726105,
+      "learning_rate": 8e-05,
+      "loss": 1.4417,
+      "step": 5014
+    },
+    {
+      "epoch": 0.6844547563805105,
+      "grad_norm": 0.22855287790298462,
+      "learning_rate": 8e-05,
+      "loss": 1.4811,
+      "step": 5015
+    },
+    {
+      "epoch": 0.6845912378872663,
+      "grad_norm": 0.22891730070114136,
+      "learning_rate": 8e-05,
+      "loss": 1.439,
+      "step": 5016
+    },
+    {
+      "epoch": 0.6847277193940221,
+      "grad_norm": 0.23642709851264954,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 5017
+    },
+    {
+      "epoch": 0.6848642009007779,
+      "grad_norm": 0.22723734378814697,
+      "learning_rate": 8e-05,
+      "loss": 1.4127,
+      "step": 5018
+    },
+    {
+      "epoch": 0.6850006824075338,
+      "grad_norm": 0.22933165729045868,
+      "learning_rate": 8e-05,
+      "loss": 1.4133,
+      "step": 5019
+    },
+    {
+      "epoch": 0.6851371639142896,
+      "grad_norm": 0.2300548106431961,
+      "learning_rate": 8e-05,
+      "loss": 1.4386,
+      "step": 5020
+    },
+    {
+      "epoch": 0.6852736454210454,
+      "grad_norm": 0.24369780719280243,
+      "learning_rate": 8e-05,
+      "loss": 1.4879,
+      "step": 5021
+    },
+    {
+      "epoch": 0.6854101269278012,
+      "grad_norm": 0.2261851578950882,
+      "learning_rate": 8e-05,
+      "loss": 1.4612,
+      "step": 5022
+    },
+    {
+      "epoch": 0.6855466084345572,
+      "grad_norm": 0.23346692323684692,
+      "learning_rate": 8e-05,
+      "loss": 1.4026,
+      "step": 5023
+    },
+    {
+      "epoch": 0.685683089941313,
+      "grad_norm": 0.22375231981277466,
+      "learning_rate": 8e-05,
+      "loss": 1.4465,
+      "step": 5024
+    },
+    {
+      "epoch": 0.6858195714480688,
+      "grad_norm": 0.22330830991268158,
+      "learning_rate": 8e-05,
+      "loss": 1.4014,
+      "step": 5025
+    },
+    {
+      "epoch": 0.6859560529548246,
+      "grad_norm": 0.24908140301704407,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 5026
+    },
+    {
+      "epoch": 0.6860925344615805,
+      "grad_norm": 0.22809389233589172,
+      "learning_rate": 8e-05,
+      "loss": 1.454,
+      "step": 5027
+    },
+    {
+      "epoch": 0.6862290159683363,
+      "grad_norm": 0.22787150740623474,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 5028
+    },
+    {
+      "epoch": 0.6863654974750921,
+      "grad_norm": 0.23242376744747162,
+      "learning_rate": 8e-05,
+      "loss": 1.4976,
+      "step": 5029
+    },
+    {
+      "epoch": 0.686501978981848,
+      "grad_norm": 0.24245303869247437,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 5030
+    },
+    {
+      "epoch": 0.6866384604886038,
+      "grad_norm": 0.23010410368442535,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 5031
+    },
+    {
+      "epoch": 0.6867749419953596,
+      "grad_norm": 0.23047325015068054,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 5032
+    },
+    {
+      "epoch": 0.6869114235021154,
+      "grad_norm": 0.24399243295192719,
+      "learning_rate": 8e-05,
+      "loss": 1.475,
+      "step": 5033
+    },
+    {
+      "epoch": 0.6870479050088713,
+      "grad_norm": 0.22931353747844696,
+      "learning_rate": 8e-05,
+      "loss": 1.4758,
+      "step": 5034
+    },
+    {
+      "epoch": 0.6871843865156272,
+      "grad_norm": 0.22713154554367065,
+      "learning_rate": 8e-05,
+      "loss": 1.4207,
+      "step": 5035
+    },
+    {
+      "epoch": 0.687320868022383,
+      "grad_norm": 0.23670606315135956,
+      "learning_rate": 8e-05,
+      "loss": 1.4731,
+      "step": 5036
+    },
+    {
+      "epoch": 0.6874573495291388,
+      "grad_norm": 0.23109430074691772,
+      "learning_rate": 8e-05,
+      "loss": 1.4922,
+      "step": 5037
+    },
+    {
+      "epoch": 0.6875938310358947,
+      "grad_norm": 0.2223534882068634,
+      "learning_rate": 8e-05,
+      "loss": 1.4225,
+      "step": 5038
+    },
+    {
+      "epoch": 0.6877303125426505,
+      "grad_norm": 0.23939067125320435,
+      "learning_rate": 8e-05,
+      "loss": 1.4359,
+      "step": 5039
+    },
+    {
+      "epoch": 0.6878667940494063,
+      "grad_norm": 0.23268179595470428,
+      "learning_rate": 8e-05,
+      "loss": 1.4389,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6880032755561621,
+      "grad_norm": 0.23072464764118195,
+      "learning_rate": 8e-05,
+      "loss": 1.5093,
+      "step": 5041
+    },
+    {
+      "epoch": 0.688139757062918,
+      "grad_norm": 0.23465904593467712,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 5042
+    },
+    {
+      "epoch": 0.6882762385696738,
+      "grad_norm": 0.23594097793102264,
+      "learning_rate": 8e-05,
+      "loss": 1.4126,
+      "step": 5043
+    },
+    {
+      "epoch": 0.6884127200764296,
+      "grad_norm": 0.21398280560970306,
+      "learning_rate": 8e-05,
+      "loss": 1.3772,
+      "step": 5044
+    },
+    {
+      "epoch": 0.6885492015831854,
+      "grad_norm": 0.23254543542861938,
+      "learning_rate": 8e-05,
+      "loss": 1.4277,
+      "step": 5045
+    },
+    {
+      "epoch": 0.6886856830899413,
+      "grad_norm": 0.244389146566391,
+      "learning_rate": 8e-05,
+      "loss": 1.5489,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6888221645966971,
+      "grad_norm": 0.23840005695819855,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 5047
+    },
+    {
+      "epoch": 0.688958646103453,
+      "grad_norm": 0.22709038853645325,
+      "learning_rate": 8e-05,
+      "loss": 1.4217,
+      "step": 5048
+    },
+    {
+      "epoch": 0.6890951276102089,
+      "grad_norm": 0.23280467092990875,
+      "learning_rate": 8e-05,
+      "loss": 1.4706,
+      "step": 5049
+    },
+    {
+      "epoch": 0.6892316091169647,
+      "grad_norm": 0.23505531251430511,
+      "learning_rate": 8e-05,
+      "loss": 1.449,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6893680906237205,
+      "grad_norm": 0.23104819655418396,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 5051
+    },
+    {
+      "epoch": 0.6895045721304763,
+      "grad_norm": 0.2260597050189972,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 5052
+    },
+    {
+      "epoch": 0.6896410536372322,
+      "grad_norm": 0.24535372853279114,
+      "learning_rate": 8e-05,
+      "loss": 1.49,
+      "step": 5053
+    },
+    {
+      "epoch": 0.689777535143988,
+      "grad_norm": 0.2321591079235077,
+      "learning_rate": 8e-05,
+      "loss": 1.4402,
+      "step": 5054
+    },
+    {
+      "epoch": 0.6899140166507438,
+      "grad_norm": 0.22221209108829498,
+      "learning_rate": 8e-05,
+      "loss": 1.3016,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6900504981574996,
+      "grad_norm": 0.23392553627490997,
+      "learning_rate": 8e-05,
+      "loss": 1.5213,
+      "step": 5056
+    },
+    {
+      "epoch": 0.6901869796642555,
+      "grad_norm": 0.22481046617031097,
+      "learning_rate": 8e-05,
+      "loss": 1.4362,
+      "step": 5057
+    },
+    {
+      "epoch": 0.6903234611710113,
+      "grad_norm": 0.23036149144172668,
+      "learning_rate": 8e-05,
+      "loss": 1.4001,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6904599426777671,
+      "grad_norm": 0.22406387329101562,
+      "learning_rate": 8e-05,
+      "loss": 1.4755,
+      "step": 5059
+    },
+    {
+      "epoch": 0.690596424184523,
+      "grad_norm": 0.22814059257507324,
+      "learning_rate": 8e-05,
+      "loss": 1.5234,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6907329056912789,
+      "grad_norm": 0.24145053327083588,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 5061
+    },
+    {
+      "epoch": 0.6908693871980347,
+      "grad_norm": 0.2258089929819107,
+      "learning_rate": 8e-05,
+      "loss": 1.3877,
+      "step": 5062
+    },
+    {
+      "epoch": 0.6910058687047905,
+      "grad_norm": 0.22434434294700623,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 5063
+    },
+    {
+      "epoch": 0.6911423502115464,
+      "grad_norm": 0.2281690090894699,
+      "learning_rate": 8e-05,
+      "loss": 1.4974,
+      "step": 5064
+    },
+    {
+      "epoch": 0.6912788317183022,
+      "grad_norm": 0.23077541589736938,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 5065
+    },
+    {
+      "epoch": 0.691415313225058,
+      "grad_norm": 0.22752626240253448,
+      "learning_rate": 8e-05,
+      "loss": 1.4852,
+      "step": 5066
+    },
+    {
+      "epoch": 0.6915517947318138,
+      "grad_norm": 0.23466192185878754,
+      "learning_rate": 8e-05,
+      "loss": 1.4615,
+      "step": 5067
+    },
+    {
+      "epoch": 0.6916882762385697,
+      "grad_norm": 0.23212094604969025,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 5068
+    },
+    {
+      "epoch": 0.6918247577453255,
+      "grad_norm": 0.23421894013881683,
+      "learning_rate": 8e-05,
+      "loss": 1.4712,
+      "step": 5069
+    },
+    {
+      "epoch": 0.6919612392520813,
+      "grad_norm": 0.2413264513015747,
+      "learning_rate": 8e-05,
+      "loss": 1.4996,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6920977207588371,
+      "grad_norm": 0.22626392543315887,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 5071
+    },
+    {
+      "epoch": 0.692234202265593,
+      "grad_norm": 0.22106380760669708,
+      "learning_rate": 8e-05,
+      "loss": 1.3622,
+      "step": 5072
+    },
+    {
+      "epoch": 0.6923706837723489,
+      "grad_norm": 0.22447755932807922,
+      "learning_rate": 8e-05,
+      "loss": 1.4501,
+      "step": 5073
+    },
+    {
+      "epoch": 0.6925071652791047,
+      "grad_norm": 0.23246055841445923,
+      "learning_rate": 8e-05,
+      "loss": 1.4439,
+      "step": 5074
+    },
+    {
+      "epoch": 0.6926436467858605,
+      "grad_norm": 0.23439981043338776,
+      "learning_rate": 8e-05,
+      "loss": 1.4852,
+      "step": 5075
+    },
+    {
+      "epoch": 0.6927801282926164,
+      "grad_norm": 0.23108838498592377,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 5076
+    },
+    {
+      "epoch": 0.6929166097993722,
+      "grad_norm": 0.23152801394462585,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 5077
+    },
+    {
+      "epoch": 0.693053091306128,
+      "grad_norm": 0.23705871403217316,
+      "learning_rate": 8e-05,
+      "loss": 1.5306,
+      "step": 5078
+    },
+    {
+      "epoch": 0.6931895728128838,
+      "grad_norm": 0.2270851880311966,
+      "learning_rate": 8e-05,
+      "loss": 1.4094,
+      "step": 5079
+    },
+    {
+      "epoch": 0.6933260543196397,
+      "grad_norm": 0.2301025688648224,
+      "learning_rate": 8e-05,
+      "loss": 1.4674,
+      "step": 5080
+    },
+    {
+      "epoch": 0.6934625358263955,
+      "grad_norm": 0.22957973182201385,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 5081
+    },
+    {
+      "epoch": 0.6935990173331513,
+      "grad_norm": 0.2330951988697052,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 5082
+    },
+    {
+      "epoch": 0.6937354988399071,
+      "grad_norm": 0.2390143722295761,
+      "learning_rate": 8e-05,
+      "loss": 1.5102,
+      "step": 5083
+    },
+    {
+      "epoch": 0.693871980346663,
+      "grad_norm": 0.22995489835739136,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 5084
+    },
+    {
+      "epoch": 0.6940084618534189,
+      "grad_norm": 0.22315309941768646,
+      "learning_rate": 8e-05,
+      "loss": 1.4708,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6941449433601747,
+      "grad_norm": 0.23628775775432587,
+      "learning_rate": 8e-05,
+      "loss": 1.426,
+      "step": 5086
+    },
+    {
+      "epoch": 0.6942814248669306,
+      "grad_norm": 0.22730235755443573,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 5087
+    },
+    {
+      "epoch": 0.6944179063736864,
+      "grad_norm": 0.22582364082336426,
+      "learning_rate": 8e-05,
+      "loss": 1.3844,
+      "step": 5088
+    },
+    {
+      "epoch": 0.6945543878804422,
+      "grad_norm": 0.22879810631275177,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 5089
+    },
+    {
+      "epoch": 0.694690869387198,
+      "grad_norm": 0.232257679104805,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 5090
+    },
+    {
+      "epoch": 0.6948273508939539,
+      "grad_norm": 0.23786935210227966,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 5091
+    },
+    {
+      "epoch": 0.6949638324007097,
+      "grad_norm": 0.22878646850585938,
+      "learning_rate": 8e-05,
+      "loss": 1.4477,
+      "step": 5092
+    },
+    {
+      "epoch": 0.6951003139074655,
+      "grad_norm": 0.22936205565929413,
+      "learning_rate": 8e-05,
+      "loss": 1.4595,
+      "step": 5093
+    },
+    {
+      "epoch": 0.6952367954142213,
+      "grad_norm": 0.24095378816127777,
+      "learning_rate": 8e-05,
+      "loss": 1.4435,
+      "step": 5094
+    },
+    {
+      "epoch": 0.6953732769209772,
+      "grad_norm": 0.23267251253128052,
+      "learning_rate": 8e-05,
+      "loss": 1.5348,
+      "step": 5095
+    },
+    {
+      "epoch": 0.695509758427733,
+      "grad_norm": 0.2193121463060379,
+      "learning_rate": 8e-05,
+      "loss": 1.4508,
+      "step": 5096
+    },
+    {
+      "epoch": 0.6956462399344889,
+      "grad_norm": 0.22852331399917603,
+      "learning_rate": 8e-05,
+      "loss": 1.4203,
+      "step": 5097
+    },
+    {
+      "epoch": 0.6957827214412448,
+      "grad_norm": 0.23829145729541779,
+      "learning_rate": 8e-05,
+      "loss": 1.469,
+      "step": 5098
+    },
+    {
+      "epoch": 0.6959192029480006,
+      "grad_norm": 0.22598522901535034,
+      "learning_rate": 8e-05,
+      "loss": 1.4179,
+      "step": 5099
+    },
+    {
+      "epoch": 0.6960556844547564,
+      "grad_norm": 0.2247435748577118,
+      "learning_rate": 8e-05,
+      "loss": 1.4394,
+      "step": 5100
+    },
+    {
+      "epoch": 0.6961921659615122,
+      "grad_norm": 0.23520605266094208,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 5101
+    },
+    {
+      "epoch": 0.6963286474682681,
+      "grad_norm": 0.23005187511444092,
+      "learning_rate": 8e-05,
+      "loss": 1.4174,
+      "step": 5102
+    },
+    {
+      "epoch": 0.6964651289750239,
+      "grad_norm": 0.2307494729757309,
+      "learning_rate": 8e-05,
+      "loss": 1.4501,
+      "step": 5103
+    },
+    {
+      "epoch": 0.6966016104817797,
+      "grad_norm": 0.23237387835979462,
+      "learning_rate": 8e-05,
+      "loss": 1.404,
+      "step": 5104
+    },
+    {
+      "epoch": 0.6967380919885355,
+      "grad_norm": 0.22133669257164001,
+      "learning_rate": 8e-05,
+      "loss": 1.4313,
+      "step": 5105
+    },
+    {
+      "epoch": 0.6968745734952914,
+      "grad_norm": 0.22346685826778412,
+      "learning_rate": 8e-05,
+      "loss": 1.4596,
+      "step": 5106
+    },
+    {
+      "epoch": 0.6970110550020472,
+      "grad_norm": 0.2341427057981491,
+      "learning_rate": 8e-05,
+      "loss": 1.4498,
+      "step": 5107
+    },
+    {
+      "epoch": 0.697147536508803,
+      "grad_norm": 0.23348917067050934,
+      "learning_rate": 8e-05,
+      "loss": 1.5226,
+      "step": 5108
+    },
+    {
+      "epoch": 0.6972840180155588,
+      "grad_norm": 0.23101168870925903,
+      "learning_rate": 8e-05,
+      "loss": 1.3767,
+      "step": 5109
+    },
+    {
+      "epoch": 0.6974204995223148,
+      "grad_norm": 0.2369215339422226,
+      "learning_rate": 8e-05,
+      "loss": 1.467,
+      "step": 5110
+    },
+    {
+      "epoch": 0.6975569810290706,
+      "grad_norm": 0.22682523727416992,
+      "learning_rate": 8e-05,
+      "loss": 1.3766,
+      "step": 5111
+    },
+    {
+      "epoch": 0.6976934625358264,
+      "grad_norm": 0.23639442026615143,
+      "learning_rate": 8e-05,
+      "loss": 1.4254,
+      "step": 5112
+    },
+    {
+      "epoch": 0.6978299440425823,
+      "grad_norm": 0.23330754041671753,
+      "learning_rate": 8e-05,
+      "loss": 1.4597,
+      "step": 5113
+    },
+    {
+      "epoch": 0.6979664255493381,
+      "grad_norm": 0.22796671092510223,
+      "learning_rate": 8e-05,
+      "loss": 1.5055,
+      "step": 5114
+    },
+    {
+      "epoch": 0.6981029070560939,
+      "grad_norm": 0.22095134854316711,
+      "learning_rate": 8e-05,
+      "loss": 1.4509,
+      "step": 5115
+    },
+    {
+      "epoch": 0.6982393885628497,
+      "grad_norm": 0.23071900010108948,
+      "learning_rate": 8e-05,
+      "loss": 1.4699,
+      "step": 5116
+    },
+    {
+      "epoch": 0.6983758700696056,
+      "grad_norm": 0.23079606890678406,
+      "learning_rate": 8e-05,
+      "loss": 1.4547,
+      "step": 5117
+    },
+    {
+      "epoch": 0.6985123515763614,
+      "grad_norm": 0.2219523787498474,
+      "learning_rate": 8e-05,
+      "loss": 1.4046,
+      "step": 5118
+    },
+    {
+      "epoch": 0.6986488330831172,
+      "grad_norm": 0.2280607372522354,
+      "learning_rate": 8e-05,
+      "loss": 1.4232,
+      "step": 5119
+    },
+    {
+      "epoch": 0.698785314589873,
+      "grad_norm": 0.22948622703552246,
+      "learning_rate": 8e-05,
+      "loss": 1.4758,
+      "step": 5120
+    },
+    {
+      "epoch": 0.6989217960966289,
+      "grad_norm": 0.22942166030406952,
+      "learning_rate": 8e-05,
+      "loss": 1.4399,
+      "step": 5121
+    },
+    {
+      "epoch": 0.6990582776033848,
+      "grad_norm": 0.23508676886558533,
+      "learning_rate": 8e-05,
+      "loss": 1.4247,
+      "step": 5122
+    },
+    {
+      "epoch": 0.6991947591101406,
+      "grad_norm": 0.23840664327144623,
+      "learning_rate": 8e-05,
+      "loss": 1.4924,
+      "step": 5123
+    },
+    {
+      "epoch": 0.6993312406168964,
+      "grad_norm": 0.23543450236320496,
+      "learning_rate": 8e-05,
+      "loss": 1.4932,
+      "step": 5124
+    },
+    {
+      "epoch": 0.6994677221236523,
+      "grad_norm": 0.2232559323310852,
+      "learning_rate": 8e-05,
+      "loss": 1.4551,
+      "step": 5125
+    },
+    {
+      "epoch": 0.6996042036304081,
+      "grad_norm": 0.2291235327720642,
+      "learning_rate": 8e-05,
+      "loss": 1.4353,
+      "step": 5126
+    },
+    {
+      "epoch": 0.6997406851371639,
+      "grad_norm": 0.22774386405944824,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 5127
+    },
+    {
+      "epoch": 0.6998771666439197,
+      "grad_norm": 0.2351008653640747,
+      "learning_rate": 8e-05,
+      "loss": 1.5581,
+      "step": 5128
+    },
+    {
+      "epoch": 0.7000136481506756,
+      "grad_norm": 0.23261822760105133,
+      "learning_rate": 8e-05,
+      "loss": 1.4796,
+      "step": 5129
+    },
+    {
+      "epoch": 0.7001501296574314,
+      "grad_norm": 0.22495388984680176,
+      "learning_rate": 8e-05,
+      "loss": 1.4327,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7002866111641872,
+      "grad_norm": 0.22172313928604126,
+      "learning_rate": 8e-05,
+      "loss": 1.435,
+      "step": 5131
+    },
+    {
+      "epoch": 0.700423092670943,
+      "grad_norm": 0.2331259548664093,
+      "learning_rate": 8e-05,
+      "loss": 1.5436,
+      "step": 5132
+    },
+    {
+      "epoch": 0.7005595741776989,
+      "grad_norm": 0.2200707495212555,
+      "learning_rate": 8e-05,
+      "loss": 1.435,
+      "step": 5133
+    },
+    {
+      "epoch": 0.7006960556844548,
+      "grad_norm": 0.235848069190979,
+      "learning_rate": 8e-05,
+      "loss": 1.4955,
+      "step": 5134
+    },
+    {
+      "epoch": 0.7008325371912106,
+      "grad_norm": 0.23125916719436646,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 5135
+    },
+    {
+      "epoch": 0.7009690186979665,
+      "grad_norm": 0.23091630637645721,
+      "learning_rate": 8e-05,
+      "loss": 1.4348,
+      "step": 5136
+    },
+    {
+      "epoch": 0.7011055002047223,
+      "grad_norm": 0.22759422659873962,
+      "learning_rate": 8e-05,
+      "loss": 1.342,
+      "step": 5137
+    },
+    {
+      "epoch": 0.7012419817114781,
+      "grad_norm": 0.2229062169790268,
+      "learning_rate": 8e-05,
+      "loss": 1.4567,
+      "step": 5138
+    },
+    {
+      "epoch": 0.7013784632182339,
+      "grad_norm": 0.23501718044281006,
+      "learning_rate": 8e-05,
+      "loss": 1.5329,
+      "step": 5139
+    },
+    {
+      "epoch": 0.7015149447249898,
+      "grad_norm": 0.22614824771881104,
+      "learning_rate": 8e-05,
+      "loss": 1.4227,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7016514262317456,
+      "grad_norm": 0.22627761960029602,
+      "learning_rate": 8e-05,
+      "loss": 1.5146,
+      "step": 5141
+    },
+    {
+      "epoch": 0.7017879077385014,
+      "grad_norm": 0.23504354059696198,
+      "learning_rate": 8e-05,
+      "loss": 1.3903,
+      "step": 5142
+    },
+    {
+      "epoch": 0.7019243892452572,
+      "grad_norm": 0.22979801893234253,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 5143
+    },
+    {
+      "epoch": 0.7020608707520131,
+      "grad_norm": 0.2267935425043106,
+      "learning_rate": 8e-05,
+      "loss": 1.4591,
+      "step": 5144
+    },
+    {
+      "epoch": 0.7021973522587689,
+      "grad_norm": 0.2312496155500412,
+      "learning_rate": 8e-05,
+      "loss": 1.4378,
+      "step": 5145
+    },
+    {
+      "epoch": 0.7023338337655247,
+      "grad_norm": 0.23789344727993011,
+      "learning_rate": 8e-05,
+      "loss": 1.5066,
+      "step": 5146
+    },
+    {
+      "epoch": 0.7024703152722807,
+      "grad_norm": 0.23923785984516144,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 5147
+    },
+    {
+      "epoch": 0.7026067967790365,
+      "grad_norm": 0.22657272219657898,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 5148
+    },
+    {
+      "epoch": 0.7027432782857923,
+      "grad_norm": 0.22614091634750366,
+      "learning_rate": 8e-05,
+      "loss": 1.4455,
+      "step": 5149
+    },
+    {
+      "epoch": 0.7028797597925481,
+      "grad_norm": 0.2270379513502121,
+      "learning_rate": 8e-05,
+      "loss": 1.3821,
+      "step": 5150
+    },
+    {
+      "epoch": 0.703016241299304,
+      "grad_norm": 0.2266760766506195,
+      "learning_rate": 8e-05,
+      "loss": 1.422,
+      "step": 5151
+    },
+    {
+      "epoch": 0.7031527228060598,
+      "grad_norm": 0.2303757220506668,
+      "learning_rate": 8e-05,
+      "loss": 1.4953,
+      "step": 5152
+    },
+    {
+      "epoch": 0.7032892043128156,
+      "grad_norm": 0.23265528678894043,
+      "learning_rate": 8e-05,
+      "loss": 1.4871,
+      "step": 5153
+    },
+    {
+      "epoch": 0.7034256858195714,
+      "grad_norm": 0.23369629681110382,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 5154
+    },
+    {
+      "epoch": 0.7035621673263273,
+      "grad_norm": 0.23297664523124695,
+      "learning_rate": 8e-05,
+      "loss": 1.4636,
+      "step": 5155
+    },
+    {
+      "epoch": 0.7036986488330831,
+      "grad_norm": 0.23646555840969086,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 5156
+    },
+    {
+      "epoch": 0.7038351303398389,
+      "grad_norm": 0.23214687407016754,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 5157
+    },
+    {
+      "epoch": 0.7039716118465947,
+      "grad_norm": 0.22202420234680176,
+      "learning_rate": 8e-05,
+      "loss": 1.3915,
+      "step": 5158
+    },
+    {
+      "epoch": 0.7041080933533507,
+      "grad_norm": 0.23451454937458038,
+      "learning_rate": 8e-05,
+      "loss": 1.4219,
+      "step": 5159
+    },
+    {
+      "epoch": 0.7042445748601065,
+      "grad_norm": 0.226332426071167,
+      "learning_rate": 8e-05,
+      "loss": 1.4657,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7043810563668623,
+      "grad_norm": 0.2316066324710846,
+      "learning_rate": 8e-05,
+      "loss": 1.4743,
+      "step": 5161
+    },
+    {
+      "epoch": 0.7045175378736181,
+      "grad_norm": 0.22673174738883972,
+      "learning_rate": 8e-05,
+      "loss": 1.424,
+      "step": 5162
+    },
+    {
+      "epoch": 0.704654019380374,
+      "grad_norm": 0.22019809484481812,
+      "learning_rate": 8e-05,
+      "loss": 1.4272,
+      "step": 5163
+    },
+    {
+      "epoch": 0.7047905008871298,
+      "grad_norm": 0.2284545749425888,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 5164
+    },
+    {
+      "epoch": 0.7049269823938856,
+      "grad_norm": 0.23567375540733337,
+      "learning_rate": 8e-05,
+      "loss": 1.4536,
+      "step": 5165
+    },
+    {
+      "epoch": 0.7050634639006415,
+      "grad_norm": 0.23670494556427002,
+      "learning_rate": 8e-05,
+      "loss": 1.5239,
+      "step": 5166
+    },
+    {
+      "epoch": 0.7051999454073973,
+      "grad_norm": 0.22246654331684113,
+      "learning_rate": 8e-05,
+      "loss": 1.453,
+      "step": 5167
+    },
+    {
+      "epoch": 0.7053364269141531,
+      "grad_norm": 0.2293093353509903,
+      "learning_rate": 8e-05,
+      "loss": 1.4277,
+      "step": 5168
+    },
+    {
+      "epoch": 0.7054729084209089,
+      "grad_norm": 0.23489068448543549,
+      "learning_rate": 8e-05,
+      "loss": 1.4116,
+      "step": 5169
+    },
+    {
+      "epoch": 0.7056093899276648,
+      "grad_norm": 0.23213842511177063,
+      "learning_rate": 8e-05,
+      "loss": 1.546,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7057458714344207,
+      "grad_norm": 0.2341366857290268,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 5171
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.22749823331832886,
+      "learning_rate": 8e-05,
+      "loss": 1.4477,
+      "step": 5172
+    },
+    {
+      "epoch": 0.7060188344479323,
+      "grad_norm": 0.22021502256393433,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 5173
+    },
+    {
+      "epoch": 0.7061553159546882,
+      "grad_norm": 0.22996768355369568,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 5174
+    },
+    {
+      "epoch": 0.706291797461444,
+      "grad_norm": 0.2389015108346939,
+      "learning_rate": 8e-05,
+      "loss": 1.4699,
+      "step": 5175
+    },
+    {
+      "epoch": 0.7064282789681998,
+      "grad_norm": 0.21842928230762482,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 5176
+    },
+    {
+      "epoch": 0.7065647604749556,
+      "grad_norm": 0.22123534977436066,
+      "learning_rate": 8e-05,
+      "loss": 1.4186,
+      "step": 5177
+    },
+    {
+      "epoch": 0.7067012419817115,
+      "grad_norm": 0.2291850745677948,
+      "learning_rate": 8e-05,
+      "loss": 1.4664,
+      "step": 5178
+    },
+    {
+      "epoch": 0.7068377234884673,
+      "grad_norm": 0.22739991545677185,
+      "learning_rate": 8e-05,
+      "loss": 1.4828,
+      "step": 5179
+    },
+    {
+      "epoch": 0.7069742049952231,
+      "grad_norm": 0.23837895691394806,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 5180
+    },
+    {
+      "epoch": 0.707110686501979,
+      "grad_norm": 0.23669789731502533,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 5181
+    },
+    {
+      "epoch": 0.7072471680087348,
+      "grad_norm": 0.21973150968551636,
+      "learning_rate": 8e-05,
+      "loss": 1.4214,
+      "step": 5182
+    },
+    {
+      "epoch": 0.7073836495154906,
+      "grad_norm": 0.22442220151424408,
+      "learning_rate": 8e-05,
+      "loss": 1.4484,
+      "step": 5183
+    },
+    {
+      "epoch": 0.7075201310222465,
+      "grad_norm": 0.2263316661119461,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 5184
+    },
+    {
+      "epoch": 0.7076566125290024,
+      "grad_norm": 0.2239358276128769,
+      "learning_rate": 8e-05,
+      "loss": 1.4181,
+      "step": 5185
+    },
+    {
+      "epoch": 0.7077930940357582,
+      "grad_norm": 0.22882649302482605,
+      "learning_rate": 8e-05,
+      "loss": 1.4479,
+      "step": 5186
+    },
+    {
+      "epoch": 0.707929575542514,
+      "grad_norm": 0.22912001609802246,
+      "learning_rate": 8e-05,
+      "loss": 1.4405,
+      "step": 5187
+    },
+    {
+      "epoch": 0.7080660570492698,
+      "grad_norm": 0.22843018174171448,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 5188
+    },
+    {
+      "epoch": 0.7082025385560257,
+      "grad_norm": 0.2302873283624649,
+      "learning_rate": 8e-05,
+      "loss": 1.4956,
+      "step": 5189
+    },
+    {
+      "epoch": 0.7083390200627815,
+      "grad_norm": 0.22532938420772552,
+      "learning_rate": 8e-05,
+      "loss": 1.4246,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7084755015695373,
+      "grad_norm": 0.2349798083305359,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 5191
+    },
+    {
+      "epoch": 0.7086119830762931,
+      "grad_norm": 0.22965744137763977,
+      "learning_rate": 8e-05,
+      "loss": 1.4447,
+      "step": 5192
+    },
+    {
+      "epoch": 0.708748464583049,
+      "grad_norm": 0.22220103442668915,
+      "learning_rate": 8e-05,
+      "loss": 1.4446,
+      "step": 5193
+    },
+    {
+      "epoch": 0.7088849460898048,
+      "grad_norm": 0.22468991577625275,
+      "learning_rate": 8e-05,
+      "loss": 1.4164,
+      "step": 5194
+    },
+    {
+      "epoch": 0.7090214275965606,
+      "grad_norm": 0.22321423888206482,
+      "learning_rate": 8e-05,
+      "loss": 1.5125,
+      "step": 5195
+    },
+    {
+      "epoch": 0.7091579091033166,
+      "grad_norm": 0.2211775928735733,
+      "learning_rate": 8e-05,
+      "loss": 1.4781,
+      "step": 5196
+    },
+    {
+      "epoch": 0.7092943906100724,
+      "grad_norm": 0.2264275997877121,
+      "learning_rate": 8e-05,
+      "loss": 1.4121,
+      "step": 5197
+    },
+    {
+      "epoch": 0.7094308721168282,
+      "grad_norm": 0.2266429215669632,
+      "learning_rate": 8e-05,
+      "loss": 1.5035,
+      "step": 5198
+    },
+    {
+      "epoch": 0.709567353623584,
+      "grad_norm": 0.23563487827777863,
+      "learning_rate": 8e-05,
+      "loss": 1.4796,
+      "step": 5199
+    },
+    {
+      "epoch": 0.7097038351303399,
+      "grad_norm": 0.22286291420459747,
+      "learning_rate": 8e-05,
+      "loss": 1.4827,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7098403166370957,
+      "grad_norm": 0.22390563786029816,
+      "learning_rate": 8e-05,
+      "loss": 1.432,
+      "step": 5201
+    },
+    {
+      "epoch": 0.7099767981438515,
+      "grad_norm": 0.2372536063194275,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 5202
+    },
+    {
+      "epoch": 0.7101132796506073,
+      "grad_norm": 0.22790288925170898,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 5203
+    },
+    {
+      "epoch": 0.7102497611573632,
+      "grad_norm": 0.2222626954317093,
+      "learning_rate": 8e-05,
+      "loss": 1.4116,
+      "step": 5204
+    },
+    {
+      "epoch": 0.710386242664119,
+      "grad_norm": 0.230938121676445,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 5205
+    },
+    {
+      "epoch": 0.7105227241708748,
+      "grad_norm": 0.22708426415920258,
+      "learning_rate": 8e-05,
+      "loss": 1.4972,
+      "step": 5206
+    },
+    {
+      "epoch": 0.7106592056776306,
+      "grad_norm": 0.2290227711200714,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 5207
+    },
+    {
+      "epoch": 0.7107956871843866,
+      "grad_norm": 0.2210305631160736,
+      "learning_rate": 8e-05,
+      "loss": 1.4449,
+      "step": 5208
+    },
+    {
+      "epoch": 0.7109321686911424,
+      "grad_norm": 0.22625501453876495,
+      "learning_rate": 8e-05,
+      "loss": 1.4549,
+      "step": 5209
+    },
+    {
+      "epoch": 0.7110686501978982,
+      "grad_norm": 0.2209835946559906,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 5210
+    },
+    {
+      "epoch": 0.711205131704654,
+      "grad_norm": 0.23020268976688385,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 5211
+    },
+    {
+      "epoch": 0.7113416132114099,
+      "grad_norm": 0.22533197700977325,
+      "learning_rate": 8e-05,
+      "loss": 1.4067,
+      "step": 5212
+    },
+    {
+      "epoch": 0.7114780947181657,
+      "grad_norm": 0.22457125782966614,
+      "learning_rate": 8e-05,
+      "loss": 1.4041,
+      "step": 5213
+    },
+    {
+      "epoch": 0.7116145762249215,
+      "grad_norm": 0.22782807052135468,
+      "learning_rate": 8e-05,
+      "loss": 1.4318,
+      "step": 5214
+    },
+    {
+      "epoch": 0.7117510577316774,
+      "grad_norm": 0.2307622879743576,
+      "learning_rate": 8e-05,
+      "loss": 1.4582,
+      "step": 5215
+    },
+    {
+      "epoch": 0.7118875392384332,
+      "grad_norm": 0.23082606494426727,
+      "learning_rate": 8e-05,
+      "loss": 1.4382,
+      "step": 5216
+    },
+    {
+      "epoch": 0.712024020745189,
+      "grad_norm": 0.23211349546909332,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 5217
+    },
+    {
+      "epoch": 0.7121605022519448,
+      "grad_norm": 0.2375563681125641,
+      "learning_rate": 8e-05,
+      "loss": 1.4792,
+      "step": 5218
+    },
+    {
+      "epoch": 0.7122969837587007,
+      "grad_norm": 0.22742272913455963,
+      "learning_rate": 8e-05,
+      "loss": 1.4712,
+      "step": 5219
+    },
+    {
+      "epoch": 0.7124334652654565,
+      "grad_norm": 0.22805550694465637,
+      "learning_rate": 8e-05,
+      "loss": 1.4546,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7125699467722124,
+      "grad_norm": 0.22987519204616547,
+      "learning_rate": 8e-05,
+      "loss": 1.459,
+      "step": 5221
+    },
+    {
+      "epoch": 0.7127064282789682,
+      "grad_norm": 0.22578541934490204,
+      "learning_rate": 8e-05,
+      "loss": 1.4734,
+      "step": 5222
+    },
+    {
+      "epoch": 0.7128429097857241,
+      "grad_norm": 0.2283637672662735,
+      "learning_rate": 8e-05,
+      "loss": 1.4528,
+      "step": 5223
+    },
+    {
+      "epoch": 0.7129793912924799,
+      "grad_norm": 0.2316460758447647,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 5224
+    },
+    {
+      "epoch": 0.7131158727992357,
+      "grad_norm": 0.23687529563903809,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 5225
+    },
+    {
+      "epoch": 0.7132523543059915,
+      "grad_norm": 0.2246427834033966,
+      "learning_rate": 8e-05,
+      "loss": 1.4126,
+      "step": 5226
+    },
+    {
+      "epoch": 0.7133888358127474,
+      "grad_norm": 0.22945912182331085,
+      "learning_rate": 8e-05,
+      "loss": 1.4939,
+      "step": 5227
+    },
+    {
+      "epoch": 0.7135253173195032,
+      "grad_norm": 0.23460642993450165,
+      "learning_rate": 8e-05,
+      "loss": 1.5016,
+      "step": 5228
+    },
+    {
+      "epoch": 0.713661798826259,
+      "grad_norm": 0.23665134608745575,
+      "learning_rate": 8e-05,
+      "loss": 1.4426,
+      "step": 5229
+    },
+    {
+      "epoch": 0.7137982803330148,
+      "grad_norm": 0.22802813351154327,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7139347618397707,
+      "grad_norm": 0.2241210639476776,
+      "learning_rate": 8e-05,
+      "loss": 1.4899,
+      "step": 5231
+    },
+    {
+      "epoch": 0.7140712433465265,
+      "grad_norm": 0.2219220995903015,
+      "learning_rate": 8e-05,
+      "loss": 1.4422,
+      "step": 5232
+    },
+    {
+      "epoch": 0.7142077248532824,
+      "grad_norm": 0.22590434551239014,
+      "learning_rate": 8e-05,
+      "loss": 1.4024,
+      "step": 5233
+    },
+    {
+      "epoch": 0.7143442063600383,
+      "grad_norm": 0.24560531973838806,
+      "learning_rate": 8e-05,
+      "loss": 1.4571,
+      "step": 5234
+    },
+    {
+      "epoch": 0.7144806878667941,
+      "grad_norm": 0.22415179014205933,
+      "learning_rate": 8e-05,
+      "loss": 1.4628,
+      "step": 5235
+    },
+    {
+      "epoch": 0.7146171693735499,
+      "grad_norm": 0.22969307005405426,
+      "learning_rate": 8e-05,
+      "loss": 1.3836,
+      "step": 5236
+    },
+    {
+      "epoch": 0.7147536508803057,
+      "grad_norm": 0.22613166272640228,
+      "learning_rate": 8e-05,
+      "loss": 1.4849,
+      "step": 5237
+    },
+    {
+      "epoch": 0.7148901323870616,
+      "grad_norm": 0.22589626908302307,
+      "learning_rate": 8e-05,
+      "loss": 1.4258,
+      "step": 5238
+    },
+    {
+      "epoch": 0.7150266138938174,
+      "grad_norm": 0.2188994586467743,
+      "learning_rate": 8e-05,
+      "loss": 1.459,
+      "step": 5239
+    },
+    {
+      "epoch": 0.7151630954005732,
+      "grad_norm": 0.2353687733411789,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 5240
+    },
+    {
+      "epoch": 0.715299576907329,
+      "grad_norm": 0.22919295728206635,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 5241
+    },
+    {
+      "epoch": 0.7154360584140849,
+      "grad_norm": 0.2287200391292572,
+      "learning_rate": 8e-05,
+      "loss": 1.4556,
+      "step": 5242
+    },
+    {
+      "epoch": 0.7155725399208407,
+      "grad_norm": 0.22957560420036316,
+      "learning_rate": 8e-05,
+      "loss": 1.4795,
+      "step": 5243
+    },
+    {
+      "epoch": 0.7157090214275965,
+      "grad_norm": 0.2190137803554535,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 5244
+    },
+    {
+      "epoch": 0.7158455029343523,
+      "grad_norm": 0.2272428274154663,
+      "learning_rate": 8e-05,
+      "loss": 1.4394,
+      "step": 5245
+    },
+    {
+      "epoch": 0.7159819844411083,
+      "grad_norm": 0.22814325988292694,
+      "learning_rate": 8e-05,
+      "loss": 1.4357,
+      "step": 5246
+    },
+    {
+      "epoch": 0.7161184659478641,
+      "grad_norm": 0.22830340266227722,
+      "learning_rate": 8e-05,
+      "loss": 1.4706,
+      "step": 5247
+    },
+    {
+      "epoch": 0.7162549474546199,
+      "grad_norm": 0.23010452091693878,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 5248
+    },
+    {
+      "epoch": 0.7163914289613758,
+      "grad_norm": 0.2403780221939087,
+      "learning_rate": 8e-05,
+      "loss": 1.4173,
+      "step": 5249
+    },
+    {
+      "epoch": 0.7165279104681316,
+      "grad_norm": 0.23053985834121704,
+      "learning_rate": 8e-05,
+      "loss": 1.4399,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7166643919748874,
+      "grad_norm": 0.22619538009166718,
+      "learning_rate": 8e-05,
+      "loss": 1.3881,
+      "step": 5251
+    },
+    {
+      "epoch": 0.7168008734816432,
+      "grad_norm": 0.22301580011844635,
+      "learning_rate": 8e-05,
+      "loss": 1.4726,
+      "step": 5252
+    },
+    {
+      "epoch": 0.7169373549883991,
+      "grad_norm": 0.23423010110855103,
+      "learning_rate": 8e-05,
+      "loss": 1.4662,
+      "step": 5253
+    },
+    {
+      "epoch": 0.7170738364951549,
+      "grad_norm": 0.22831673920154572,
+      "learning_rate": 8e-05,
+      "loss": 1.4705,
+      "step": 5254
+    },
+    {
+      "epoch": 0.7172103180019107,
+      "grad_norm": 0.22351470589637756,
+      "learning_rate": 8e-05,
+      "loss": 1.3727,
+      "step": 5255
+    },
+    {
+      "epoch": 0.7173467995086665,
+      "grad_norm": 0.22754192352294922,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 5256
+    },
+    {
+      "epoch": 0.7174832810154224,
+      "grad_norm": 0.23520877957344055,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 5257
+    },
+    {
+      "epoch": 0.7176197625221783,
+      "grad_norm": 0.22898343205451965,
+      "learning_rate": 8e-05,
+      "loss": 1.4936,
+      "step": 5258
+    },
+    {
+      "epoch": 0.7177562440289341,
+      "grad_norm": 0.22621594369411469,
+      "learning_rate": 8e-05,
+      "loss": 1.428,
+      "step": 5259
+    },
+    {
+      "epoch": 0.71789272553569,
+      "grad_norm": 0.2237178385257721,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7180292070424458,
+      "grad_norm": 0.2254728525876999,
+      "learning_rate": 8e-05,
+      "loss": 1.4489,
+      "step": 5261
+    },
+    {
+      "epoch": 0.7181656885492016,
+      "grad_norm": 0.22403603792190552,
+      "learning_rate": 8e-05,
+      "loss": 1.4413,
+      "step": 5262
+    },
+    {
+      "epoch": 0.7183021700559574,
+      "grad_norm": 0.23477238416671753,
+      "learning_rate": 8e-05,
+      "loss": 1.4863,
+      "step": 5263
+    },
+    {
+      "epoch": 0.7184386515627132,
+      "grad_norm": 0.23135671019554138,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 5264
+    },
+    {
+      "epoch": 0.7185751330694691,
+      "grad_norm": 0.22708939015865326,
+      "learning_rate": 8e-05,
+      "loss": 1.4665,
+      "step": 5265
+    },
+    {
+      "epoch": 0.7187116145762249,
+      "grad_norm": 0.2184794843196869,
+      "learning_rate": 8e-05,
+      "loss": 1.4556,
+      "step": 5266
+    },
+    {
+      "epoch": 0.7188480960829807,
+      "grad_norm": 0.23387610912322998,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 5267
+    },
+    {
+      "epoch": 0.7189845775897366,
+      "grad_norm": 0.23199856281280518,
+      "learning_rate": 8e-05,
+      "loss": 1.4596,
+      "step": 5268
+    },
+    {
+      "epoch": 0.7191210590964924,
+      "grad_norm": 0.23771816492080688,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 5269
+    },
+    {
+      "epoch": 0.7192575406032483,
+      "grad_norm": 0.2327917218208313,
+      "learning_rate": 8e-05,
+      "loss": 1.5286,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7193940221100041,
+      "grad_norm": 0.2535395324230194,
+      "learning_rate": 8e-05,
+      "loss": 1.4092,
+      "step": 5271
+    },
+    {
+      "epoch": 0.71953050361676,
+      "grad_norm": 0.23165638744831085,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 5272
+    },
+    {
+      "epoch": 0.7196669851235158,
+      "grad_norm": 0.22785091400146484,
+      "learning_rate": 8e-05,
+      "loss": 1.4575,
+      "step": 5273
+    },
+    {
+      "epoch": 0.7198034666302716,
+      "grad_norm": 0.22400890290737152,
+      "learning_rate": 8e-05,
+      "loss": 1.418,
+      "step": 5274
+    },
+    {
+      "epoch": 0.7199399481370274,
+      "grad_norm": 0.23173034191131592,
+      "learning_rate": 8e-05,
+      "loss": 1.4596,
+      "step": 5275
+    },
+    {
+      "epoch": 0.7200764296437833,
+      "grad_norm": 0.23364093899726868,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 5276
+    },
+    {
+      "epoch": 0.7202129111505391,
+      "grad_norm": 0.2318466454744339,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 5277
+    },
+    {
+      "epoch": 0.7203493926572949,
+      "grad_norm": 0.2363099604845047,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 5278
+    },
+    {
+      "epoch": 0.7204858741640507,
+      "grad_norm": 0.22352993488311768,
+      "learning_rate": 8e-05,
+      "loss": 1.4525,
+      "step": 5279
+    },
+    {
+      "epoch": 0.7206223556708066,
+      "grad_norm": 0.23048235476016998,
+      "learning_rate": 8e-05,
+      "loss": 1.4913,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7207588371775624,
+      "grad_norm": 0.23505954444408417,
+      "learning_rate": 8e-05,
+      "loss": 1.4876,
+      "step": 5281
+    },
+    {
+      "epoch": 0.7208953186843182,
+      "grad_norm": 0.24082958698272705,
+      "learning_rate": 8e-05,
+      "loss": 1.482,
+      "step": 5282
+    },
+    {
+      "epoch": 0.7210318001910742,
+      "grad_norm": 0.23025617003440857,
+      "learning_rate": 8e-05,
+      "loss": 1.4682,
+      "step": 5283
+    },
+    {
+      "epoch": 0.72116828169783,
+      "grad_norm": 0.23293665051460266,
+      "learning_rate": 8e-05,
+      "loss": 1.4987,
+      "step": 5284
+    },
+    {
+      "epoch": 0.7213047632045858,
+      "grad_norm": 0.22668147087097168,
+      "learning_rate": 8e-05,
+      "loss": 1.4206,
+      "step": 5285
+    },
+    {
+      "epoch": 0.7214412447113416,
+      "grad_norm": 0.23604561388492584,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 5286
+    },
+    {
+      "epoch": 0.7215777262180975,
+      "grad_norm": 0.22835572063922882,
+      "learning_rate": 8e-05,
+      "loss": 1.4464,
+      "step": 5287
+    },
+    {
+      "epoch": 0.7217142077248533,
+      "grad_norm": 0.23009514808654785,
+      "learning_rate": 8e-05,
+      "loss": 1.4808,
+      "step": 5288
+    },
+    {
+      "epoch": 0.7218506892316091,
+      "grad_norm": 0.228826105594635,
+      "learning_rate": 8e-05,
+      "loss": 1.3958,
+      "step": 5289
+    },
+    {
+      "epoch": 0.7219871707383649,
+      "grad_norm": 0.23606547713279724,
+      "learning_rate": 8e-05,
+      "loss": 1.4428,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7221236522451208,
+      "grad_norm": 0.230474054813385,
+      "learning_rate": 8e-05,
+      "loss": 1.4446,
+      "step": 5291
+    },
+    {
+      "epoch": 0.7222601337518766,
+      "grad_norm": 0.23692047595977783,
+      "learning_rate": 8e-05,
+      "loss": 1.4852,
+      "step": 5292
+    },
+    {
+      "epoch": 0.7223966152586324,
+      "grad_norm": 0.2328113317489624,
+      "learning_rate": 8e-05,
+      "loss": 1.4437,
+      "step": 5293
+    },
+    {
+      "epoch": 0.7225330967653882,
+      "grad_norm": 0.22825218737125397,
+      "learning_rate": 8e-05,
+      "loss": 1.3887,
+      "step": 5294
+    },
+    {
+      "epoch": 0.7226695782721442,
+      "grad_norm": 0.23564966022968292,
+      "learning_rate": 8e-05,
+      "loss": 1.4058,
+      "step": 5295
+    },
+    {
+      "epoch": 0.7228060597789,
+      "grad_norm": 0.23616553843021393,
+      "learning_rate": 8e-05,
+      "loss": 1.4228,
+      "step": 5296
+    },
+    {
+      "epoch": 0.7229425412856558,
+      "grad_norm": 0.23622408509254456,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 5297
+    },
+    {
+      "epoch": 0.7230790227924117,
+      "grad_norm": 0.2299550473690033,
+      "learning_rate": 8e-05,
+      "loss": 1.4817,
+      "step": 5298
+    },
+    {
+      "epoch": 0.7232155042991675,
+      "grad_norm": 0.22760528326034546,
+      "learning_rate": 8e-05,
+      "loss": 1.4367,
+      "step": 5299
+    },
+    {
+      "epoch": 0.7233519858059233,
+      "grad_norm": 0.2318352311849594,
+      "learning_rate": 8e-05,
+      "loss": 1.421,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7234884673126791,
+      "grad_norm": 0.23360735177993774,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 5301
+    },
+    {
+      "epoch": 0.723624948819435,
+      "grad_norm": 0.23550571501255035,
+      "learning_rate": 8e-05,
+      "loss": 1.4587,
+      "step": 5302
+    },
+    {
+      "epoch": 0.7237614303261908,
+      "grad_norm": 0.23339222371578217,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 5303
+    },
+    {
+      "epoch": 0.7238979118329466,
+      "grad_norm": 0.22132667899131775,
+      "learning_rate": 8e-05,
+      "loss": 1.4549,
+      "step": 5304
+    },
+    {
+      "epoch": 0.7240343933397024,
+      "grad_norm": 0.22522932291030884,
+      "learning_rate": 8e-05,
+      "loss": 1.4377,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7241708748464583,
+      "grad_norm": 0.2330664098262787,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7243073563532142,
+      "grad_norm": 0.24181297421455383,
+      "learning_rate": 8e-05,
+      "loss": 1.535,
+      "step": 5307
+    },
+    {
+      "epoch": 0.72444383785997,
+      "grad_norm": 0.23423512279987335,
+      "learning_rate": 8e-05,
+      "loss": 1.5116,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7245803193667258,
+      "grad_norm": 0.22781367599964142,
+      "learning_rate": 8e-05,
+      "loss": 1.4697,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7247168008734817,
+      "grad_norm": 0.2365158647298813,
+      "learning_rate": 8e-05,
+      "loss": 1.3836,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7248532823802375,
+      "grad_norm": 0.22710949182510376,
+      "learning_rate": 8e-05,
+      "loss": 1.4236,
+      "step": 5311
+    },
+    {
+      "epoch": 0.7249897638869933,
+      "grad_norm": 0.23008687794208527,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7251262453937491,
+      "grad_norm": 0.23265910148620605,
+      "learning_rate": 8e-05,
+      "loss": 1.4328,
+      "step": 5313
+    },
+    {
+      "epoch": 0.725262726900505,
+      "grad_norm": 0.24594861268997192,
+      "learning_rate": 8e-05,
+      "loss": 1.5129,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7253992084072608,
+      "grad_norm": 0.22055862843990326,
+      "learning_rate": 8e-05,
+      "loss": 1.414,
+      "step": 5315
+    },
+    {
+      "epoch": 0.7255356899140166,
+      "grad_norm": 0.24171856045722961,
+      "learning_rate": 8e-05,
+      "loss": 1.5024,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7256721714207724,
+      "grad_norm": 0.23419013619422913,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7258086529275283,
+      "grad_norm": 0.24209046363830566,
+      "learning_rate": 8e-05,
+      "loss": 1.4844,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7259451344342841,
+      "grad_norm": 0.2303965985774994,
+      "learning_rate": 8e-05,
+      "loss": 1.5032,
+      "step": 5319
+    },
+    {
+      "epoch": 0.72608161594104,
+      "grad_norm": 0.22313524782657623,
+      "learning_rate": 8e-05,
+      "loss": 1.455,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7262180974477959,
+      "grad_norm": 0.24665449559688568,
+      "learning_rate": 8e-05,
+      "loss": 1.5182,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7263545789545517,
+      "grad_norm": 0.22862735390663147,
+      "learning_rate": 8e-05,
+      "loss": 1.427,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7264910604613075,
+      "grad_norm": 0.23640479147434235,
+      "learning_rate": 8e-05,
+      "loss": 1.4216,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7266275419680633,
+      "grad_norm": 0.23321279883384705,
+      "learning_rate": 8e-05,
+      "loss": 1.4868,
+      "step": 5324
+    },
+    {
+      "epoch": 0.7267640234748192,
+      "grad_norm": 0.24157695472240448,
+      "learning_rate": 8e-05,
+      "loss": 1.488,
+      "step": 5325
+    },
+    {
+      "epoch": 0.726900504981575,
+      "grad_norm": 0.2316184937953949,
+      "learning_rate": 8e-05,
+      "loss": 1.4296,
+      "step": 5326
+    },
+    {
+      "epoch": 0.7270369864883308,
+      "grad_norm": 0.23264643549919128,
+      "learning_rate": 8e-05,
+      "loss": 1.4218,
+      "step": 5327
+    },
+    {
+      "epoch": 0.7271734679950866,
+      "grad_norm": 0.23049168288707733,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7273099495018425,
+      "grad_norm": 0.23966392874717712,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7274464310085983,
+      "grad_norm": 0.23343442380428314,
+      "learning_rate": 8e-05,
+      "loss": 1.5026,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7275829125153541,
+      "grad_norm": 0.2502121031284332,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 5331
+    },
+    {
+      "epoch": 0.72771939402211,
+      "grad_norm": 0.2411840260028839,
+      "learning_rate": 8e-05,
+      "loss": 1.4821,
+      "step": 5332
+    },
+    {
+      "epoch": 0.7278558755288659,
+      "grad_norm": 0.22902125120162964,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7279923570356217,
+      "grad_norm": 0.23181374371051788,
+      "learning_rate": 8e-05,
+      "loss": 1.4741,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7281288385423775,
+      "grad_norm": 0.23361146450042725,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7282653200491334,
+      "grad_norm": 0.22817663848400116,
+      "learning_rate": 8e-05,
+      "loss": 1.438,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7284018015558892,
+      "grad_norm": 0.23050130903720856,
+      "learning_rate": 8e-05,
+      "loss": 1.4782,
+      "step": 5337
+    },
+    {
+      "epoch": 0.728538283062645,
+      "grad_norm": 0.22238372266292572,
+      "learning_rate": 8e-05,
+      "loss": 1.3356,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7286747645694008,
+      "grad_norm": 0.24139800667762756,
+      "learning_rate": 8e-05,
+      "loss": 1.4926,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7288112460761567,
+      "grad_norm": 0.2403285801410675,
+      "learning_rate": 8e-05,
+      "loss": 1.4118,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7289477275829125,
+      "grad_norm": 0.23192352056503296,
+      "learning_rate": 8e-05,
+      "loss": 1.4249,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7290842090896683,
+      "grad_norm": 0.23252302408218384,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7292206905964241,
+      "grad_norm": 0.22691796720027924,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7293571721031801,
+      "grad_norm": 0.23446527123451233,
+      "learning_rate": 8e-05,
+      "loss": 1.5143,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7294936536099359,
+      "grad_norm": 0.23514872789382935,
+      "learning_rate": 8e-05,
+      "loss": 1.4314,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7296301351166917,
+      "grad_norm": 0.23052138090133667,
+      "learning_rate": 8e-05,
+      "loss": 1.41,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7297666166234476,
+      "grad_norm": 0.22796691954135895,
+      "learning_rate": 8e-05,
+      "loss": 1.4602,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7299030981302034,
+      "grad_norm": 0.22415000200271606,
+      "learning_rate": 8e-05,
+      "loss": 1.472,
+      "step": 5348
+    },
+    {
+      "epoch": 0.7300395796369592,
+      "grad_norm": 0.22221362590789795,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 5349
+    },
+    {
+      "epoch": 0.730176061143715,
+      "grad_norm": 0.22684365510940552,
+      "learning_rate": 8e-05,
+      "loss": 1.4489,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7303125426504709,
+      "grad_norm": 0.2365785390138626,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 5351
+    },
+    {
+      "epoch": 0.7304490241572267,
+      "grad_norm": 0.2281636893749237,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 5352
+    },
+    {
+      "epoch": 0.7305855056639825,
+      "grad_norm": 0.23404833674430847,
+      "learning_rate": 8e-05,
+      "loss": 1.4003,
+      "step": 5353
+    },
+    {
+      "epoch": 0.7307219871707383,
+      "grad_norm": 0.24050290882587433,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 5354
+    },
+    {
+      "epoch": 0.7308584686774942,
+      "grad_norm": 0.24188768863677979,
+      "learning_rate": 8e-05,
+      "loss": 1.507,
+      "step": 5355
+    },
+    {
+      "epoch": 0.73099495018425,
+      "grad_norm": 0.24535927176475525,
+      "learning_rate": 8e-05,
+      "loss": 1.5357,
+      "step": 5356
+    },
+    {
+      "epoch": 0.7311314316910059,
+      "grad_norm": 0.22273485362529755,
+      "learning_rate": 8e-05,
+      "loss": 1.401,
+      "step": 5357
+    },
+    {
+      "epoch": 0.7312679131977617,
+      "grad_norm": 0.22941969335079193,
+      "learning_rate": 8e-05,
+      "loss": 1.4562,
+      "step": 5358
+    },
+    {
+      "epoch": 0.7314043947045176,
+      "grad_norm": 0.24198800325393677,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 5359
+    },
+    {
+      "epoch": 0.7315408762112734,
+      "grad_norm": 0.22070355713367462,
+      "learning_rate": 8e-05,
+      "loss": 1.3686,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7316773577180292,
+      "grad_norm": 0.2219826877117157,
+      "learning_rate": 8e-05,
+      "loss": 1.436,
+      "step": 5361
+    },
+    {
+      "epoch": 0.731813839224785,
+      "grad_norm": 0.23151229321956635,
+      "learning_rate": 8e-05,
+      "loss": 1.4796,
+      "step": 5362
+    },
+    {
+      "epoch": 0.7319503207315409,
+      "grad_norm": 0.2341909110546112,
+      "learning_rate": 8e-05,
+      "loss": 1.4341,
+      "step": 5363
+    },
+    {
+      "epoch": 0.7320868022382967,
+      "grad_norm": 0.23004098236560822,
+      "learning_rate": 8e-05,
+      "loss": 1.4468,
+      "step": 5364
+    },
+    {
+      "epoch": 0.7322232837450525,
+      "grad_norm": 0.23812565207481384,
+      "learning_rate": 8e-05,
+      "loss": 1.499,
+      "step": 5365
+    },
+    {
+      "epoch": 0.7323597652518083,
+      "grad_norm": 0.23968151211738586,
+      "learning_rate": 8e-05,
+      "loss": 1.4617,
+      "step": 5366
+    },
+    {
+      "epoch": 0.7324962467585642,
+      "grad_norm": 0.25700920820236206,
+      "learning_rate": 8e-05,
+      "loss": 1.364,
+      "step": 5367
+    },
+    {
+      "epoch": 0.73263272826532,
+      "grad_norm": 0.2526964545249939,
+      "learning_rate": 8e-05,
+      "loss": 1.4054,
+      "step": 5368
+    },
+    {
+      "epoch": 0.7327692097720759,
+      "grad_norm": 0.22964175045490265,
+      "learning_rate": 8e-05,
+      "loss": 1.4354,
+      "step": 5369
+    },
+    {
+      "epoch": 0.7329056912788318,
+      "grad_norm": 0.24745650589466095,
+      "learning_rate": 8e-05,
+      "loss": 1.4954,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7330421727855876,
+      "grad_norm": 0.23828831315040588,
+      "learning_rate": 8e-05,
+      "loss": 1.3966,
+      "step": 5371
+    },
+    {
+      "epoch": 0.7331786542923434,
+      "grad_norm": 0.23327599465847015,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 5372
+    },
+    {
+      "epoch": 0.7333151357990992,
+      "grad_norm": 0.24403904378414154,
+      "learning_rate": 8e-05,
+      "loss": 1.4735,
+      "step": 5373
+    },
+    {
+      "epoch": 0.7334516173058551,
+      "grad_norm": 0.24699600040912628,
+      "learning_rate": 8e-05,
+      "loss": 1.4378,
+      "step": 5374
+    },
+    {
+      "epoch": 0.7335880988126109,
+      "grad_norm": 0.23798972368240356,
+      "learning_rate": 8e-05,
+      "loss": 1.4612,
+      "step": 5375
+    },
+    {
+      "epoch": 0.7337245803193667,
+      "grad_norm": 0.2450876235961914,
+      "learning_rate": 8e-05,
+      "loss": 1.4835,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7338610618261225,
+      "grad_norm": 0.24554628133773804,
+      "learning_rate": 8e-05,
+      "loss": 1.4479,
+      "step": 5377
+    },
+    {
+      "epoch": 0.7339975433328784,
+      "grad_norm": 0.22485744953155518,
+      "learning_rate": 8e-05,
+      "loss": 1.4683,
+      "step": 5378
+    },
+    {
+      "epoch": 0.7341340248396342,
+      "grad_norm": 0.23831279575824738,
+      "learning_rate": 8e-05,
+      "loss": 1.4735,
+      "step": 5379
+    },
+    {
+      "epoch": 0.73427050634639,
+      "grad_norm": 0.22268791496753693,
+      "learning_rate": 8e-05,
+      "loss": 1.4448,
+      "step": 5380
+    },
+    {
+      "epoch": 0.734406987853146,
+      "grad_norm": 0.23480865359306335,
+      "learning_rate": 8e-05,
+      "loss": 1.4468,
+      "step": 5381
+    },
+    {
+      "epoch": 0.7345434693599018,
+      "grad_norm": 0.22699372470378876,
+      "learning_rate": 8e-05,
+      "loss": 1.4347,
+      "step": 5382
+    },
+    {
+      "epoch": 0.7346799508666576,
+      "grad_norm": 0.23383373022079468,
+      "learning_rate": 8e-05,
+      "loss": 1.4088,
+      "step": 5383
+    },
+    {
+      "epoch": 0.7348164323734134,
+      "grad_norm": 0.22804340720176697,
+      "learning_rate": 8e-05,
+      "loss": 1.5105,
+      "step": 5384
+    },
+    {
+      "epoch": 0.7349529138801693,
+      "grad_norm": 0.22474032640457153,
+      "learning_rate": 8e-05,
+      "loss": 1.3943,
+      "step": 5385
+    },
+    {
+      "epoch": 0.7350893953869251,
+      "grad_norm": 0.2366325855255127,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 5386
+    },
+    {
+      "epoch": 0.7352258768936809,
+      "grad_norm": 0.2294752597808838,
+      "learning_rate": 8e-05,
+      "loss": 1.5151,
+      "step": 5387
+    },
+    {
+      "epoch": 0.7353623584004367,
+      "grad_norm": 0.23280516266822815,
+      "learning_rate": 8e-05,
+      "loss": 1.4692,
+      "step": 5388
+    },
+    {
+      "epoch": 0.7354988399071926,
+      "grad_norm": 0.23477290570735931,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 5389
+    },
+    {
+      "epoch": 0.7356353214139484,
+      "grad_norm": 0.22921539843082428,
+      "learning_rate": 8e-05,
+      "loss": 1.3693,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7357718029207042,
+      "grad_norm": 0.23668691515922546,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 5391
+    },
+    {
+      "epoch": 0.73590828442746,
+      "grad_norm": 0.22349058091640472,
+      "learning_rate": 8e-05,
+      "loss": 1.3654,
+      "step": 5392
+    },
+    {
+      "epoch": 0.7360447659342159,
+      "grad_norm": 0.23499074578285217,
+      "learning_rate": 8e-05,
+      "loss": 1.4176,
+      "step": 5393
+    },
+    {
+      "epoch": 0.7361812474409718,
+      "grad_norm": 0.23919731378555298,
+      "learning_rate": 8e-05,
+      "loss": 1.4556,
+      "step": 5394
+    },
+    {
+      "epoch": 0.7363177289477276,
+      "grad_norm": 0.22556236386299133,
+      "learning_rate": 8e-05,
+      "loss": 1.4111,
+      "step": 5395
+    },
+    {
+      "epoch": 0.7364542104544834,
+      "grad_norm": 0.2295866310596466,
+      "learning_rate": 8e-05,
+      "loss": 1.4284,
+      "step": 5396
+    },
+    {
+      "epoch": 0.7365906919612393,
+      "grad_norm": 0.23728801310062408,
+      "learning_rate": 8e-05,
+      "loss": 1.4455,
+      "step": 5397
+    },
+    {
+      "epoch": 0.7367271734679951,
+      "grad_norm": 0.2346675843000412,
+      "learning_rate": 8e-05,
+      "loss": 1.479,
+      "step": 5398
+    },
+    {
+      "epoch": 0.7368636549747509,
+      "grad_norm": 0.2296382188796997,
+      "learning_rate": 8e-05,
+      "loss": 1.3871,
+      "step": 5399
+    },
+    {
+      "epoch": 0.7370001364815068,
+      "grad_norm": 0.23015180230140686,
+      "learning_rate": 8e-05,
+      "loss": 1.4386,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7371366179882626,
+      "grad_norm": 0.22741682827472687,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 5401
+    },
+    {
+      "epoch": 0.7372730994950184,
+      "grad_norm": 0.2277875393629074,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 5402
+    },
+    {
+      "epoch": 0.7374095810017742,
+      "grad_norm": 0.2306147962808609,
+      "learning_rate": 8e-05,
+      "loss": 1.4095,
+      "step": 5403
+    },
+    {
+      "epoch": 0.73754606250853,
+      "grad_norm": 0.24104784429073334,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 5404
+    },
+    {
+      "epoch": 0.7376825440152859,
+      "grad_norm": 0.22377142310142517,
+      "learning_rate": 8e-05,
+      "loss": 1.4198,
+      "step": 5405
+    },
+    {
+      "epoch": 0.7378190255220418,
+      "grad_norm": 0.22960709035396576,
+      "learning_rate": 8e-05,
+      "loss": 1.4555,
+      "step": 5406
+    },
+    {
+      "epoch": 0.7379555070287976,
+      "grad_norm": 0.24812842905521393,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 5407
+    },
+    {
+      "epoch": 0.7380919885355535,
+      "grad_norm": 0.23960065841674805,
+      "learning_rate": 8e-05,
+      "loss": 1.4699,
+      "step": 5408
+    },
+    {
+      "epoch": 0.7382284700423093,
+      "grad_norm": 0.23155266046524048,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 5409
+    },
+    {
+      "epoch": 0.7383649515490651,
+      "grad_norm": 0.23349325358867645,
+      "learning_rate": 8e-05,
+      "loss": 1.4345,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7385014330558209,
+      "grad_norm": 0.23217438161373138,
+      "learning_rate": 8e-05,
+      "loss": 1.3821,
+      "step": 5411
+    },
+    {
+      "epoch": 0.7386379145625768,
+      "grad_norm": 0.2326105386018753,
+      "learning_rate": 8e-05,
+      "loss": 1.3979,
+      "step": 5412
+    },
+    {
+      "epoch": 0.7387743960693326,
+      "grad_norm": 0.23590391874313354,
+      "learning_rate": 8e-05,
+      "loss": 1.4757,
+      "step": 5413
+    },
+    {
+      "epoch": 0.7389108775760884,
+      "grad_norm": 0.2365354746580124,
+      "learning_rate": 8e-05,
+      "loss": 1.4366,
+      "step": 5414
+    },
+    {
+      "epoch": 0.7390473590828442,
+      "grad_norm": 0.2290075719356537,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 5415
+    },
+    {
+      "epoch": 0.7391838405896001,
+      "grad_norm": 0.2313995510339737,
+      "learning_rate": 8e-05,
+      "loss": 1.4226,
+      "step": 5416
+    },
+    {
+      "epoch": 0.7393203220963559,
+      "grad_norm": 0.23147158324718475,
+      "learning_rate": 8e-05,
+      "loss": 1.3896,
+      "step": 5417
+    },
+    {
+      "epoch": 0.7394568036031117,
+      "grad_norm": 0.22454842925071716,
+      "learning_rate": 8e-05,
+      "loss": 1.403,
+      "step": 5418
+    },
+    {
+      "epoch": 0.7395932851098677,
+      "grad_norm": 0.23005039989948273,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 5419
+    },
+    {
+      "epoch": 0.7397297666166235,
+      "grad_norm": 0.23498544096946716,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 5420
+    },
+    {
+      "epoch": 0.7398662481233793,
+      "grad_norm": 0.2305397093296051,
+      "learning_rate": 8e-05,
+      "loss": 1.4183,
+      "step": 5421
+    },
+    {
+      "epoch": 0.7400027296301351,
+      "grad_norm": 0.2323109656572342,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 5422
+    },
+    {
+      "epoch": 0.740139211136891,
+      "grad_norm": 0.2385823130607605,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 5423
+    },
+    {
+      "epoch": 0.7402756926436468,
+      "grad_norm": 0.24378721415996552,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 5424
+    },
+    {
+      "epoch": 0.7404121741504026,
+      "grad_norm": 0.22923608124256134,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 5425
+    },
+    {
+      "epoch": 0.7405486556571584,
+      "grad_norm": 0.23621587455272675,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 5426
+    },
+    {
+      "epoch": 0.7406851371639143,
+      "grad_norm": 0.22832398116588593,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 5427
+    },
+    {
+      "epoch": 0.7408216186706701,
+      "grad_norm": 0.2290707677602768,
+      "learning_rate": 8e-05,
+      "loss": 1.3588,
+      "step": 5428
+    },
+    {
+      "epoch": 0.7409581001774259,
+      "grad_norm": 0.22947391867637634,
+      "learning_rate": 8e-05,
+      "loss": 1.4978,
+      "step": 5429
+    },
+    {
+      "epoch": 0.7410945816841817,
+      "grad_norm": 0.23131950199604034,
+      "learning_rate": 8e-05,
+      "loss": 1.4231,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7412310631909377,
+      "grad_norm": 0.23132729530334473,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 5431
+    },
+    {
+      "epoch": 0.7413675446976935,
+      "grad_norm": 0.23361927270889282,
+      "learning_rate": 8e-05,
+      "loss": 1.3946,
+      "step": 5432
+    },
+    {
+      "epoch": 0.7415040262044493,
+      "grad_norm": 0.23760920763015747,
+      "learning_rate": 8e-05,
+      "loss": 1.4802,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7416405077112052,
+      "grad_norm": 0.23963181674480438,
+      "learning_rate": 8e-05,
+      "loss": 1.5049,
+      "step": 5434
+    },
+    {
+      "epoch": 0.741776989217961,
+      "grad_norm": 0.2447018325328827,
+      "learning_rate": 8e-05,
+      "loss": 1.5161,
+      "step": 5435
+    },
+    {
+      "epoch": 0.7419134707247168,
+      "grad_norm": 0.2312150001525879,
+      "learning_rate": 8e-05,
+      "loss": 1.4247,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7420499522314726,
+      "grad_norm": 0.2264053374528885,
+      "learning_rate": 8e-05,
+      "loss": 1.4915,
+      "step": 5437
+    },
+    {
+      "epoch": 0.7421864337382285,
+      "grad_norm": 0.22064529359340668,
+      "learning_rate": 8e-05,
+      "loss": 1.378,
+      "step": 5438
+    },
+    {
+      "epoch": 0.7423229152449843,
+      "grad_norm": 0.2286280393600464,
+      "learning_rate": 8e-05,
+      "loss": 1.4908,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7424593967517401,
+      "grad_norm": 0.23819568753242493,
+      "learning_rate": 8e-05,
+      "loss": 1.492,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7425958782584959,
+      "grad_norm": 0.22619980573654175,
+      "learning_rate": 8e-05,
+      "loss": 1.3947,
+      "step": 5441
+    },
+    {
+      "epoch": 0.7427323597652518,
+      "grad_norm": 0.23072820901870728,
+      "learning_rate": 8e-05,
+      "loss": 1.4505,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7428688412720077,
+      "grad_norm": 0.2268265187740326,
+      "learning_rate": 8e-05,
+      "loss": 1.4687,
+      "step": 5443
+    },
+    {
+      "epoch": 0.7430053227787635,
+      "grad_norm": 0.23408031463623047,
+      "learning_rate": 8e-05,
+      "loss": 1.4624,
+      "step": 5444
+    },
+    {
+      "epoch": 0.7431418042855193,
+      "grad_norm": 0.22525933384895325,
+      "learning_rate": 8e-05,
+      "loss": 1.4429,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7432782857922752,
+      "grad_norm": 0.2279914766550064,
+      "learning_rate": 8e-05,
+      "loss": 1.4176,
+      "step": 5446
+    },
+    {
+      "epoch": 0.743414767299031,
+      "grad_norm": 0.22487708926200867,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 5447
+    },
+    {
+      "epoch": 0.7435512488057868,
+      "grad_norm": 0.2260654866695404,
+      "learning_rate": 8e-05,
+      "loss": 1.4824,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7436877303125426,
+      "grad_norm": 0.23101775348186493,
+      "learning_rate": 8e-05,
+      "loss": 1.4132,
+      "step": 5449
+    },
+    {
+      "epoch": 0.7438242118192985,
+      "grad_norm": 0.2334427684545517,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7439606933260543,
+      "grad_norm": 0.2328389436006546,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7440971748328101,
+      "grad_norm": 0.21947164833545685,
+      "learning_rate": 8e-05,
+      "loss": 1.403,
+      "step": 5452
+    },
+    {
+      "epoch": 0.744233656339566,
+      "grad_norm": 0.23512783646583557,
+      "learning_rate": 8e-05,
+      "loss": 1.4572,
+      "step": 5453
+    },
+    {
+      "epoch": 0.7443701378463218,
+      "grad_norm": 0.2520940899848938,
+      "learning_rate": 8e-05,
+      "loss": 1.4301,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7445066193530776,
+      "grad_norm": 0.23764021694660187,
+      "learning_rate": 8e-05,
+      "loss": 1.4144,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7446431008598335,
+      "grad_norm": 0.24062025547027588,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7447795823665894,
+      "grad_norm": 0.23779985308647156,
+      "learning_rate": 8e-05,
+      "loss": 1.4325,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7449160638733452,
+      "grad_norm": 0.22479653358459473,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 5458
+    },
+    {
+      "epoch": 0.745052545380101,
+      "grad_norm": 0.22232404351234436,
+      "learning_rate": 8e-05,
+      "loss": 1.3752,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7451890268868568,
+      "grad_norm": 0.23947718739509583,
+      "learning_rate": 8e-05,
+      "loss": 1.4902,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7453255083936127,
+      "grad_norm": 0.24322591722011566,
+      "learning_rate": 8e-05,
+      "loss": 1.4603,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7454619899003685,
+      "grad_norm": 0.23918889462947845,
+      "learning_rate": 8e-05,
+      "loss": 1.4642,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7455984714071243,
+      "grad_norm": 0.22220833599567413,
+      "learning_rate": 8e-05,
+      "loss": 1.3353,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7457349529138801,
+      "grad_norm": 0.23220521211624146,
+      "learning_rate": 8e-05,
+      "loss": 1.4376,
+      "step": 5464
+    },
+    {
+      "epoch": 0.745871434420636,
+      "grad_norm": 0.22554932534694672,
+      "learning_rate": 8e-05,
+      "loss": 1.4544,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7460079159273918,
+      "grad_norm": 0.2210371196269989,
+      "learning_rate": 8e-05,
+      "loss": 1.4403,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7461443974341476,
+      "grad_norm": 0.23625583946704865,
+      "learning_rate": 8e-05,
+      "loss": 1.4542,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7462808789409036,
+      "grad_norm": 0.23824359476566315,
+      "learning_rate": 8e-05,
+      "loss": 1.4148,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7464173604476594,
+      "grad_norm": 0.22984260320663452,
+      "learning_rate": 8e-05,
+      "loss": 1.4721,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7465538419544152,
+      "grad_norm": 0.23944777250289917,
+      "learning_rate": 8e-05,
+      "loss": 1.5524,
+      "step": 5470
+    },
+    {
+      "epoch": 0.746690323461171,
+      "grad_norm": 0.2285875380039215,
+      "learning_rate": 8e-05,
+      "loss": 1.4443,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7468268049679269,
+      "grad_norm": 0.23165689408779144,
+      "learning_rate": 8e-05,
+      "loss": 1.4115,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7469632864746827,
+      "grad_norm": 0.2408268004655838,
+      "learning_rate": 8e-05,
+      "loss": 1.5446,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7470997679814385,
+      "grad_norm": 0.24573947489261627,
+      "learning_rate": 8e-05,
+      "loss": 1.3973,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7472362494881943,
+      "grad_norm": 0.22331851720809937,
+      "learning_rate": 8e-05,
+      "loss": 1.4241,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7473727309949502,
+      "grad_norm": 0.24434417486190796,
+      "learning_rate": 8e-05,
+      "loss": 1.529,
+      "step": 5476
+    },
+    {
+      "epoch": 0.747509212501706,
+      "grad_norm": 0.2336418330669403,
+      "learning_rate": 8e-05,
+      "loss": 1.5074,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7476456940084618,
+      "grad_norm": 0.2304091900587082,
+      "learning_rate": 8e-05,
+      "loss": 1.4056,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7477821755152176,
+      "grad_norm": 0.2245134860277176,
+      "learning_rate": 8e-05,
+      "loss": 1.3748,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7479186570219736,
+      "grad_norm": 0.23465248942375183,
+      "learning_rate": 8e-05,
+      "loss": 1.4656,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7480551385287294,
+      "grad_norm": 0.2328892946243286,
+      "learning_rate": 8e-05,
+      "loss": 1.441,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7481916200354852,
+      "grad_norm": 0.24208420515060425,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 5482
+    },
+    {
+      "epoch": 0.748328101542241,
+      "grad_norm": 0.2266128957271576,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7484645830489969,
+      "grad_norm": 0.2226395457983017,
+      "learning_rate": 8e-05,
+      "loss": 1.4232,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7486010645557527,
+      "grad_norm": 0.22303183376789093,
+      "learning_rate": 8e-05,
+      "loss": 1.4097,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7487375460625085,
+      "grad_norm": 0.24082913994789124,
+      "learning_rate": 8e-05,
+      "loss": 1.5058,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7488740275692644,
+      "grad_norm": 0.22559095919132233,
+      "learning_rate": 8e-05,
+      "loss": 1.4065,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7490105090760202,
+      "grad_norm": 0.22808629274368286,
+      "learning_rate": 8e-05,
+      "loss": 1.3575,
+      "step": 5488
+    },
+    {
+      "epoch": 0.749146990582776,
+      "grad_norm": 0.23182958364486694,
+      "learning_rate": 8e-05,
+      "loss": 1.4711,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7492834720895318,
+      "grad_norm": 0.22415268421173096,
+      "learning_rate": 8e-05,
+      "loss": 1.4239,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7494199535962877,
+      "grad_norm": 0.24841336905956268,
+      "learning_rate": 8e-05,
+      "loss": 1.4711,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7495564351030435,
+      "grad_norm": 0.23971742391586304,
+      "learning_rate": 8e-05,
+      "loss": 1.5148,
+      "step": 5492
+    },
+    {
+      "epoch": 0.7496929166097994,
+      "grad_norm": 0.2354128062725067,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7498293981165552,
+      "grad_norm": 0.2383018136024475,
+      "learning_rate": 8e-05,
+      "loss": 1.463,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7499658796233111,
+      "grad_norm": 0.2353178709745407,
+      "learning_rate": 8e-05,
+      "loss": 1.4126,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7501023611300669,
+      "grad_norm": 0.23781658709049225,
+      "learning_rate": 8e-05,
+      "loss": 1.478,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7502388426368227,
+      "grad_norm": 0.23770399391651154,
+      "learning_rate": 8e-05,
+      "loss": 1.4369,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7503753241435785,
+      "grad_norm": 0.23702123761177063,
+      "learning_rate": 8e-05,
+      "loss": 1.4261,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7505118056503344,
+      "grad_norm": 0.24078671634197235,
+      "learning_rate": 8e-05,
+      "loss": 1.4462,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7506482871570902,
+      "grad_norm": 0.2312447875738144,
+      "learning_rate": 8e-05,
+      "loss": 1.4066,
+      "step": 5500
+    },
+    {
+      "epoch": 0.750784768663846,
+      "grad_norm": 0.24043422937393188,
+      "learning_rate": 8e-05,
+      "loss": 1.462,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7509212501706019,
+      "grad_norm": 0.23676154017448425,
+      "learning_rate": 8e-05,
+      "loss": 1.5009,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7510577316773577,
+      "grad_norm": 0.23995369672775269,
+      "learning_rate": 8e-05,
+      "loss": 1.5265,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7511942131841135,
+      "grad_norm": 0.2344968020915985,
+      "learning_rate": 8e-05,
+      "loss": 1.4269,
+      "step": 5504
+    },
+    {
+      "epoch": 0.7513306946908694,
+      "grad_norm": 0.2372904270887375,
+      "learning_rate": 8e-05,
+      "loss": 1.3962,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7514671761976253,
+      "grad_norm": 0.22881458699703217,
+      "learning_rate": 8e-05,
+      "loss": 1.4453,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7516036577043811,
+      "grad_norm": 0.23556549847126007,
+      "learning_rate": 8e-05,
+      "loss": 1.4125,
+      "step": 5507
+    },
+    {
+      "epoch": 0.7517401392111369,
+      "grad_norm": 0.23557212948799133,
+      "learning_rate": 8e-05,
+      "loss": 1.3881,
+      "step": 5508
+    },
+    {
+      "epoch": 0.7518766207178927,
+      "grad_norm": 0.2407660335302353,
+      "learning_rate": 8e-05,
+      "loss": 1.4783,
+      "step": 5509
+    },
+    {
+      "epoch": 0.7520131022246486,
+      "grad_norm": 0.23004890978336334,
+      "learning_rate": 8e-05,
+      "loss": 1.3628,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7521495837314044,
+      "grad_norm": 0.23470093309879303,
+      "learning_rate": 8e-05,
+      "loss": 1.4402,
+      "step": 5511
+    },
+    {
+      "epoch": 0.7522860652381602,
+      "grad_norm": 0.23707321286201477,
+      "learning_rate": 8e-05,
+      "loss": 1.411,
+      "step": 5512
+    },
+    {
+      "epoch": 0.752422546744916,
+      "grad_norm": 0.2390184849500656,
+      "learning_rate": 8e-05,
+      "loss": 1.4362,
+      "step": 5513
+    },
+    {
+      "epoch": 0.7525590282516719,
+      "grad_norm": 0.22809483110904694,
+      "learning_rate": 8e-05,
+      "loss": 1.4173,
+      "step": 5514
+    },
+    {
+      "epoch": 0.7526955097584277,
+      "grad_norm": 0.23081955313682556,
+      "learning_rate": 8e-05,
+      "loss": 1.4913,
+      "step": 5515
+    },
+    {
+      "epoch": 0.7528319912651835,
+      "grad_norm": 0.24276885390281677,
+      "learning_rate": 8e-05,
+      "loss": 1.4564,
+      "step": 5516
+    },
+    {
+      "epoch": 0.7529684727719395,
+      "grad_norm": 0.23636262118816376,
+      "learning_rate": 8e-05,
+      "loss": 1.4392,
+      "step": 5517
+    },
+    {
+      "epoch": 0.7531049542786953,
+      "grad_norm": 0.232864648103714,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 5518
+    },
+    {
+      "epoch": 0.7532414357854511,
+      "grad_norm": 0.23204819858074188,
+      "learning_rate": 8e-05,
+      "loss": 1.4283,
+      "step": 5519
+    },
+    {
+      "epoch": 0.7533779172922069,
+      "grad_norm": 0.24035769701004028,
+      "learning_rate": 8e-05,
+      "loss": 1.4814,
+      "step": 5520
+    },
+    {
+      "epoch": 0.7535143987989628,
+      "grad_norm": 0.2342885434627533,
+      "learning_rate": 8e-05,
+      "loss": 1.3729,
+      "step": 5521
+    },
+    {
+      "epoch": 0.7536508803057186,
+      "grad_norm": 0.2284945398569107,
+      "learning_rate": 8e-05,
+      "loss": 1.436,
+      "step": 5522
+    },
+    {
+      "epoch": 0.7537873618124744,
+      "grad_norm": 0.23148933053016663,
+      "learning_rate": 8e-05,
+      "loss": 1.4784,
+      "step": 5523
+    },
+    {
+      "epoch": 0.7539238433192302,
+      "grad_norm": 0.23114363849163055,
+      "learning_rate": 8e-05,
+      "loss": 1.4384,
+      "step": 5524
+    },
+    {
+      "epoch": 0.7540603248259861,
+      "grad_norm": 0.2480083405971527,
+      "learning_rate": 8e-05,
+      "loss": 1.4388,
+      "step": 5525
+    },
+    {
+      "epoch": 0.7541968063327419,
+      "grad_norm": 0.22956795990467072,
+      "learning_rate": 8e-05,
+      "loss": 1.4084,
+      "step": 5526
+    },
+    {
+      "epoch": 0.7543332878394977,
+      "grad_norm": 0.2277536690235138,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 5527
+    },
+    {
+      "epoch": 0.7544697693462535,
+      "grad_norm": 0.2351599633693695,
+      "learning_rate": 8e-05,
+      "loss": 1.4234,
+      "step": 5528
+    },
+    {
+      "epoch": 0.7546062508530094,
+      "grad_norm": 0.24506905674934387,
+      "learning_rate": 8e-05,
+      "loss": 1.4337,
+      "step": 5529
+    },
+    {
+      "epoch": 0.7547427323597653,
+      "grad_norm": 0.23541918396949768,
+      "learning_rate": 8e-05,
+      "loss": 1.4375,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7548792138665211,
+      "grad_norm": 0.23244623839855194,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 5531
+    },
+    {
+      "epoch": 0.755015695373277,
+      "grad_norm": 0.23257069289684296,
+      "learning_rate": 8e-05,
+      "loss": 1.427,
+      "step": 5532
+    },
+    {
+      "epoch": 0.7551521768800328,
+      "grad_norm": 0.24327880144119263,
+      "learning_rate": 8e-05,
+      "loss": 1.4318,
+      "step": 5533
+    },
+    {
+      "epoch": 0.7552886583867886,
+      "grad_norm": 0.2355663925409317,
+      "learning_rate": 8e-05,
+      "loss": 1.4087,
+      "step": 5534
+    },
+    {
+      "epoch": 0.7554251398935444,
+      "grad_norm": 0.22469916939735413,
+      "learning_rate": 8e-05,
+      "loss": 1.4106,
+      "step": 5535
+    },
+    {
+      "epoch": 0.7555616214003003,
+      "grad_norm": 0.22179201245307922,
+      "learning_rate": 8e-05,
+      "loss": 1.5001,
+      "step": 5536
+    },
+    {
+      "epoch": 0.7556981029070561,
+      "grad_norm": 0.2329602688550949,
+      "learning_rate": 8e-05,
+      "loss": 1.4299,
+      "step": 5537
+    },
+    {
+      "epoch": 0.7558345844138119,
+      "grad_norm": 0.23746338486671448,
+      "learning_rate": 8e-05,
+      "loss": 1.4265,
+      "step": 5538
+    },
+    {
+      "epoch": 0.7559710659205677,
+      "grad_norm": 0.2400079220533371,
+      "learning_rate": 8e-05,
+      "loss": 1.4602,
+      "step": 5539
+    },
+    {
+      "epoch": 0.7561075474273236,
+      "grad_norm": 0.23185886442661285,
+      "learning_rate": 8e-05,
+      "loss": 1.4367,
+      "step": 5540
+    },
+    {
+      "epoch": 0.7562440289340794,
+      "grad_norm": 0.23354509472846985,
+      "learning_rate": 8e-05,
+      "loss": 1.4401,
+      "step": 5541
+    },
+    {
+      "epoch": 0.7563805104408353,
+      "grad_norm": 0.23393850028514862,
+      "learning_rate": 8e-05,
+      "loss": 1.4732,
+      "step": 5542
+    },
+    {
+      "epoch": 0.7565169919475911,
+      "grad_norm": 0.24462071061134338,
+      "learning_rate": 8e-05,
+      "loss": 1.5128,
+      "step": 5543
+    },
+    {
+      "epoch": 0.756653473454347,
+      "grad_norm": 0.23847441375255585,
+      "learning_rate": 8e-05,
+      "loss": 1.5163,
+      "step": 5544
+    },
+    {
+      "epoch": 0.7567899549611028,
+      "grad_norm": 0.23977838456630707,
+      "learning_rate": 8e-05,
+      "loss": 1.4759,
+      "step": 5545
+    },
+    {
+      "epoch": 0.7569264364678586,
+      "grad_norm": 0.22594425082206726,
+      "learning_rate": 8e-05,
+      "loss": 1.4486,
+      "step": 5546
+    },
+    {
+      "epoch": 0.7570629179746144,
+      "grad_norm": 0.22592028975486755,
+      "learning_rate": 8e-05,
+      "loss": 1.4175,
+      "step": 5547
+    },
+    {
+      "epoch": 0.7571993994813703,
+      "grad_norm": 0.23018395900726318,
+      "learning_rate": 8e-05,
+      "loss": 1.5055,
+      "step": 5548
+    },
+    {
+      "epoch": 0.7573358809881261,
+      "grad_norm": 0.23036062717437744,
+      "learning_rate": 8e-05,
+      "loss": 1.3831,
+      "step": 5549
+    },
+    {
+      "epoch": 0.7574723624948819,
+      "grad_norm": 0.2336152344942093,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7576088440016377,
+      "grad_norm": 0.24702188372612,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 5551
+    },
+    {
+      "epoch": 0.7577453255083936,
+      "grad_norm": 0.23044301569461823,
+      "learning_rate": 8e-05,
+      "loss": 1.4932,
+      "step": 5552
+    },
+    {
+      "epoch": 0.7578818070151494,
+      "grad_norm": 0.23151962459087372,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 5553
+    },
+    {
+      "epoch": 0.7580182885219053,
+      "grad_norm": 0.2313581109046936,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 5554
+    },
+    {
+      "epoch": 0.7581547700286612,
+      "grad_norm": 0.2368733286857605,
+      "learning_rate": 8e-05,
+      "loss": 1.466,
+      "step": 5555
+    },
+    {
+      "epoch": 0.758291251535417,
+      "grad_norm": 0.23817236721515656,
+      "learning_rate": 8e-05,
+      "loss": 1.4798,
+      "step": 5556
+    },
+    {
+      "epoch": 0.7584277330421728,
+      "grad_norm": 0.22685439884662628,
+      "learning_rate": 8e-05,
+      "loss": 1.4202,
+      "step": 5557
+    },
+    {
+      "epoch": 0.7585642145489286,
+      "grad_norm": 0.2357075810432434,
+      "learning_rate": 8e-05,
+      "loss": 1.4679,
+      "step": 5558
+    },
+    {
+      "epoch": 0.7587006960556845,
+      "grad_norm": 0.22319969534873962,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 5559
+    },
+    {
+      "epoch": 0.7588371775624403,
+      "grad_norm": 0.23965150117874146,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 5560
+    },
+    {
+      "epoch": 0.7589736590691961,
+      "grad_norm": 0.2285008579492569,
+      "learning_rate": 8e-05,
+      "loss": 1.4787,
+      "step": 5561
+    },
+    {
+      "epoch": 0.7591101405759519,
+      "grad_norm": 0.26351869106292725,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 5562
+    },
+    {
+      "epoch": 0.7592466220827078,
+      "grad_norm": 0.23487111926078796,
+      "learning_rate": 8e-05,
+      "loss": 1.4344,
+      "step": 5563
+    },
+    {
+      "epoch": 0.7593831035894636,
+      "grad_norm": 0.2296253740787506,
+      "learning_rate": 8e-05,
+      "loss": 1.3947,
+      "step": 5564
+    },
+    {
+      "epoch": 0.7595195850962194,
+      "grad_norm": 0.234577476978302,
+      "learning_rate": 8e-05,
+      "loss": 1.4392,
+      "step": 5565
+    },
+    {
+      "epoch": 0.7596560666029752,
+      "grad_norm": 0.23296037316322327,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 5566
+    },
+    {
+      "epoch": 0.7597925481097312,
+      "grad_norm": 0.23680196702480316,
+      "learning_rate": 8e-05,
+      "loss": 1.4446,
+      "step": 5567
+    },
+    {
+      "epoch": 0.759929029616487,
+      "grad_norm": 0.22917480766773224,
+      "learning_rate": 8e-05,
+      "loss": 1.4617,
+      "step": 5568
+    },
+    {
+      "epoch": 0.7600655111232428,
+      "grad_norm": 0.2350766658782959,
+      "learning_rate": 8e-05,
+      "loss": 1.445,
+      "step": 5569
+    },
+    {
+      "epoch": 0.7602019926299987,
+      "grad_norm": 0.2342926561832428,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 5570
+    },
+    {
+      "epoch": 0.7603384741367545,
+      "grad_norm": 0.2328982949256897,
+      "learning_rate": 8e-05,
+      "loss": 1.4398,
+      "step": 5571
+    },
+    {
+      "epoch": 0.7604749556435103,
+      "grad_norm": 0.24676910042762756,
+      "learning_rate": 8e-05,
+      "loss": 1.4355,
+      "step": 5572
+    },
+    {
+      "epoch": 0.7606114371502661,
+      "grad_norm": 0.24163806438446045,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 5573
+    },
+    {
+      "epoch": 0.760747918657022,
+      "grad_norm": 0.23059327900409698,
+      "learning_rate": 8e-05,
+      "loss": 1.3655,
+      "step": 5574
+    },
+    {
+      "epoch": 0.7608844001637778,
+      "grad_norm": 0.2266145795583725,
+      "learning_rate": 8e-05,
+      "loss": 1.3812,
+      "step": 5575
+    },
+    {
+      "epoch": 0.7610208816705336,
+      "grad_norm": 0.22920896112918854,
+      "learning_rate": 8e-05,
+      "loss": 1.4125,
+      "step": 5576
+    },
+    {
+      "epoch": 0.7611573631772894,
+      "grad_norm": 0.23637108504772186,
+      "learning_rate": 8e-05,
+      "loss": 1.4725,
+      "step": 5577
+    },
+    {
+      "epoch": 0.7612938446840453,
+      "grad_norm": 0.22649773955345154,
+      "learning_rate": 8e-05,
+      "loss": 1.4226,
+      "step": 5578
+    },
+    {
+      "epoch": 0.7614303261908012,
+      "grad_norm": 0.23809078335762024,
+      "learning_rate": 8e-05,
+      "loss": 1.4471,
+      "step": 5579
+    },
+    {
+      "epoch": 0.761566807697557,
+      "grad_norm": 0.23114313185214996,
+      "learning_rate": 8e-05,
+      "loss": 1.4383,
+      "step": 5580
+    },
+    {
+      "epoch": 0.7617032892043129,
+      "grad_norm": 0.23516236245632172,
+      "learning_rate": 8e-05,
+      "loss": 1.4248,
+      "step": 5581
+    },
+    {
+      "epoch": 0.7618397707110687,
+      "grad_norm": 0.24090434610843658,
+      "learning_rate": 8e-05,
+      "loss": 1.4772,
+      "step": 5582
+    },
+    {
+      "epoch": 0.7619762522178245,
+      "grad_norm": 0.24062126874923706,
+      "learning_rate": 8e-05,
+      "loss": 1.4741,
+      "step": 5583
+    },
+    {
+      "epoch": 0.7621127337245803,
+      "grad_norm": 0.2393125742673874,
+      "learning_rate": 8e-05,
+      "loss": 1.4725,
+      "step": 5584
+    },
+    {
+      "epoch": 0.7622492152313362,
+      "grad_norm": 0.24033060669898987,
+      "learning_rate": 8e-05,
+      "loss": 1.4362,
+      "step": 5585
+    },
+    {
+      "epoch": 0.762385696738092,
+      "grad_norm": 0.2344217449426651,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 5586
+    },
+    {
+      "epoch": 0.7625221782448478,
+      "grad_norm": 0.22848385572433472,
+      "learning_rate": 8e-05,
+      "loss": 1.4073,
+      "step": 5587
+    },
+    {
+      "epoch": 0.7626586597516036,
+      "grad_norm": 0.22943542897701263,
+      "learning_rate": 8e-05,
+      "loss": 1.4558,
+      "step": 5588
+    },
+    {
+      "epoch": 0.7627951412583595,
+      "grad_norm": 0.24466146528720856,
+      "learning_rate": 8e-05,
+      "loss": 1.473,
+      "step": 5589
+    },
+    {
+      "epoch": 0.7629316227651153,
+      "grad_norm": 0.23269988596439362,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 5590
+    },
+    {
+      "epoch": 0.7630681042718711,
+      "grad_norm": 0.24303074181079865,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 5591
+    },
+    {
+      "epoch": 0.763204585778627,
+      "grad_norm": 0.23157356679439545,
+      "learning_rate": 8e-05,
+      "loss": 1.3407,
+      "step": 5592
+    },
+    {
+      "epoch": 0.7633410672853829,
+      "grad_norm": 0.24543976783752441,
+      "learning_rate": 8e-05,
+      "loss": 1.4786,
+      "step": 5593
+    },
+    {
+      "epoch": 0.7634775487921387,
+      "grad_norm": 0.23173443973064423,
+      "learning_rate": 8e-05,
+      "loss": 1.4225,
+      "step": 5594
+    },
+    {
+      "epoch": 0.7636140302988945,
+      "grad_norm": 0.23874112963676453,
+      "learning_rate": 8e-05,
+      "loss": 1.3724,
+      "step": 5595
+    },
+    {
+      "epoch": 0.7637505118056503,
+      "grad_norm": 0.23418624699115753,
+      "learning_rate": 8e-05,
+      "loss": 1.3807,
+      "step": 5596
+    },
+    {
+      "epoch": 0.7638869933124062,
+      "grad_norm": 0.23073336482048035,
+      "learning_rate": 8e-05,
+      "loss": 1.4673,
+      "step": 5597
+    },
+    {
+      "epoch": 0.764023474819162,
+      "grad_norm": 0.23574082553386688,
+      "learning_rate": 8e-05,
+      "loss": 1.4335,
+      "step": 5598
+    },
+    {
+      "epoch": 0.7641599563259178,
+      "grad_norm": 0.2334708720445633,
+      "learning_rate": 8e-05,
+      "loss": 1.4122,
+      "step": 5599
+    },
+    {
+      "epoch": 0.7642964378326736,
+      "grad_norm": 0.2397107183933258,
+      "learning_rate": 8e-05,
+      "loss": 1.4233,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7644329193394295,
+      "grad_norm": 0.24064216017723083,
+      "learning_rate": 8e-05,
+      "loss": 1.4608,
+      "step": 5601
+    },
+    {
+      "epoch": 0.7645694008461853,
+      "grad_norm": 0.24223579466342926,
+      "learning_rate": 8e-05,
+      "loss": 1.5542,
+      "step": 5602
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 0.23448175191879272,
+      "learning_rate": 8e-05,
+      "loss": 1.4491,
+      "step": 5603
+    },
+    {
+      "epoch": 0.7648423638596971,
+      "grad_norm": 0.2318779081106186,
+      "learning_rate": 8e-05,
+      "loss": 1.4085,
+      "step": 5604
+    },
+    {
+      "epoch": 0.7649788453664529,
+      "grad_norm": 0.25499752163887024,
+      "learning_rate": 8e-05,
+      "loss": 1.4255,
+      "step": 5605
+    },
+    {
+      "epoch": 0.7651153268732087,
+      "grad_norm": 0.2482333928346634,
+      "learning_rate": 8e-05,
+      "loss": 1.449,
+      "step": 5606
+    },
+    {
+      "epoch": 0.7652518083799645,
+      "grad_norm": 0.23240843415260315,
+      "learning_rate": 8e-05,
+      "loss": 1.4588,
+      "step": 5607
+    },
+    {
+      "epoch": 0.7653882898867204,
+      "grad_norm": 0.23425838351249695,
+      "learning_rate": 8e-05,
+      "loss": 1.4097,
+      "step": 5608
+    },
+    {
+      "epoch": 0.7655247713934762,
+      "grad_norm": 0.22358208894729614,
+      "learning_rate": 8e-05,
+      "loss": 1.4147,
+      "step": 5609
+    },
+    {
+      "epoch": 0.765661252900232,
+      "grad_norm": 0.2454998642206192,
+      "learning_rate": 8e-05,
+      "loss": 1.5202,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7657977344069878,
+      "grad_norm": 0.240730419754982,
+      "learning_rate": 8e-05,
+      "loss": 1.4494,
+      "step": 5611
+    },
+    {
+      "epoch": 0.7659342159137437,
+      "grad_norm": 0.2335236817598343,
+      "learning_rate": 8e-05,
+      "loss": 1.5027,
+      "step": 5612
+    },
+    {
+      "epoch": 0.7660706974204995,
+      "grad_norm": 0.23862071335315704,
+      "learning_rate": 8e-05,
+      "loss": 1.5036,
+      "step": 5613
+    },
+    {
+      "epoch": 0.7662071789272553,
+      "grad_norm": 0.24564708769321442,
+      "learning_rate": 8e-05,
+      "loss": 1.4168,
+      "step": 5614
+    },
+    {
+      "epoch": 0.7663436604340111,
+      "grad_norm": 0.23729662597179413,
+      "learning_rate": 8e-05,
+      "loss": 1.4141,
+      "step": 5615
+    },
+    {
+      "epoch": 0.7664801419407671,
+      "grad_norm": 0.24129271507263184,
+      "learning_rate": 8e-05,
+      "loss": 1.3664,
+      "step": 5616
+    },
+    {
+      "epoch": 0.7666166234475229,
+      "grad_norm": 0.23258242011070251,
+      "learning_rate": 8e-05,
+      "loss": 1.4111,
+      "step": 5617
+    },
+    {
+      "epoch": 0.7667531049542787,
+      "grad_norm": 0.23705439269542694,
+      "learning_rate": 8e-05,
+      "loss": 1.4453,
+      "step": 5618
+    },
+    {
+      "epoch": 0.7668895864610346,
+      "grad_norm": 0.2408711463212967,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 5619
+    },
+    {
+      "epoch": 0.7670260679677904,
+      "grad_norm": 0.2217932790517807,
+      "learning_rate": 8e-05,
+      "loss": 1.3924,
+      "step": 5620
+    },
+    {
+      "epoch": 0.7671625494745462,
+      "grad_norm": 0.23366065323352814,
+      "learning_rate": 8e-05,
+      "loss": 1.4182,
+      "step": 5621
+    },
+    {
+      "epoch": 0.767299030981302,
+      "grad_norm": 0.24863748252391815,
+      "learning_rate": 8e-05,
+      "loss": 1.4476,
+      "step": 5622
+    },
+    {
+      "epoch": 0.7674355124880579,
+      "grad_norm": 0.23968449234962463,
+      "learning_rate": 8e-05,
+      "loss": 1.4032,
+      "step": 5623
+    },
+    {
+      "epoch": 0.7675719939948137,
+      "grad_norm": 0.23493170738220215,
+      "learning_rate": 8e-05,
+      "loss": 1.4457,
+      "step": 5624
+    },
+    {
+      "epoch": 0.7677084755015695,
+      "grad_norm": 0.23007884621620178,
+      "learning_rate": 8e-05,
+      "loss": 1.5188,
+      "step": 5625
+    },
+    {
+      "epoch": 0.7678449570083253,
+      "grad_norm": 0.23192553222179413,
+      "learning_rate": 8e-05,
+      "loss": 1.4293,
+      "step": 5626
+    },
+    {
+      "epoch": 0.7679814385150812,
+      "grad_norm": 0.2348790019750595,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 5627
+    },
+    {
+      "epoch": 0.768117920021837,
+      "grad_norm": 0.2361660897731781,
+      "learning_rate": 8e-05,
+      "loss": 1.4908,
+      "step": 5628
+    },
+    {
+      "epoch": 0.7682544015285929,
+      "grad_norm": 0.2330189347267151,
+      "learning_rate": 8e-05,
+      "loss": 1.4654,
+      "step": 5629
+    },
+    {
+      "epoch": 0.7683908830353487,
+      "grad_norm": 0.23572468757629395,
+      "learning_rate": 8e-05,
+      "loss": 1.4528,
+      "step": 5630
+    },
+    {
+      "epoch": 0.7685273645421046,
+      "grad_norm": 0.238274484872818,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 5631
+    },
+    {
+      "epoch": 0.7686638460488604,
+      "grad_norm": 0.23053410649299622,
+      "learning_rate": 8e-05,
+      "loss": 1.413,
+      "step": 5632
+    },
+    {
+      "epoch": 0.7688003275556162,
+      "grad_norm": 0.23708906769752502,
+      "learning_rate": 8e-05,
+      "loss": 1.4098,
+      "step": 5633
+    },
+    {
+      "epoch": 0.768936809062372,
+      "grad_norm": 0.2415619194507599,
+      "learning_rate": 8e-05,
+      "loss": 1.364,
+      "step": 5634
+    },
+    {
+      "epoch": 0.7690732905691279,
+      "grad_norm": 0.24315476417541504,
+      "learning_rate": 8e-05,
+      "loss": 1.4573,
+      "step": 5635
+    },
+    {
+      "epoch": 0.7692097720758837,
+      "grad_norm": 0.22907719016075134,
+      "learning_rate": 8e-05,
+      "loss": 1.4112,
+      "step": 5636
+    },
+    {
+      "epoch": 0.7693462535826395,
+      "grad_norm": 0.2439599186182022,
+      "learning_rate": 8e-05,
+      "loss": 1.4219,
+      "step": 5637
+    },
+    {
+      "epoch": 0.7694827350893954,
+      "grad_norm": 0.23360145092010498,
+      "learning_rate": 8e-05,
+      "loss": 1.4618,
+      "step": 5638
+    },
+    {
+      "epoch": 0.7696192165961512,
+      "grad_norm": 0.2358836531639099,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 5639
+    },
+    {
+      "epoch": 0.769755698102907,
+      "grad_norm": 0.22917979955673218,
+      "learning_rate": 8e-05,
+      "loss": 1.5067,
+      "step": 5640
+    },
+    {
+      "epoch": 0.7698921796096629,
+      "grad_norm": 0.22038787603378296,
+      "learning_rate": 8e-05,
+      "loss": 1.4085,
+      "step": 5641
+    },
+    {
+      "epoch": 0.7700286611164188,
+      "grad_norm": 0.23641133308410645,
+      "learning_rate": 8e-05,
+      "loss": 1.4813,
+      "step": 5642
+    },
+    {
+      "epoch": 0.7701651426231746,
+      "grad_norm": 0.2468692809343338,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 5643
+    },
+    {
+      "epoch": 0.7703016241299304,
+      "grad_norm": 0.23734407126903534,
+      "learning_rate": 8e-05,
+      "loss": 1.4122,
+      "step": 5644
+    },
+    {
+      "epoch": 0.7704381056366862,
+      "grad_norm": 0.24895983934402466,
+      "learning_rate": 8e-05,
+      "loss": 1.4075,
+      "step": 5645
+    },
+    {
+      "epoch": 0.7705745871434421,
+      "grad_norm": 0.23009754717350006,
+      "learning_rate": 8e-05,
+      "loss": 1.3741,
+      "step": 5646
+    },
+    {
+      "epoch": 0.7707110686501979,
+      "grad_norm": 0.23996686935424805,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 5647
+    },
+    {
+      "epoch": 0.7708475501569537,
+      "grad_norm": 0.23327364027500153,
+      "learning_rate": 8e-05,
+      "loss": 1.4329,
+      "step": 5648
+    },
+    {
+      "epoch": 0.7709840316637095,
+      "grad_norm": 0.23154279589653015,
+      "learning_rate": 8e-05,
+      "loss": 1.4216,
+      "step": 5649
+    },
+    {
+      "epoch": 0.7711205131704654,
+      "grad_norm": 0.2311466485261917,
+      "learning_rate": 8e-05,
+      "loss": 1.4872,
+      "step": 5650
+    },
+    {
+      "epoch": 0.7712569946772212,
+      "grad_norm": 0.2321779429912567,
+      "learning_rate": 8e-05,
+      "loss": 1.439,
+      "step": 5651
+    },
+    {
+      "epoch": 0.771393476183977,
+      "grad_norm": 0.23593227565288544,
+      "learning_rate": 8e-05,
+      "loss": 1.4213,
+      "step": 5652
+    },
+    {
+      "epoch": 0.771529957690733,
+      "grad_norm": 0.23477531969547272,
+      "learning_rate": 8e-05,
+      "loss": 1.4634,
+      "step": 5653
+    },
+    {
+      "epoch": 0.7716664391974888,
+      "grad_norm": 0.2579036355018616,
+      "learning_rate": 8e-05,
+      "loss": 1.4445,
+      "step": 5654
+    },
+    {
+      "epoch": 0.7718029207042446,
+      "grad_norm": 0.23362836241722107,
+      "learning_rate": 8e-05,
+      "loss": 1.418,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7719394022110004,
+      "grad_norm": 0.24268262088298798,
+      "learning_rate": 8e-05,
+      "loss": 1.4537,
+      "step": 5656
+    },
+    {
+      "epoch": 0.7720758837177563,
+      "grad_norm": 0.2433621734380722,
+      "learning_rate": 8e-05,
+      "loss": 1.4536,
+      "step": 5657
+    },
+    {
+      "epoch": 0.7722123652245121,
+      "grad_norm": 0.22754450142383575,
+      "learning_rate": 8e-05,
+      "loss": 1.4487,
+      "step": 5658
+    },
+    {
+      "epoch": 0.7723488467312679,
+      "grad_norm": 0.24308529496192932,
+      "learning_rate": 8e-05,
+      "loss": 1.4534,
+      "step": 5659
+    },
+    {
+      "epoch": 0.7724853282380237,
+      "grad_norm": 0.2411012500524521,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 5660
+    },
+    {
+      "epoch": 0.7726218097447796,
+      "grad_norm": 0.24843473732471466,
+      "learning_rate": 8e-05,
+      "loss": 1.4702,
+      "step": 5661
+    },
+    {
+      "epoch": 0.7727582912515354,
+      "grad_norm": 0.24544332921504974,
+      "learning_rate": 8e-05,
+      "loss": 1.4372,
+      "step": 5662
+    },
+    {
+      "epoch": 0.7728947727582912,
+      "grad_norm": 0.22988121211528778,
+      "learning_rate": 8e-05,
+      "loss": 1.421,
+      "step": 5663
+    },
+    {
+      "epoch": 0.773031254265047,
+      "grad_norm": 0.23389500379562378,
+      "learning_rate": 8e-05,
+      "loss": 1.3946,
+      "step": 5664
+    },
+    {
+      "epoch": 0.7731677357718029,
+      "grad_norm": 0.24111759662628174,
+      "learning_rate": 8e-05,
+      "loss": 1.4396,
+      "step": 5665
+    },
+    {
+      "epoch": 0.7733042172785588,
+      "grad_norm": 0.2369510531425476,
+      "learning_rate": 8e-05,
+      "loss": 1.4239,
+      "step": 5666
+    },
+    {
+      "epoch": 0.7734406987853146,
+      "grad_norm": 0.237503319978714,
+      "learning_rate": 8e-05,
+      "loss": 1.436,
+      "step": 5667
+    },
+    {
+      "epoch": 0.7735771802920705,
+      "grad_norm": 0.24653634428977966,
+      "learning_rate": 8e-05,
+      "loss": 1.4855,
+      "step": 5668
+    },
+    {
+      "epoch": 0.7737136617988263,
+      "grad_norm": 0.23189572989940643,
+      "learning_rate": 8e-05,
+      "loss": 1.4411,
+      "step": 5669
+    },
+    {
+      "epoch": 0.7738501433055821,
+      "grad_norm": 0.22882738709449768,
+      "learning_rate": 8e-05,
+      "loss": 1.4562,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7739866248123379,
+      "grad_norm": 0.23600970208644867,
+      "learning_rate": 8e-05,
+      "loss": 1.4833,
+      "step": 5671
+    },
+    {
+      "epoch": 0.7741231063190938,
+      "grad_norm": 0.2194918394088745,
+      "learning_rate": 8e-05,
+      "loss": 1.3665,
+      "step": 5672
+    },
+    {
+      "epoch": 0.7742595878258496,
+      "grad_norm": 0.2363158017396927,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 5673
+    },
+    {
+      "epoch": 0.7743960693326054,
+      "grad_norm": 0.2377772182226181,
+      "learning_rate": 8e-05,
+      "loss": 1.4212,
+      "step": 5674
+    },
+    {
+      "epoch": 0.7745325508393612,
+      "grad_norm": 0.24195361137390137,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 5675
+    },
+    {
+      "epoch": 0.7746690323461171,
+      "grad_norm": 0.23967291414737701,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 5676
+    },
+    {
+      "epoch": 0.7748055138528729,
+      "grad_norm": 0.24107171595096588,
+      "learning_rate": 8e-05,
+      "loss": 1.4645,
+      "step": 5677
+    },
+    {
+      "epoch": 0.7749419953596288,
+      "grad_norm": 0.23616956174373627,
+      "learning_rate": 8e-05,
+      "loss": 1.4338,
+      "step": 5678
+    },
+    {
+      "epoch": 0.7750784768663846,
+      "grad_norm": 0.24143461883068085,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 5679
+    },
+    {
+      "epoch": 0.7752149583731405,
+      "grad_norm": 0.22714421153068542,
+      "learning_rate": 8e-05,
+      "loss": 1.4414,
+      "step": 5680
+    },
+    {
+      "epoch": 0.7753514398798963,
+      "grad_norm": 0.232874795794487,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 5681
+    },
+    {
+      "epoch": 0.7754879213866521,
+      "grad_norm": 0.24673987925052643,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 5682
+    },
+    {
+      "epoch": 0.775624402893408,
+      "grad_norm": 0.23305775225162506,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 5683
+    },
+    {
+      "epoch": 0.7757608844001638,
+      "grad_norm": 0.22705726325511932,
+      "learning_rate": 8e-05,
+      "loss": 1.4363,
+      "step": 5684
+    },
+    {
+      "epoch": 0.7758973659069196,
+      "grad_norm": 0.23369674384593964,
+      "learning_rate": 8e-05,
+      "loss": 1.398,
+      "step": 5685
+    },
+    {
+      "epoch": 0.7760338474136754,
+      "grad_norm": 0.23789799213409424,
+      "learning_rate": 8e-05,
+      "loss": 1.4561,
+      "step": 5686
+    },
+    {
+      "epoch": 0.7761703289204313,
+      "grad_norm": 0.2333407700061798,
+      "learning_rate": 8e-05,
+      "loss": 1.393,
+      "step": 5687
+    },
+    {
+      "epoch": 0.7763068104271871,
+      "grad_norm": 0.2326601892709732,
+      "learning_rate": 8e-05,
+      "loss": 1.3852,
+      "step": 5688
+    },
+    {
+      "epoch": 0.7764432919339429,
+      "grad_norm": 0.23092429339885712,
+      "learning_rate": 8e-05,
+      "loss": 1.4325,
+      "step": 5689
+    },
+    {
+      "epoch": 0.7765797734406988,
+      "grad_norm": 0.24055393040180206,
+      "learning_rate": 8e-05,
+      "loss": 1.4251,
+      "step": 5690
+    },
+    {
+      "epoch": 0.7767162549474547,
+      "grad_norm": 0.23387575149536133,
+      "learning_rate": 8e-05,
+      "loss": 1.4462,
+      "step": 5691
+    },
+    {
+      "epoch": 0.7768527364542105,
+      "grad_norm": 0.2318478226661682,
+      "learning_rate": 8e-05,
+      "loss": 1.5058,
+      "step": 5692
+    },
+    {
+      "epoch": 0.7769892179609663,
+      "grad_norm": 0.23904550075531006,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 5693
+    },
+    {
+      "epoch": 0.7771256994677221,
+      "grad_norm": 0.23743632435798645,
+      "learning_rate": 8e-05,
+      "loss": 1.3661,
+      "step": 5694
+    },
+    {
+      "epoch": 0.777262180974478,
+      "grad_norm": 0.23476843535900116,
+      "learning_rate": 8e-05,
+      "loss": 1.4702,
+      "step": 5695
+    },
+    {
+      "epoch": 0.7773986624812338,
+      "grad_norm": 0.2407900094985962,
+      "learning_rate": 8e-05,
+      "loss": 1.4776,
+      "step": 5696
+    },
+    {
+      "epoch": 0.7775351439879896,
+      "grad_norm": 0.2450813502073288,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 5697
+    },
+    {
+      "epoch": 0.7776716254947454,
+      "grad_norm": 0.24269409477710724,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 5698
+    },
+    {
+      "epoch": 0.7778081070015013,
+      "grad_norm": 0.23397736251354218,
+      "learning_rate": 8e-05,
+      "loss": 1.4238,
+      "step": 5699
+    },
+    {
+      "epoch": 0.7779445885082571,
+      "grad_norm": 0.24252445995807648,
+      "learning_rate": 8e-05,
+      "loss": 1.4749,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7780810700150129,
+      "grad_norm": 0.23396548628807068,
+      "learning_rate": 8e-05,
+      "loss": 1.4336,
+      "step": 5701
+    },
+    {
+      "epoch": 0.7782175515217687,
+      "grad_norm": 0.23370850086212158,
+      "learning_rate": 8e-05,
+      "loss": 1.4516,
+      "step": 5702
+    },
+    {
+      "epoch": 0.7783540330285247,
+      "grad_norm": 0.23984333872795105,
+      "learning_rate": 8e-05,
+      "loss": 1.4726,
+      "step": 5703
+    },
+    {
+      "epoch": 0.7784905145352805,
+      "grad_norm": 0.24317516386508942,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 5704
+    },
+    {
+      "epoch": 0.7786269960420363,
+      "grad_norm": 0.2452613115310669,
+      "learning_rate": 8e-05,
+      "loss": 1.4498,
+      "step": 5705
+    },
+    {
+      "epoch": 0.7787634775487922,
+      "grad_norm": 0.23710697889328003,
+      "learning_rate": 8e-05,
+      "loss": 1.3645,
+      "step": 5706
+    },
+    {
+      "epoch": 0.778899959055548,
+      "grad_norm": 0.22427794337272644,
+      "learning_rate": 8e-05,
+      "loss": 1.4195,
+      "step": 5707
+    },
+    {
+      "epoch": 0.7790364405623038,
+      "grad_norm": 0.23783913254737854,
+      "learning_rate": 8e-05,
+      "loss": 1.3569,
+      "step": 5708
+    },
+    {
+      "epoch": 0.7791729220690596,
+      "grad_norm": 0.25410500168800354,
+      "learning_rate": 8e-05,
+      "loss": 1.3909,
+      "step": 5709
+    },
+    {
+      "epoch": 0.7793094035758155,
+      "grad_norm": 0.24704043567180634,
+      "learning_rate": 8e-05,
+      "loss": 1.4314,
+      "step": 5710
+    },
+    {
+      "epoch": 0.7794458850825713,
+      "grad_norm": 0.24660030007362366,
+      "learning_rate": 8e-05,
+      "loss": 1.4998,
+      "step": 5711
+    },
+    {
+      "epoch": 0.7795823665893271,
+      "grad_norm": 0.23867273330688477,
+      "learning_rate": 8e-05,
+      "loss": 1.4869,
+      "step": 5712
+    },
+    {
+      "epoch": 0.7797188480960829,
+      "grad_norm": 0.23602385818958282,
+      "learning_rate": 8e-05,
+      "loss": 1.4554,
+      "step": 5713
+    },
+    {
+      "epoch": 0.7798553296028388,
+      "grad_norm": 0.23208419978618622,
+      "learning_rate": 8e-05,
+      "loss": 1.4558,
+      "step": 5714
+    },
+    {
+      "epoch": 0.7799918111095947,
+      "grad_norm": 0.23959524929523468,
+      "learning_rate": 8e-05,
+      "loss": 1.4352,
+      "step": 5715
+    },
+    {
+      "epoch": 0.7801282926163505,
+      "grad_norm": 0.24005185067653656,
+      "learning_rate": 8e-05,
+      "loss": 1.3903,
+      "step": 5716
+    },
+    {
+      "epoch": 0.7802647741231064,
+      "grad_norm": 0.23749057948589325,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 5717
+    },
+    {
+      "epoch": 0.7804012556298622,
+      "grad_norm": 0.23761336505413055,
+      "learning_rate": 8e-05,
+      "loss": 1.3876,
+      "step": 5718
+    },
+    {
+      "epoch": 0.780537737136618,
+      "grad_norm": 0.2384224534034729,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 5719
+    },
+    {
+      "epoch": 0.7806742186433738,
+      "grad_norm": 0.23048044741153717,
+      "learning_rate": 8e-05,
+      "loss": 1.3681,
+      "step": 5720
+    },
+    {
+      "epoch": 0.7808107001501297,
+      "grad_norm": 0.24642930924892426,
+      "learning_rate": 8e-05,
+      "loss": 1.4614,
+      "step": 5721
+    },
+    {
+      "epoch": 0.7809471816568855,
+      "grad_norm": 0.23711124062538147,
+      "learning_rate": 8e-05,
+      "loss": 1.4458,
+      "step": 5722
+    },
+    {
+      "epoch": 0.7810836631636413,
+      "grad_norm": 0.2385333627462387,
+      "learning_rate": 8e-05,
+      "loss": 1.4467,
+      "step": 5723
+    },
+    {
+      "epoch": 0.7812201446703971,
+      "grad_norm": 0.2434920370578766,
+      "learning_rate": 8e-05,
+      "loss": 1.4035,
+      "step": 5724
+    },
+    {
+      "epoch": 0.781356626177153,
+      "grad_norm": 0.22612306475639343,
+      "learning_rate": 8e-05,
+      "loss": 1.4451,
+      "step": 5725
+    },
+    {
+      "epoch": 0.7814931076839088,
+      "grad_norm": 0.23159319162368774,
+      "learning_rate": 8e-05,
+      "loss": 1.3625,
+      "step": 5726
+    },
+    {
+      "epoch": 0.7816295891906647,
+      "grad_norm": 0.25096338987350464,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 5727
+    },
+    {
+      "epoch": 0.7817660706974205,
+      "grad_norm": 0.245771586894989,
+      "learning_rate": 8e-05,
+      "loss": 1.4348,
+      "step": 5728
+    },
+    {
+      "epoch": 0.7819025522041764,
+      "grad_norm": 0.25852423906326294,
+      "learning_rate": 8e-05,
+      "loss": 1.3989,
+      "step": 5729
+    },
+    {
+      "epoch": 0.7820390337109322,
+      "grad_norm": 0.23343418538570404,
+      "learning_rate": 8e-05,
+      "loss": 1.4398,
+      "step": 5730
+    },
+    {
+      "epoch": 0.782175515217688,
+      "grad_norm": 0.23580682277679443,
+      "learning_rate": 8e-05,
+      "loss": 1.4248,
+      "step": 5731
+    },
+    {
+      "epoch": 0.7823119967244438,
+      "grad_norm": 0.24060769379138947,
+      "learning_rate": 8e-05,
+      "loss": 1.357,
+      "step": 5732
+    },
+    {
+      "epoch": 0.7824484782311997,
+      "grad_norm": 0.2467980533838272,
+      "learning_rate": 8e-05,
+      "loss": 1.4721,
+      "step": 5733
+    },
+    {
+      "epoch": 0.7825849597379555,
+      "grad_norm": 0.23783157765865326,
+      "learning_rate": 8e-05,
+      "loss": 1.4425,
+      "step": 5734
+    },
+    {
+      "epoch": 0.7827214412447113,
+      "grad_norm": 0.2356017678976059,
+      "learning_rate": 8e-05,
+      "loss": 1.3779,
+      "step": 5735
+    },
+    {
+      "epoch": 0.7828579227514672,
+      "grad_norm": 0.23383989930152893,
+      "learning_rate": 8e-05,
+      "loss": 1.4456,
+      "step": 5736
+    },
+    {
+      "epoch": 0.782994404258223,
+      "grad_norm": 0.2379591017961502,
+      "learning_rate": 8e-05,
+      "loss": 1.4702,
+      "step": 5737
+    },
+    {
+      "epoch": 0.7831308857649788,
+      "grad_norm": 0.23061588406562805,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 5738
+    },
+    {
+      "epoch": 0.7832673672717346,
+      "grad_norm": 0.2344280630350113,
+      "learning_rate": 8e-05,
+      "loss": 1.4278,
+      "step": 5739
+    },
+    {
+      "epoch": 0.7834038487784906,
+      "grad_norm": 0.2349577397108078,
+      "learning_rate": 8e-05,
+      "loss": 1.4269,
+      "step": 5740
+    },
+    {
+      "epoch": 0.7835403302852464,
+      "grad_norm": 0.23601274192333221,
+      "learning_rate": 8e-05,
+      "loss": 1.4177,
+      "step": 5741
+    },
+    {
+      "epoch": 0.7836768117920022,
+      "grad_norm": 0.2434464991092682,
+      "learning_rate": 8e-05,
+      "loss": 1.4496,
+      "step": 5742
+    },
+    {
+      "epoch": 0.783813293298758,
+      "grad_norm": 0.2316821664571762,
+      "learning_rate": 8e-05,
+      "loss": 1.4508,
+      "step": 5743
+    },
+    {
+      "epoch": 0.7839497748055139,
+      "grad_norm": 0.22899247705936432,
+      "learning_rate": 8e-05,
+      "loss": 1.4463,
+      "step": 5744
+    },
+    {
+      "epoch": 0.7840862563122697,
+      "grad_norm": 0.24308471381664276,
+      "learning_rate": 8e-05,
+      "loss": 1.5011,
+      "step": 5745
+    },
+    {
+      "epoch": 0.7842227378190255,
+      "grad_norm": 0.23294997215270996,
+      "learning_rate": 8e-05,
+      "loss": 1.4024,
+      "step": 5746
+    },
+    {
+      "epoch": 0.7843592193257813,
+      "grad_norm": 0.2277994304895401,
+      "learning_rate": 8e-05,
+      "loss": 1.4456,
+      "step": 5747
+    },
+    {
+      "epoch": 0.7844957008325372,
+      "grad_norm": 0.23825617134571075,
+      "learning_rate": 8e-05,
+      "loss": 1.3822,
+      "step": 5748
+    },
+    {
+      "epoch": 0.784632182339293,
+      "grad_norm": 0.24636642634868622,
+      "learning_rate": 8e-05,
+      "loss": 1.4211,
+      "step": 5749
+    },
+    {
+      "epoch": 0.7847686638460488,
+      "grad_norm": 0.2409563660621643,
+      "learning_rate": 8e-05,
+      "loss": 1.4095,
+      "step": 5750
+    },
+    {
+      "epoch": 0.7849051453528046,
+      "grad_norm": 0.2474474310874939,
+      "learning_rate": 8e-05,
+      "loss": 1.4486,
+      "step": 5751
+    },
+    {
+      "epoch": 0.7850416268595606,
+      "grad_norm": 0.24080826342105865,
+      "learning_rate": 8e-05,
+      "loss": 1.4531,
+      "step": 5752
+    },
+    {
+      "epoch": 0.7851781083663164,
+      "grad_norm": 0.2336939126253128,
+      "learning_rate": 8e-05,
+      "loss": 1.4528,
+      "step": 5753
+    },
+    {
+      "epoch": 0.7853145898730722,
+      "grad_norm": 0.23342564702033997,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 5754
+    },
+    {
+      "epoch": 0.7854510713798281,
+      "grad_norm": 0.22775305807590485,
+      "learning_rate": 8e-05,
+      "loss": 1.3998,
+      "step": 5755
+    },
+    {
+      "epoch": 0.7855875528865839,
+      "grad_norm": 0.2438337206840515,
+      "learning_rate": 8e-05,
+      "loss": 1.5008,
+      "step": 5756
+    },
+    {
+      "epoch": 0.7857240343933397,
+      "grad_norm": 0.2393786758184433,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 5757
+    },
+    {
+      "epoch": 0.7858605159000955,
+      "grad_norm": 0.23313851654529572,
+      "learning_rate": 8e-05,
+      "loss": 1.3964,
+      "step": 5758
+    },
+    {
+      "epoch": 0.7859969974068514,
+      "grad_norm": 0.23362167179584503,
+      "learning_rate": 8e-05,
+      "loss": 1.4476,
+      "step": 5759
+    },
+    {
+      "epoch": 0.7861334789136072,
+      "grad_norm": 0.24982048571109772,
+      "learning_rate": 8e-05,
+      "loss": 1.4717,
+      "step": 5760
+    },
+    {
+      "epoch": 0.786269960420363,
+      "grad_norm": 0.24634385108947754,
+      "learning_rate": 8e-05,
+      "loss": 1.425,
+      "step": 5761
+    },
+    {
+      "epoch": 0.7864064419271188,
+      "grad_norm": 0.24184654653072357,
+      "learning_rate": 8e-05,
+      "loss": 1.4611,
+      "step": 5762
+    },
+    {
+      "epoch": 0.7865429234338747,
+      "grad_norm": 0.22984904050827026,
+      "learning_rate": 8e-05,
+      "loss": 1.4161,
+      "step": 5763
+    },
+    {
+      "epoch": 0.7866794049406306,
+      "grad_norm": 0.2341347485780716,
+      "learning_rate": 8e-05,
+      "loss": 1.4024,
+      "step": 5764
+    },
+    {
+      "epoch": 0.7868158864473864,
+      "grad_norm": 0.22549642622470856,
+      "learning_rate": 8e-05,
+      "loss": 1.3682,
+      "step": 5765
+    },
+    {
+      "epoch": 0.7869523679541423,
+      "grad_norm": 0.2493850290775299,
+      "learning_rate": 8e-05,
+      "loss": 1.4396,
+      "step": 5766
+    },
+    {
+      "epoch": 0.7870888494608981,
+      "grad_norm": 0.22926899790763855,
+      "learning_rate": 8e-05,
+      "loss": 1.3776,
+      "step": 5767
+    },
+    {
+      "epoch": 0.7872253309676539,
+      "grad_norm": 0.2457343190908432,
+      "learning_rate": 8e-05,
+      "loss": 1.4704,
+      "step": 5768
+    },
+    {
+      "epoch": 0.7873618124744097,
+      "grad_norm": 0.2367512583732605,
+      "learning_rate": 8e-05,
+      "loss": 1.3657,
+      "step": 5769
+    },
+    {
+      "epoch": 0.7874982939811656,
+      "grad_norm": 0.23465313017368317,
+      "learning_rate": 8e-05,
+      "loss": 1.4309,
+      "step": 5770
+    },
+    {
+      "epoch": 0.7876347754879214,
+      "grad_norm": 0.2475089132785797,
+      "learning_rate": 8e-05,
+      "loss": 1.5076,
+      "step": 5771
+    },
+    {
+      "epoch": 0.7877712569946772,
+      "grad_norm": 0.2477979063987732,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 5772
+    },
+    {
+      "epoch": 0.787907738501433,
+      "grad_norm": 0.22576430439949036,
+      "learning_rate": 8e-05,
+      "loss": 1.4507,
+      "step": 5773
+    },
+    {
+      "epoch": 0.7880442200081889,
+      "grad_norm": 0.23471754789352417,
+      "learning_rate": 8e-05,
+      "loss": 1.4922,
+      "step": 5774
+    },
+    {
+      "epoch": 0.7881807015149447,
+      "grad_norm": 0.24101491272449493,
+      "learning_rate": 8e-05,
+      "loss": 1.5019,
+      "step": 5775
+    },
+    {
+      "epoch": 0.7883171830217005,
+      "grad_norm": 0.24841561913490295,
+      "learning_rate": 8e-05,
+      "loss": 1.4695,
+      "step": 5776
+    },
+    {
+      "epoch": 0.7884536645284564,
+      "grad_norm": 0.25206390023231506,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 5777
+    },
+    {
+      "epoch": 0.7885901460352123,
+      "grad_norm": 0.26844990253448486,
+      "learning_rate": 8e-05,
+      "loss": 1.5151,
+      "step": 5778
+    },
+    {
+      "epoch": 0.7887266275419681,
+      "grad_norm": 0.22879324853420258,
+      "learning_rate": 8e-05,
+      "loss": 1.37,
+      "step": 5779
+    },
+    {
+      "epoch": 0.7888631090487239,
+      "grad_norm": 0.23788219690322876,
+      "learning_rate": 8e-05,
+      "loss": 1.4148,
+      "step": 5780
+    },
+    {
+      "epoch": 0.7889995905554797,
+      "grad_norm": 0.2271324247121811,
+      "learning_rate": 8e-05,
+      "loss": 1.4103,
+      "step": 5781
+    },
+    {
+      "epoch": 0.7891360720622356,
+      "grad_norm": 0.24694578349590302,
+      "learning_rate": 8e-05,
+      "loss": 1.4345,
+      "step": 5782
+    },
+    {
+      "epoch": 0.7892725535689914,
+      "grad_norm": 0.23529550433158875,
+      "learning_rate": 8e-05,
+      "loss": 1.4663,
+      "step": 5783
+    },
+    {
+      "epoch": 0.7894090350757472,
+      "grad_norm": 0.24664567410945892,
+      "learning_rate": 8e-05,
+      "loss": 1.4482,
+      "step": 5784
+    },
+    {
+      "epoch": 0.789545516582503,
+      "grad_norm": 0.2197350710630417,
+      "learning_rate": 8e-05,
+      "loss": 1.3375,
+      "step": 5785
+    },
+    {
+      "epoch": 0.7896819980892589,
+      "grad_norm": 0.23956067860126495,
+      "learning_rate": 8e-05,
+      "loss": 1.4265,
+      "step": 5786
+    },
+    {
+      "epoch": 0.7898184795960147,
+      "grad_norm": 0.23393844068050385,
+      "learning_rate": 8e-05,
+      "loss": 1.4376,
+      "step": 5787
+    },
+    {
+      "epoch": 0.7899549611027705,
+      "grad_norm": 0.23544368147850037,
+      "learning_rate": 8e-05,
+      "loss": 1.4456,
+      "step": 5788
+    },
+    {
+      "epoch": 0.7900914426095265,
+      "grad_norm": 0.2364151030778885,
+      "learning_rate": 8e-05,
+      "loss": 1.4842,
+      "step": 5789
+    },
+    {
+      "epoch": 0.7902279241162823,
+      "grad_norm": 0.23336894810199738,
+      "learning_rate": 8e-05,
+      "loss": 1.4068,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7903644056230381,
+      "grad_norm": 0.23475079238414764,
+      "learning_rate": 8e-05,
+      "loss": 1.4086,
+      "step": 5791
+    },
+    {
+      "epoch": 0.7905008871297939,
+      "grad_norm": 0.24072013795375824,
+      "learning_rate": 8e-05,
+      "loss": 1.4448,
+      "step": 5792
+    },
+    {
+      "epoch": 0.7906373686365498,
+      "grad_norm": 0.24194838106632233,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 5793
+    },
+    {
+      "epoch": 0.7907738501433056,
+      "grad_norm": 0.2304919958114624,
+      "learning_rate": 8e-05,
+      "loss": 1.432,
+      "step": 5794
+    },
+    {
+      "epoch": 0.7909103316500614,
+      "grad_norm": 0.23333878815174103,
+      "learning_rate": 8e-05,
+      "loss": 1.4229,
+      "step": 5795
+    },
+    {
+      "epoch": 0.7910468131568172,
+      "grad_norm": 0.22615814208984375,
+      "learning_rate": 8e-05,
+      "loss": 1.4414,
+      "step": 5796
+    },
+    {
+      "epoch": 0.7911832946635731,
+      "grad_norm": 0.23355723917484283,
+      "learning_rate": 8e-05,
+      "loss": 1.4421,
+      "step": 5797
+    },
+    {
+      "epoch": 0.7913197761703289,
+      "grad_norm": 0.25367406010627747,
+      "learning_rate": 8e-05,
+      "loss": 1.4863,
+      "step": 5798
+    },
+    {
+      "epoch": 0.7914562576770847,
+      "grad_norm": 0.23473483324050903,
+      "learning_rate": 8e-05,
+      "loss": 1.4234,
+      "step": 5799
+    },
+    {
+      "epoch": 0.7915927391838405,
+      "grad_norm": 0.23245465755462646,
+      "learning_rate": 8e-05,
+      "loss": 1.4088,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7917292206905964,
+      "grad_norm": 0.23365731537342072,
+      "learning_rate": 8e-05,
+      "loss": 1.4332,
+      "step": 5801
+    },
+    {
+      "epoch": 0.7918657021973523,
+      "grad_norm": 0.24205321073532104,
+      "learning_rate": 8e-05,
+      "loss": 1.4949,
+      "step": 5802
+    },
+    {
+      "epoch": 0.7920021837041081,
+      "grad_norm": 0.2506902515888214,
+      "learning_rate": 8e-05,
+      "loss": 1.5303,
+      "step": 5803
+    },
+    {
+      "epoch": 0.792138665210864,
+      "grad_norm": 0.2402496486902237,
+      "learning_rate": 8e-05,
+      "loss": 1.4584,
+      "step": 5804
+    },
+    {
+      "epoch": 0.7922751467176198,
+      "grad_norm": 0.240568608045578,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7924116282243756,
+      "grad_norm": 0.2441534847021103,
+      "learning_rate": 8e-05,
+      "loss": 1.4719,
+      "step": 5806
+    },
+    {
+      "epoch": 0.7925481097311314,
+      "grad_norm": 0.2453349232673645,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 5807
+    },
+    {
+      "epoch": 0.7926845912378873,
+      "grad_norm": 0.22573626041412354,
+      "learning_rate": 8e-05,
+      "loss": 1.3912,
+      "step": 5808
+    },
+    {
+      "epoch": 0.7928210727446431,
+      "grad_norm": 0.24056889116764069,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 5809
+    },
+    {
+      "epoch": 0.7929575542513989,
+      "grad_norm": 0.24801048636436462,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 5810
+    },
+    {
+      "epoch": 0.7930940357581547,
+      "grad_norm": 0.23811832070350647,
+      "learning_rate": 8e-05,
+      "loss": 1.3905,
+      "step": 5811
+    },
+    {
+      "epoch": 0.7932305172649106,
+      "grad_norm": 0.23620720207691193,
+      "learning_rate": 8e-05,
+      "loss": 1.4383,
+      "step": 5812
+    },
+    {
+      "epoch": 0.7933669987716664,
+      "grad_norm": 0.22912539541721344,
+      "learning_rate": 8e-05,
+      "loss": 1.4018,
+      "step": 5813
+    },
+    {
+      "epoch": 0.7935034802784223,
+      "grad_norm": 0.24289178848266602,
+      "learning_rate": 8e-05,
+      "loss": 1.4613,
+      "step": 5814
+    },
+    {
+      "epoch": 0.7936399617851781,
+      "grad_norm": 0.2459285706281662,
+      "learning_rate": 8e-05,
+      "loss": 1.4901,
+      "step": 5815
+    },
+    {
+      "epoch": 0.793776443291934,
+      "grad_norm": 0.239617258310318,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 5816
+    },
+    {
+      "epoch": 0.7939129247986898,
+      "grad_norm": 0.24111562967300415,
+      "learning_rate": 8e-05,
+      "loss": 1.4356,
+      "step": 5817
+    },
+    {
+      "epoch": 0.7940494063054456,
+      "grad_norm": 0.23637214303016663,
+      "learning_rate": 8e-05,
+      "loss": 1.4121,
+      "step": 5818
+    },
+    {
+      "epoch": 0.7941858878122015,
+      "grad_norm": 0.24557903409004211,
+      "learning_rate": 8e-05,
+      "loss": 1.4186,
+      "step": 5819
+    },
+    {
+      "epoch": 0.7943223693189573,
+      "grad_norm": 0.24201151728630066,
+      "learning_rate": 8e-05,
+      "loss": 1.477,
+      "step": 5820
+    },
+    {
+      "epoch": 0.7944588508257131,
+      "grad_norm": 0.23630541563034058,
+      "learning_rate": 8e-05,
+      "loss": 1.3717,
+      "step": 5821
+    },
+    {
+      "epoch": 0.7945953323324689,
+      "grad_norm": 0.23564426600933075,
+      "learning_rate": 8e-05,
+      "loss": 1.4142,
+      "step": 5822
+    },
+    {
+      "epoch": 0.7947318138392248,
+      "grad_norm": 0.24715179204940796,
+      "learning_rate": 8e-05,
+      "loss": 1.4402,
+      "step": 5823
+    },
+    {
+      "epoch": 0.7948682953459806,
+      "grad_norm": 0.24091872572898865,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 5824
+    },
+    {
+      "epoch": 0.7950047768527364,
+      "grad_norm": 0.2430531084537506,
+      "learning_rate": 8e-05,
+      "loss": 1.4111,
+      "step": 5825
+    },
+    {
+      "epoch": 0.7951412583594923,
+      "grad_norm": 0.2375899702310562,
+      "learning_rate": 8e-05,
+      "loss": 1.4413,
+      "step": 5826
+    },
+    {
+      "epoch": 0.7952777398662482,
+      "grad_norm": 0.25006020069122314,
+      "learning_rate": 8e-05,
+      "loss": 1.4227,
+      "step": 5827
+    },
+    {
+      "epoch": 0.795414221373004,
+      "grad_norm": 0.23300251364707947,
+      "learning_rate": 8e-05,
+      "loss": 1.427,
+      "step": 5828
+    },
+    {
+      "epoch": 0.7955507028797598,
+      "grad_norm": 0.23727747797966003,
+      "learning_rate": 8e-05,
+      "loss": 1.4795,
+      "step": 5829
+    },
+    {
+      "epoch": 0.7956871843865156,
+      "grad_norm": 0.22807112336158752,
+      "learning_rate": 8e-05,
+      "loss": 1.3708,
+      "step": 5830
+    },
+    {
+      "epoch": 0.7958236658932715,
+      "grad_norm": 0.22905097901821136,
+      "learning_rate": 8e-05,
+      "loss": 1.4038,
+      "step": 5831
+    },
+    {
+      "epoch": 0.7959601474000273,
+      "grad_norm": 0.2557128071784973,
+      "learning_rate": 8e-05,
+      "loss": 1.4473,
+      "step": 5832
+    },
+    {
+      "epoch": 0.7960966289067831,
+      "grad_norm": 0.23994258046150208,
+      "learning_rate": 8e-05,
+      "loss": 1.3798,
+      "step": 5833
+    },
+    {
+      "epoch": 0.796233110413539,
+      "grad_norm": 0.24385660886764526,
+      "learning_rate": 8e-05,
+      "loss": 1.4311,
+      "step": 5834
+    },
+    {
+      "epoch": 0.7963695919202948,
+      "grad_norm": 0.23541179299354553,
+      "learning_rate": 8e-05,
+      "loss": 1.392,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7965060734270506,
+      "grad_norm": 0.248274564743042,
+      "learning_rate": 8e-05,
+      "loss": 1.4952,
+      "step": 5836
+    },
+    {
+      "epoch": 0.7966425549338064,
+      "grad_norm": 0.23888981342315674,
+      "learning_rate": 8e-05,
+      "loss": 1.4273,
+      "step": 5837
+    },
+    {
+      "epoch": 0.7967790364405622,
+      "grad_norm": 0.24138200283050537,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 5838
+    },
+    {
+      "epoch": 0.7969155179473182,
+      "grad_norm": 0.2429000735282898,
+      "learning_rate": 8e-05,
+      "loss": 1.5035,
+      "step": 5839
+    },
+    {
+      "epoch": 0.797051999454074,
+      "grad_norm": 0.23733998835086823,
+      "learning_rate": 8e-05,
+      "loss": 1.4668,
+      "step": 5840
+    },
+    {
+      "epoch": 0.7971884809608298,
+      "grad_norm": 0.23367184400558472,
+      "learning_rate": 8e-05,
+      "loss": 1.4765,
+      "step": 5841
+    },
+    {
+      "epoch": 0.7973249624675857,
+      "grad_norm": 0.22999729216098785,
+      "learning_rate": 8e-05,
+      "loss": 1.412,
+      "step": 5842
+    },
+    {
+      "epoch": 0.7974614439743415,
+      "grad_norm": 0.22672493755817413,
+      "learning_rate": 8e-05,
+      "loss": 1.3822,
+      "step": 5843
+    },
+    {
+      "epoch": 0.7975979254810973,
+      "grad_norm": 0.23378343880176544,
+      "learning_rate": 8e-05,
+      "loss": 1.3647,
+      "step": 5844
+    },
+    {
+      "epoch": 0.7977344069878531,
+      "grad_norm": 0.24148738384246826,
+      "learning_rate": 8e-05,
+      "loss": 1.4245,
+      "step": 5845
+    },
+    {
+      "epoch": 0.797870888494609,
+      "grad_norm": 0.2250978648662567,
+      "learning_rate": 8e-05,
+      "loss": 1.3769,
+      "step": 5846
+    },
+    {
+      "epoch": 0.7980073700013648,
+      "grad_norm": 0.2321162074804306,
+      "learning_rate": 8e-05,
+      "loss": 1.403,
+      "step": 5847
+    },
+    {
+      "epoch": 0.7981438515081206,
+      "grad_norm": 0.24674364924430847,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 5848
+    },
+    {
+      "epoch": 0.7982803330148764,
+      "grad_norm": 0.24724207818508148,
+      "learning_rate": 8e-05,
+      "loss": 1.4511,
+      "step": 5849
+    },
+    {
+      "epoch": 0.7984168145216323,
+      "grad_norm": 0.23696810007095337,
+      "learning_rate": 8e-05,
+      "loss": 1.4158,
+      "step": 5850
+    },
+    {
+      "epoch": 0.7985532960283882,
+      "grad_norm": 0.2390514463186264,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 5851
+    },
+    {
+      "epoch": 0.798689777535144,
+      "grad_norm": 0.23427166044712067,
+      "learning_rate": 8e-05,
+      "loss": 1.3764,
+      "step": 5852
+    },
+    {
+      "epoch": 0.7988262590418999,
+      "grad_norm": 0.23149864375591278,
+      "learning_rate": 8e-05,
+      "loss": 1.4326,
+      "step": 5853
+    },
+    {
+      "epoch": 0.7989627405486557,
+      "grad_norm": 0.23544394969940186,
+      "learning_rate": 8e-05,
+      "loss": 1.5142,
+      "step": 5854
+    },
+    {
+      "epoch": 0.7990992220554115,
+      "grad_norm": 0.2442428022623062,
+      "learning_rate": 8e-05,
+      "loss": 1.4347,
+      "step": 5855
+    },
+    {
+      "epoch": 0.7992357035621673,
+      "grad_norm": 0.24647310376167297,
+      "learning_rate": 8e-05,
+      "loss": 1.4717,
+      "step": 5856
+    },
+    {
+      "epoch": 0.7993721850689232,
+      "grad_norm": 0.23924070596694946,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 5857
+    },
+    {
+      "epoch": 0.799508666575679,
+      "grad_norm": 0.236973837018013,
+      "learning_rate": 8e-05,
+      "loss": 1.3933,
+      "step": 5858
+    },
+    {
+      "epoch": 0.7996451480824348,
+      "grad_norm": 0.25349777936935425,
+      "learning_rate": 8e-05,
+      "loss": 1.4755,
+      "step": 5859
+    },
+    {
+      "epoch": 0.7997816295891906,
+      "grad_norm": 0.2330014854669571,
+      "learning_rate": 8e-05,
+      "loss": 1.486,
+      "step": 5860
+    },
+    {
+      "epoch": 0.7999181110959465,
+      "grad_norm": 0.24783207476139069,
+      "learning_rate": 8e-05,
+      "loss": 1.4333,
+      "step": 5861
+    },
+    {
+      "epoch": 0.8000545926027023,
+      "grad_norm": 0.2516767680644989,
+      "learning_rate": 8e-05,
+      "loss": 1.4884,
+      "step": 5862
+    },
+    {
+      "epoch": 0.8001910741094582,
+      "grad_norm": 0.2340412139892578,
+      "learning_rate": 8e-05,
+      "loss": 1.4092,
+      "step": 5863
+    },
+    {
+      "epoch": 0.800327555616214,
+      "grad_norm": 0.24123387038707733,
+      "learning_rate": 8e-05,
+      "loss": 1.3885,
+      "step": 5864
+    },
+    {
+      "epoch": 0.8004640371229699,
+      "grad_norm": 0.2422478049993515,
+      "learning_rate": 8e-05,
+      "loss": 1.4936,
+      "step": 5865
+    },
+    {
+      "epoch": 0.8006005186297257,
+      "grad_norm": 0.2396695464849472,
+      "learning_rate": 8e-05,
+      "loss": 1.4435,
+      "step": 5866
+    },
+    {
+      "epoch": 0.8007370001364815,
+      "grad_norm": 0.23689812421798706,
+      "learning_rate": 8e-05,
+      "loss": 1.4554,
+      "step": 5867
+    },
+    {
+      "epoch": 0.8008734816432374,
+      "grad_norm": 0.23886382579803467,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 5868
+    },
+    {
+      "epoch": 0.8010099631499932,
+      "grad_norm": 0.23431850969791412,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 5869
+    },
+    {
+      "epoch": 0.801146444656749,
+      "grad_norm": 0.2402949333190918,
+      "learning_rate": 8e-05,
+      "loss": 1.4972,
+      "step": 5870
+    },
+    {
+      "epoch": 0.8012829261635048,
+      "grad_norm": 0.25060686469078064,
+      "learning_rate": 8e-05,
+      "loss": 1.4427,
+      "step": 5871
+    },
+    {
+      "epoch": 0.8014194076702607,
+      "grad_norm": 0.23107130825519562,
+      "learning_rate": 8e-05,
+      "loss": 1.389,
+      "step": 5872
+    },
+    {
+      "epoch": 0.8015558891770165,
+      "grad_norm": 0.24028481543064117,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 5873
+    },
+    {
+      "epoch": 0.8016923706837723,
+      "grad_norm": 0.23434941470623016,
+      "learning_rate": 8e-05,
+      "loss": 1.4166,
+      "step": 5874
+    },
+    {
+      "epoch": 0.8018288521905281,
+      "grad_norm": 0.2351999282836914,
+      "learning_rate": 8e-05,
+      "loss": 1.4357,
+      "step": 5875
+    },
+    {
+      "epoch": 0.8019653336972841,
+      "grad_norm": 0.2402610331773758,
+      "learning_rate": 8e-05,
+      "loss": 1.415,
+      "step": 5876
+    },
+    {
+      "epoch": 0.8021018152040399,
+      "grad_norm": 0.23841916024684906,
+      "learning_rate": 8e-05,
+      "loss": 1.4514,
+      "step": 5877
+    },
+    {
+      "epoch": 0.8022382967107957,
+      "grad_norm": 0.24063760042190552,
+      "learning_rate": 8e-05,
+      "loss": 1.4285,
+      "step": 5878
+    },
+    {
+      "epoch": 0.8023747782175515,
+      "grad_norm": 0.2387036234140396,
+      "learning_rate": 8e-05,
+      "loss": 1.4689,
+      "step": 5879
+    },
+    {
+      "epoch": 0.8025112597243074,
+      "grad_norm": 0.23176898062229156,
+      "learning_rate": 8e-05,
+      "loss": 1.4389,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8026477412310632,
+      "grad_norm": 0.2307896763086319,
+      "learning_rate": 8e-05,
+      "loss": 1.4082,
+      "step": 5881
+    },
+    {
+      "epoch": 0.802784222737819,
+      "grad_norm": 0.2440449446439743,
+      "learning_rate": 8e-05,
+      "loss": 1.4287,
+      "step": 5882
+    },
+    {
+      "epoch": 0.8029207042445748,
+      "grad_norm": 0.23851165175437927,
+      "learning_rate": 8e-05,
+      "loss": 1.4477,
+      "step": 5883
+    },
+    {
+      "epoch": 0.8030571857513307,
+      "grad_norm": 0.23855364322662354,
+      "learning_rate": 8e-05,
+      "loss": 1.4599,
+      "step": 5884
+    },
+    {
+      "epoch": 0.8031936672580865,
+      "grad_norm": 0.24413608014583588,
+      "learning_rate": 8e-05,
+      "loss": 1.4607,
+      "step": 5885
+    },
+    {
+      "epoch": 0.8033301487648423,
+      "grad_norm": 0.23794734477996826,
+      "learning_rate": 8e-05,
+      "loss": 1.4959,
+      "step": 5886
+    },
+    {
+      "epoch": 0.8034666302715981,
+      "grad_norm": 0.237937331199646,
+      "learning_rate": 8e-05,
+      "loss": 1.466,
+      "step": 5887
+    },
+    {
+      "epoch": 0.8036031117783541,
+      "grad_norm": 0.2399560958147049,
+      "learning_rate": 8e-05,
+      "loss": 1.4419,
+      "step": 5888
+    },
+    {
+      "epoch": 0.8037395932851099,
+      "grad_norm": 0.23926006257534027,
+      "learning_rate": 8e-05,
+      "loss": 1.3714,
+      "step": 5889
+    },
+    {
+      "epoch": 0.8038760747918657,
+      "grad_norm": 0.23486462235450745,
+      "learning_rate": 8e-05,
+      "loss": 1.4651,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8040125562986216,
+      "grad_norm": 0.24520274996757507,
+      "learning_rate": 8e-05,
+      "loss": 1.4521,
+      "step": 5891
+    },
+    {
+      "epoch": 0.8041490378053774,
+      "grad_norm": 0.25203585624694824,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 5892
+    },
+    {
+      "epoch": 0.8042855193121332,
+      "grad_norm": 0.24074327945709229,
+      "learning_rate": 8e-05,
+      "loss": 1.4297,
+      "step": 5893
+    },
+    {
+      "epoch": 0.804422000818889,
+      "grad_norm": 0.23027200996875763,
+      "learning_rate": 8e-05,
+      "loss": 1.4871,
+      "step": 5894
+    },
+    {
+      "epoch": 0.8045584823256449,
+      "grad_norm": 0.24325014650821686,
+      "learning_rate": 8e-05,
+      "loss": 1.4093,
+      "step": 5895
+    },
+    {
+      "epoch": 0.8046949638324007,
+      "grad_norm": 0.23581823706626892,
+      "learning_rate": 8e-05,
+      "loss": 1.3517,
+      "step": 5896
+    },
+    {
+      "epoch": 0.8048314453391565,
+      "grad_norm": 0.2278321087360382,
+      "learning_rate": 8e-05,
+      "loss": 1.4118,
+      "step": 5897
+    },
+    {
+      "epoch": 0.8049679268459123,
+      "grad_norm": 0.2396267056465149,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 5898
+    },
+    {
+      "epoch": 0.8051044083526682,
+      "grad_norm": 0.2458927184343338,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 5899
+    },
+    {
+      "epoch": 0.8052408898594241,
+      "grad_norm": 0.23562055826187134,
+      "learning_rate": 8e-05,
+      "loss": 1.4458,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8053773713661799,
+      "grad_norm": 0.237859308719635,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 5901
+    },
+    {
+      "epoch": 0.8055138528729358,
+      "grad_norm": 0.2368531972169876,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 5902
+    },
+    {
+      "epoch": 0.8056503343796916,
+      "grad_norm": 0.23780620098114014,
+      "learning_rate": 8e-05,
+      "loss": 1.4279,
+      "step": 5903
+    },
+    {
+      "epoch": 0.8057868158864474,
+      "grad_norm": 0.2389354705810547,
+      "learning_rate": 8e-05,
+      "loss": 1.4008,
+      "step": 5904
+    },
+    {
+      "epoch": 0.8059232973932032,
+      "grad_norm": 0.23608151078224182,
+      "learning_rate": 8e-05,
+      "loss": 1.4178,
+      "step": 5905
+    },
+    {
+      "epoch": 0.8060597788999591,
+      "grad_norm": 0.23239582777023315,
+      "learning_rate": 8e-05,
+      "loss": 1.4277,
+      "step": 5906
+    },
+    {
+      "epoch": 0.8061962604067149,
+      "grad_norm": 0.2311106026172638,
+      "learning_rate": 8e-05,
+      "loss": 1.4384,
+      "step": 5907
+    },
+    {
+      "epoch": 0.8063327419134707,
+      "grad_norm": 0.23089265823364258,
+      "learning_rate": 8e-05,
+      "loss": 1.4571,
+      "step": 5908
+    },
+    {
+      "epoch": 0.8064692234202265,
+      "grad_norm": 0.2423102706670761,
+      "learning_rate": 8e-05,
+      "loss": 1.4439,
+      "step": 5909
+    },
+    {
+      "epoch": 0.8066057049269824,
+      "grad_norm": 0.23792174458503723,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8067421864337382,
+      "grad_norm": 0.237611323595047,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 5911
+    },
+    {
+      "epoch": 0.806878667940494,
+      "grad_norm": 0.23351748287677765,
+      "learning_rate": 8e-05,
+      "loss": 1.369,
+      "step": 5912
+    },
+    {
+      "epoch": 0.80701514944725,
+      "grad_norm": 0.23891650140285492,
+      "learning_rate": 8e-05,
+      "loss": 1.4121,
+      "step": 5913
+    },
+    {
+      "epoch": 0.8071516309540058,
+      "grad_norm": 0.25928664207458496,
+      "learning_rate": 8e-05,
+      "loss": 1.5029,
+      "step": 5914
+    },
+    {
+      "epoch": 0.8072881124607616,
+      "grad_norm": 0.2522110044956207,
+      "learning_rate": 8e-05,
+      "loss": 1.4541,
+      "step": 5915
+    },
+    {
+      "epoch": 0.8074245939675174,
+      "grad_norm": 0.23376616835594177,
+      "learning_rate": 8e-05,
+      "loss": 1.4315,
+      "step": 5916
+    },
+    {
+      "epoch": 0.8075610754742732,
+      "grad_norm": 0.2309272289276123,
+      "learning_rate": 8e-05,
+      "loss": 1.485,
+      "step": 5917
+    },
+    {
+      "epoch": 0.8076975569810291,
+      "grad_norm": 0.23657318949699402,
+      "learning_rate": 8e-05,
+      "loss": 1.4249,
+      "step": 5918
+    },
+    {
+      "epoch": 0.8078340384877849,
+      "grad_norm": 0.24033544957637787,
+      "learning_rate": 8e-05,
+      "loss": 1.455,
+      "step": 5919
+    },
+    {
+      "epoch": 0.8079705199945407,
+      "grad_norm": 0.22799037396907806,
+      "learning_rate": 8e-05,
+      "loss": 1.4172,
+      "step": 5920
+    },
+    {
+      "epoch": 0.8081070015012966,
+      "grad_norm": 0.23210591077804565,
+      "learning_rate": 8e-05,
+      "loss": 1.407,
+      "step": 5921
+    },
+    {
+      "epoch": 0.8082434830080524,
+      "grad_norm": 0.24740038812160492,
+      "learning_rate": 8e-05,
+      "loss": 1.4427,
+      "step": 5922
+    },
+    {
+      "epoch": 0.8083799645148082,
+      "grad_norm": 0.25189754366874695,
+      "learning_rate": 8e-05,
+      "loss": 1.4061,
+      "step": 5923
+    },
+    {
+      "epoch": 0.808516446021564,
+      "grad_norm": 0.24986734986305237,
+      "learning_rate": 8e-05,
+      "loss": 1.4796,
+      "step": 5924
+    },
+    {
+      "epoch": 0.80865292752832,
+      "grad_norm": 0.24241511523723602,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 5925
+    },
+    {
+      "epoch": 0.8087894090350758,
+      "grad_norm": 0.2458614856004715,
+      "learning_rate": 8e-05,
+      "loss": 1.4674,
+      "step": 5926
+    },
+    {
+      "epoch": 0.8089258905418316,
+      "grad_norm": 0.24609558284282684,
+      "learning_rate": 8e-05,
+      "loss": 1.4593,
+      "step": 5927
+    },
+    {
+      "epoch": 0.8090623720485874,
+      "grad_norm": 0.22759337723255157,
+      "learning_rate": 8e-05,
+      "loss": 1.3743,
+      "step": 5928
+    },
+    {
+      "epoch": 0.8091988535553433,
+      "grad_norm": 0.23357951641082764,
+      "learning_rate": 8e-05,
+      "loss": 1.4003,
+      "step": 5929
+    },
+    {
+      "epoch": 0.8093353350620991,
+      "grad_norm": 0.23718619346618652,
+      "learning_rate": 8e-05,
+      "loss": 1.3949,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8094718165688549,
+      "grad_norm": 0.23792606592178345,
+      "learning_rate": 8e-05,
+      "loss": 1.4581,
+      "step": 5931
+    },
+    {
+      "epoch": 0.8096082980756107,
+      "grad_norm": 0.23944678902626038,
+      "learning_rate": 8e-05,
+      "loss": 1.4094,
+      "step": 5932
+    },
+    {
+      "epoch": 0.8097447795823666,
+      "grad_norm": 0.24578560888767242,
+      "learning_rate": 8e-05,
+      "loss": 1.4737,
+      "step": 5933
+    },
+    {
+      "epoch": 0.8098812610891224,
+      "grad_norm": 0.23896247148513794,
+      "learning_rate": 8e-05,
+      "loss": 1.3881,
+      "step": 5934
+    },
+    {
+      "epoch": 0.8100177425958782,
+      "grad_norm": 0.2452411949634552,
+      "learning_rate": 8e-05,
+      "loss": 1.4772,
+      "step": 5935
+    },
+    {
+      "epoch": 0.810154224102634,
+      "grad_norm": 0.2374831736087799,
+      "learning_rate": 8e-05,
+      "loss": 1.454,
+      "step": 5936
+    },
+    {
+      "epoch": 0.81029070560939,
+      "grad_norm": 0.2370942085981369,
+      "learning_rate": 8e-05,
+      "loss": 1.3752,
+      "step": 5937
+    },
+    {
+      "epoch": 0.8104271871161458,
+      "grad_norm": 0.23482942581176758,
+      "learning_rate": 8e-05,
+      "loss": 1.3415,
+      "step": 5938
+    },
+    {
+      "epoch": 0.8105636686229016,
+      "grad_norm": 0.23110154271125793,
+      "learning_rate": 8e-05,
+      "loss": 1.3819,
+      "step": 5939
+    },
+    {
+      "epoch": 0.8107001501296575,
+      "grad_norm": 0.2393108457326889,
+      "learning_rate": 8e-05,
+      "loss": 1.4367,
+      "step": 5940
+    },
+    {
+      "epoch": 0.8108366316364133,
+      "grad_norm": 0.2416023164987564,
+      "learning_rate": 8e-05,
+      "loss": 1.509,
+      "step": 5941
+    },
+    {
+      "epoch": 0.8109731131431691,
+      "grad_norm": 0.24556922912597656,
+      "learning_rate": 8e-05,
+      "loss": 1.4722,
+      "step": 5942
+    },
+    {
+      "epoch": 0.8111095946499249,
+      "grad_norm": 0.24086499214172363,
+      "learning_rate": 8e-05,
+      "loss": 1.4327,
+      "step": 5943
+    },
+    {
+      "epoch": 0.8112460761566808,
+      "grad_norm": 0.23886358737945557,
+      "learning_rate": 8e-05,
+      "loss": 1.411,
+      "step": 5944
+    },
+    {
+      "epoch": 0.8113825576634366,
+      "grad_norm": 0.2562863826751709,
+      "learning_rate": 8e-05,
+      "loss": 1.4752,
+      "step": 5945
+    },
+    {
+      "epoch": 0.8115190391701924,
+      "grad_norm": 0.2509612739086151,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 5946
+    },
+    {
+      "epoch": 0.8116555206769482,
+      "grad_norm": 0.2386801689863205,
+      "learning_rate": 8e-05,
+      "loss": 1.4172,
+      "step": 5947
+    },
+    {
+      "epoch": 0.8117920021837041,
+      "grad_norm": 0.2328961342573166,
+      "learning_rate": 8e-05,
+      "loss": 1.407,
+      "step": 5948
+    },
+    {
+      "epoch": 0.8119284836904599,
+      "grad_norm": 0.23096033930778503,
+      "learning_rate": 8e-05,
+      "loss": 1.4256,
+      "step": 5949
+    },
+    {
+      "epoch": 0.8120649651972158,
+      "grad_norm": 0.2541479170322418,
+      "learning_rate": 8e-05,
+      "loss": 1.5027,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8122014467039717,
+      "grad_norm": 0.24616189301013947,
+      "learning_rate": 8e-05,
+      "loss": 1.4782,
+      "step": 5951
+    },
+    {
+      "epoch": 0.8123379282107275,
+      "grad_norm": 0.24321970343589783,
+      "learning_rate": 8e-05,
+      "loss": 1.3949,
+      "step": 5952
+    },
+    {
+      "epoch": 0.8124744097174833,
+      "grad_norm": 0.24209776520729065,
+      "learning_rate": 8e-05,
+      "loss": 1.4533,
+      "step": 5953
+    },
+    {
+      "epoch": 0.8126108912242391,
+      "grad_norm": 0.24842603504657745,
+      "learning_rate": 8e-05,
+      "loss": 1.423,
+      "step": 5954
+    },
+    {
+      "epoch": 0.812747372730995,
+      "grad_norm": 0.2409479022026062,
+      "learning_rate": 8e-05,
+      "loss": 1.3904,
+      "step": 5955
+    },
+    {
+      "epoch": 0.8128838542377508,
+      "grad_norm": 0.2370232343673706,
+      "learning_rate": 8e-05,
+      "loss": 1.466,
+      "step": 5956
+    },
+    {
+      "epoch": 0.8130203357445066,
+      "grad_norm": 0.23967401683330536,
+      "learning_rate": 8e-05,
+      "loss": 1.4501,
+      "step": 5957
+    },
+    {
+      "epoch": 0.8131568172512624,
+      "grad_norm": 0.23843559622764587,
+      "learning_rate": 8e-05,
+      "loss": 1.4027,
+      "step": 5958
+    },
+    {
+      "epoch": 0.8132932987580183,
+      "grad_norm": 0.2317013293504715,
+      "learning_rate": 8e-05,
+      "loss": 1.4067,
+      "step": 5959
+    },
+    {
+      "epoch": 0.8134297802647741,
+      "grad_norm": 0.2434590607881546,
+      "learning_rate": 8e-05,
+      "loss": 1.3923,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8135662617715299,
+      "grad_norm": 0.22869227826595306,
+      "learning_rate": 8e-05,
+      "loss": 1.3462,
+      "step": 5961
+    },
+    {
+      "epoch": 0.8137027432782858,
+      "grad_norm": 0.2440468966960907,
+      "learning_rate": 8e-05,
+      "loss": 1.392,
+      "step": 5962
+    },
+    {
+      "epoch": 0.8138392247850417,
+      "grad_norm": 0.2343759834766388,
+      "learning_rate": 8e-05,
+      "loss": 1.3712,
+      "step": 5963
+    },
+    {
+      "epoch": 0.8139757062917975,
+      "grad_norm": 0.2400575429201126,
+      "learning_rate": 8e-05,
+      "loss": 1.4022,
+      "step": 5964
+    },
+    {
+      "epoch": 0.8141121877985533,
+      "grad_norm": 0.23770837485790253,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 5965
+    },
+    {
+      "epoch": 0.8142486693053091,
+      "grad_norm": 0.23036539554595947,
+      "learning_rate": 8e-05,
+      "loss": 1.4232,
+      "step": 5966
+    },
+    {
+      "epoch": 0.814385150812065,
+      "grad_norm": 0.24349334836006165,
+      "learning_rate": 8e-05,
+      "loss": 1.4061,
+      "step": 5967
+    },
+    {
+      "epoch": 0.8145216323188208,
+      "grad_norm": 0.24997101724147797,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 5968
+    },
+    {
+      "epoch": 0.8146581138255766,
+      "grad_norm": 0.24676865339279175,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 5969
+    },
+    {
+      "epoch": 0.8147945953323324,
+      "grad_norm": 0.2559300363063812,
+      "learning_rate": 8e-05,
+      "loss": 1.5051,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8149310768390883,
+      "grad_norm": 0.23717670142650604,
+      "learning_rate": 8e-05,
+      "loss": 1.4996,
+      "step": 5971
+    },
+    {
+      "epoch": 0.8150675583458441,
+      "grad_norm": 0.24250823259353638,
+      "learning_rate": 8e-05,
+      "loss": 1.4873,
+      "step": 5972
+    },
+    {
+      "epoch": 0.8152040398525999,
+      "grad_norm": 0.23389583826065063,
+      "learning_rate": 8e-05,
+      "loss": 1.455,
+      "step": 5973
+    },
+    {
+      "epoch": 0.8153405213593558,
+      "grad_norm": 0.24449385702610016,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 5974
+    },
+    {
+      "epoch": 0.8154770028661117,
+      "grad_norm": 0.2449869066476822,
+      "learning_rate": 8e-05,
+      "loss": 1.4597,
+      "step": 5975
+    },
+    {
+      "epoch": 0.8156134843728675,
+      "grad_norm": 0.24148793518543243,
+      "learning_rate": 8e-05,
+      "loss": 1.4208,
+      "step": 5976
+    },
+    {
+      "epoch": 0.8157499658796233,
+      "grad_norm": 0.24427126348018646,
+      "learning_rate": 8e-05,
+      "loss": 1.4333,
+      "step": 5977
+    },
+    {
+      "epoch": 0.8158864473863792,
+      "grad_norm": 0.23902970552444458,
+      "learning_rate": 8e-05,
+      "loss": 1.4883,
+      "step": 5978
+    },
+    {
+      "epoch": 0.816022928893135,
+      "grad_norm": 0.2562556564807892,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 5979
+    },
+    {
+      "epoch": 0.8161594103998908,
+      "grad_norm": 0.2340439260005951,
+      "learning_rate": 8e-05,
+      "loss": 1.3745,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8162958919066466,
+      "grad_norm": 0.22747012972831726,
+      "learning_rate": 8e-05,
+      "loss": 1.4594,
+      "step": 5981
+    },
+    {
+      "epoch": 0.8164323734134025,
+      "grad_norm": 0.2271149903535843,
+      "learning_rate": 8e-05,
+      "loss": 1.3993,
+      "step": 5982
+    },
+    {
+      "epoch": 0.8165688549201583,
+      "grad_norm": 0.2450660765171051,
+      "learning_rate": 8e-05,
+      "loss": 1.414,
+      "step": 5983
+    },
+    {
+      "epoch": 0.8167053364269141,
+      "grad_norm": 0.23845788836479187,
+      "learning_rate": 8e-05,
+      "loss": 1.4452,
+      "step": 5984
+    },
+    {
+      "epoch": 0.81684181793367,
+      "grad_norm": 0.24085500836372375,
+      "learning_rate": 8e-05,
+      "loss": 1.4581,
+      "step": 5985
+    },
+    {
+      "epoch": 0.8169782994404258,
+      "grad_norm": 0.24057720601558685,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 5986
+    },
+    {
+      "epoch": 0.8171147809471817,
+      "grad_norm": 0.2450377643108368,
+      "learning_rate": 8e-05,
+      "loss": 1.49,
+      "step": 5987
+    },
+    {
+      "epoch": 0.8172512624539375,
+      "grad_norm": 0.24217212200164795,
+      "learning_rate": 8e-05,
+      "loss": 1.4426,
+      "step": 5988
+    },
+    {
+      "epoch": 0.8173877439606934,
+      "grad_norm": 0.23991000652313232,
+      "learning_rate": 8e-05,
+      "loss": 1.4512,
+      "step": 5989
+    },
+    {
+      "epoch": 0.8175242254674492,
+      "grad_norm": 0.23082904517650604,
+      "learning_rate": 8e-05,
+      "loss": 1.4379,
+      "step": 5990
+    },
+    {
+      "epoch": 0.817660706974205,
+      "grad_norm": 0.23655807971954346,
+      "learning_rate": 8e-05,
+      "loss": 1.376,
+      "step": 5991
+    },
+    {
+      "epoch": 0.8177971884809608,
+      "grad_norm": 0.2400001734495163,
+      "learning_rate": 8e-05,
+      "loss": 1.4429,
+      "step": 5992
+    },
+    {
+      "epoch": 0.8179336699877167,
+      "grad_norm": 0.24617899954319,
+      "learning_rate": 8e-05,
+      "loss": 1.5007,
+      "step": 5993
+    },
+    {
+      "epoch": 0.8180701514944725,
+      "grad_norm": 0.24426016211509705,
+      "learning_rate": 8e-05,
+      "loss": 1.4215,
+      "step": 5994
+    },
+    {
+      "epoch": 0.8182066330012283,
+      "grad_norm": 0.2322605848312378,
+      "learning_rate": 8e-05,
+      "loss": 1.4271,
+      "step": 5995
+    },
+    {
+      "epoch": 0.8183431145079841,
+      "grad_norm": 0.23960435390472412,
+      "learning_rate": 8e-05,
+      "loss": 1.4642,
+      "step": 5996
+    },
+    {
+      "epoch": 0.81847959601474,
+      "grad_norm": 0.23276013135910034,
+      "learning_rate": 8e-05,
+      "loss": 1.4466,
+      "step": 5997
+    },
+    {
+      "epoch": 0.8186160775214958,
+      "grad_norm": 0.23943625390529633,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 5998
+    },
+    {
+      "epoch": 0.8187525590282517,
+      "grad_norm": 0.23996341228485107,
+      "learning_rate": 8e-05,
+      "loss": 1.4147,
+      "step": 5999
+    },
+    {
+      "epoch": 0.8188890405350076,
+      "grad_norm": 0.24511149525642395,
+      "learning_rate": 8e-05,
+      "loss": 1.3628,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7327,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.4159781870567424e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}