{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1115,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003590664272890485,
      "grad_norm": 2.40625,
      "learning_rate": 0.0,
      "loss": 1.1722,
      "step": 1
    },
    {
      "epoch": 0.00718132854578097,
      "grad_norm": 1.9453125,
      "learning_rate": 5.000000000000001e-07,
      "loss": 1.21,
      "step": 2
    },
    {
      "epoch": 0.010771992818671455,
      "grad_norm": 2.421875,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.2156,
      "step": 3
    },
    {
      "epoch": 0.01436265709156194,
      "grad_norm": 2.53125,
      "learning_rate": 1.5e-06,
      "loss": 1.1552,
      "step": 4
    },
    {
      "epoch": 0.017953321364452424,
      "grad_norm": 2.0625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.0573,
      "step": 5
    },
    {
      "epoch": 0.02154398563734291,
      "grad_norm": 2.234375,
      "learning_rate": 2.5e-06,
      "loss": 1.0933,
      "step": 6
    },
    {
      "epoch": 0.025134649910233394,
      "grad_norm": 2.078125,
      "learning_rate": 3e-06,
      "loss": 1.2411,
      "step": 7
    },
    {
      "epoch": 0.02872531418312388,
      "grad_norm": 1.8203125,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 1.4382,
      "step": 8
    },
    {
      "epoch": 0.03231597845601436,
      "grad_norm": 1.9296875,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.2927,
      "step": 9
    },
    {
      "epoch": 0.03590664272890485,
      "grad_norm": 1.96875,
      "learning_rate": 4.5e-06,
      "loss": 1.4581,
      "step": 10
    },
    {
      "epoch": 0.03949730700179533,
      "grad_norm": 1.4765625,
      "learning_rate": 5e-06,
      "loss": 1.0057,
      "step": 11
    },
    {
      "epoch": 0.04308797127468582,
      "grad_norm": 1.6796875,
      "learning_rate": 5.500000000000001e-06,
      "loss": 1.257,
      "step": 12
    },
    {
      "epoch": 0.0466786355475763,
      "grad_norm": 1.6640625,
      "learning_rate": 6e-06,
      "loss": 1.0787,
      "step": 13
    },
    {
      "epoch": 0.05026929982046679,
      "grad_norm": 1.6328125,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.3275,
      "step": 14
    },
    {
      "epoch": 0.05385996409335727,
      "grad_norm": 1.6171875,
      "learning_rate": 7.000000000000001e-06,
      "loss": 1.2611,
      "step": 15
    },
    {
      "epoch": 0.05745062836624776,
      "grad_norm": 1.2265625,
      "learning_rate": 7.5e-06,
      "loss": 1.1558,
      "step": 16
    },
    {
      "epoch": 0.06104129263913824,
      "grad_norm": 0.94140625,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.2986,
      "step": 17
    },
    {
      "epoch": 0.06463195691202872,
      "grad_norm": 0.98828125,
      "learning_rate": 8.500000000000002e-06,
      "loss": 1.0375,
      "step": 18
    },
    {
      "epoch": 0.06822262118491922,
      "grad_norm": 0.86328125,
      "learning_rate": 9e-06,
      "loss": 1.0183,
      "step": 19
    },
    {
      "epoch": 0.0718132854578097,
      "grad_norm": 0.75390625,
      "learning_rate": 9.5e-06,
      "loss": 1.0167,
      "step": 20
    },
    {
      "epoch": 0.07540394973070018,
      "grad_norm": 0.609375,
      "learning_rate": 1e-05,
      "loss": 0.888,
      "step": 21
    },
    {
      "epoch": 0.07899461400359066,
      "grad_norm": 0.5234375,
      "learning_rate": 1.05e-05,
      "loss": 1.0882,
      "step": 22
    },
    {
      "epoch": 0.08258527827648116,
      "grad_norm": 0.5078125,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 1.281,
      "step": 23
    },
    {
      "epoch": 0.08617594254937164,
      "grad_norm": 0.349609375,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.8019,
      "step": 24
    },
    {
      "epoch": 0.08976660682226212,
      "grad_norm": 0.357421875,
      "learning_rate": 1.2e-05,
      "loss": 1.0605,
      "step": 25
    },
    {
      "epoch": 0.0933572710951526,
      "grad_norm": 0.3359375,
      "learning_rate": 1.25e-05,
      "loss": 1.2793,
      "step": 26
    },
    {
      "epoch": 0.09694793536804308,
      "grad_norm": 0.302734375,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.9709,
      "step": 27
    },
    {
      "epoch": 0.10053859964093358,
      "grad_norm": 0.24609375,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.8537,
      "step": 28
    },
    {
      "epoch": 0.10412926391382406,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.0136,
      "step": 29
    },
    {
      "epoch": 0.10771992818671454,
      "grad_norm": 0.28515625,
      "learning_rate": 1.45e-05,
      "loss": 0.9281,
      "step": 30
    },
    {
      "epoch": 0.11131059245960502,
      "grad_norm": 0.27734375,
      "learning_rate": 1.5e-05,
      "loss": 0.9482,
      "step": 31
    },
    {
      "epoch": 0.11490125673249552,
      "grad_norm": 0.244140625,
      "learning_rate": 1.55e-05,
      "loss": 0.9554,
      "step": 32
    },
    {
      "epoch": 0.118491921005386,
      "grad_norm": 0.2216796875,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.1227,
      "step": 33
    },
    {
      "epoch": 0.12208258527827648,
      "grad_norm": 0.22265625,
      "learning_rate": 1.65e-05,
      "loss": 0.8488,
      "step": 34
    },
    {
      "epoch": 0.12567324955116696,
      "grad_norm": 0.2138671875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.8447,
      "step": 35
    },
    {
      "epoch": 0.12926391382405744,
      "grad_norm": 0.201171875,
      "learning_rate": 1.75e-05,
      "loss": 0.6357,
      "step": 36
    },
    {
      "epoch": 0.13285457809694792,
      "grad_norm": 0.216796875,
      "learning_rate": 1.8e-05,
      "loss": 0.8309,
      "step": 37
    },
    {
      "epoch": 0.13644524236983843,
      "grad_norm": 0.193359375,
      "learning_rate": 1.85e-05,
      "loss": 0.8134,
      "step": 38
    },
    {
      "epoch": 0.1400359066427289,
      "grad_norm": 0.1796875,
      "learning_rate": 1.9e-05,
      "loss": 0.9591,
      "step": 39
    },
    {
      "epoch": 0.1436265709156194,
      "grad_norm": 0.1728515625,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.7686,
      "step": 40
    },
    {
      "epoch": 0.14721723518850988,
      "grad_norm": 0.169921875,
      "learning_rate": 2e-05,
      "loss": 0.9276,
      "step": 41
    },
    {
      "epoch": 0.15080789946140036,
      "grad_norm": 0.1796875,
      "learning_rate": 2.05e-05,
      "loss": 0.9531,
      "step": 42
    },
    {
      "epoch": 0.15439856373429084,
      "grad_norm": 0.26953125,
      "learning_rate": 2.1e-05,
      "loss": 1.0058,
      "step": 43
    },
    {
      "epoch": 0.15798922800718132,
      "grad_norm": 0.17578125,
      "learning_rate": 2.15e-05,
      "loss": 0.7975,
      "step": 44
    },
    {
      "epoch": 0.1615798922800718,
      "grad_norm": 0.1962890625,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.0862,
      "step": 45
    },
    {
      "epoch": 0.1651705565529623,
      "grad_norm": 0.19140625,
      "learning_rate": 2.25e-05,
      "loss": 0.9248,
      "step": 46
    },
    {
      "epoch": 0.1687612208258528,
      "grad_norm": 0.1796875,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.7745,
      "step": 47
    },
    {
      "epoch": 0.17235188509874327,
      "grad_norm": 0.1376953125,
      "learning_rate": 2.35e-05,
      "loss": 0.6916,
      "step": 48
    },
    {
      "epoch": 0.17594254937163376,
      "grad_norm": 0.470703125,
      "learning_rate": 2.4e-05,
      "loss": 0.8171,
      "step": 49
    },
    {
      "epoch": 0.17953321364452424,
      "grad_norm": 0.1640625,
      "learning_rate": 2.45e-05,
      "loss": 0.7178,
      "step": 50
    },
    {
      "epoch": 0.18312387791741472,
      "grad_norm": 0.1708984375,
      "learning_rate": 2.5e-05,
      "loss": 0.9633,
      "step": 51
    },
    {
      "epoch": 0.1867145421903052,
      "grad_norm": 0.1591796875,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.8793,
      "step": 52
    },
    {
      "epoch": 0.19030520646319568,
      "grad_norm": 0.173828125,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.7277,
      "step": 53
    },
    {
      "epoch": 0.19389587073608616,
      "grad_norm": 0.1533203125,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.715,
      "step": 54
    },
    {
      "epoch": 0.19748653500897667,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.9866,
      "step": 55
    },
    {
      "epoch": 0.20107719928186715,
      "grad_norm": 0.16796875,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.8752,
      "step": 56
    },
    {
      "epoch": 0.20466786355475763,
      "grad_norm": 0.173828125,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.7317,
      "step": 57
    },
    {
      "epoch": 0.20825852782764812,
      "grad_norm": 0.1376953125,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.7779,
      "step": 58
    },
    {
      "epoch": 0.2118491921005386,
      "grad_norm": 0.13671875,
      "learning_rate": 2.9e-05,
      "loss": 0.7609,
      "step": 59
    },
    {
      "epoch": 0.21543985637342908,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.95e-05,
      "loss": 0.9057,
      "step": 60
    },
    {
      "epoch": 0.21903052064631956,
      "grad_norm": 0.177734375,
      "learning_rate": 3e-05,
      "loss": 0.7539,
      "step": 61
    },
    {
      "epoch": 0.22262118491921004,
      "grad_norm": 0.1806640625,
      "learning_rate": 3.05e-05,
      "loss": 0.7595,
      "step": 62
    },
    {
      "epoch": 0.22621184919210055,
      "grad_norm": 0.166015625,
      "learning_rate": 3.1e-05,
      "loss": 0.8889,
      "step": 63
    },
    {
      "epoch": 0.22980251346499103,
      "grad_norm": 0.146484375,
      "learning_rate": 3.15e-05,
      "loss": 0.7323,
      "step": 64
    },
    {
      "epoch": 0.2333931777378815,
      "grad_norm": 0.18359375,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.7669,
      "step": 65
    },
    {
      "epoch": 0.236983842010772,
      "grad_norm": 0.1748046875,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.7156,
      "step": 66
    },
    {
      "epoch": 0.24057450628366248,
      "grad_norm": 0.173828125,
      "learning_rate": 3.3e-05,
      "loss": 0.7047,
      "step": 67
    },
    {
      "epoch": 0.24416517055655296,
      "grad_norm": 0.171875,
      "learning_rate": 3.35e-05,
      "loss": 0.6125,
      "step": 68
    },
    {
      "epoch": 0.24775583482944344,
      "grad_norm": 0.189453125,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.6077,
      "step": 69
    },
    {
      "epoch": 0.2513464991023339,
      "grad_norm": 0.1865234375,
      "learning_rate": 3.45e-05,
      "loss": 0.8392,
      "step": 70
    },
    {
      "epoch": 0.25493716337522443,
      "grad_norm": 0.15234375,
      "learning_rate": 3.5e-05,
      "loss": 0.681,
      "step": 71
    },
    {
      "epoch": 0.2585278276481149,
      "grad_norm": 0.171875,
      "learning_rate": 3.55e-05,
      "loss": 0.6538,
      "step": 72
    },
    {
      "epoch": 0.2621184919210054,
      "grad_norm": 0.189453125,
      "learning_rate": 3.6e-05,
      "loss": 0.8431,
      "step": 73
    },
    {
      "epoch": 0.26570915619389585,
      "grad_norm": 0.1806640625,
      "learning_rate": 3.65e-05,
      "loss": 0.5274,
      "step": 74
    },
    {
      "epoch": 0.26929982046678635,
      "grad_norm": 0.21875,
      "learning_rate": 3.7e-05,
      "loss": 0.5791,
      "step": 75
    },
    {
      "epoch": 0.27289048473967686,
      "grad_norm": 0.1787109375,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4813,
      "step": 76
    },
    {
      "epoch": 0.2764811490125673,
      "grad_norm": 0.19140625,
      "learning_rate": 3.8e-05,
      "loss": 0.8457,
      "step": 77
    },
    {
      "epoch": 0.2800718132854578,
      "grad_norm": 0.2001953125,
      "learning_rate": 3.85e-05,
      "loss": 0.6608,
      "step": 78
    },
    {
      "epoch": 0.2836624775583483,
      "grad_norm": 0.1376953125,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.5428,
      "step": 79
    },
    {
      "epoch": 0.2872531418312388,
      "grad_norm": 0.1826171875,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.5525,
      "step": 80
    },
    {
      "epoch": 0.29084380610412924,
      "grad_norm": 0.1962890625,
      "learning_rate": 4e-05,
      "loss": 0.4874,
      "step": 81
    },
    {
      "epoch": 0.29443447037701975,
      "grad_norm": 0.25,
      "learning_rate": 4.05e-05,
      "loss": 0.7057,
      "step": 82
    },
    {
      "epoch": 0.2980251346499102,
      "grad_norm": 0.1787109375,
      "learning_rate": 4.1e-05,
      "loss": 0.4528,
      "step": 83
    },
    {
      "epoch": 0.3016157989228007,
      "grad_norm": 0.1884765625,
      "learning_rate": 4.15e-05,
      "loss": 0.8576,
      "step": 84
    },
    {
      "epoch": 0.3052064631956912,
      "grad_norm": 0.173828125,
      "learning_rate": 4.2e-05,
      "loss": 0.7524,
      "step": 85
    },
    {
      "epoch": 0.3087971274685817,
      "grad_norm": 0.1875,
      "learning_rate": 4.25e-05,
      "loss": 0.7139,
      "step": 86
    },
    {
      "epoch": 0.3123877917414722,
      "grad_norm": 0.1630859375,
      "learning_rate": 4.3e-05,
      "loss": 0.4864,
      "step": 87
    },
    {
      "epoch": 0.31597845601436264,
      "grad_norm": 0.1552734375,
      "learning_rate": 4.35e-05,
      "loss": 0.4428,
      "step": 88
    },
    {
      "epoch": 0.31956912028725315,
      "grad_norm": 0.13671875,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6394,
      "step": 89
    },
    {
      "epoch": 0.3231597845601436,
      "grad_norm": 0.1484375,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 0.4399,
      "step": 90
    },
    {
      "epoch": 0.3267504488330341,
      "grad_norm": 0.15625,
      "learning_rate": 4.5e-05,
      "loss": 0.5438,
      "step": 91
    },
    {
      "epoch": 0.3303411131059246,
      "grad_norm": 0.12158203125,
      "learning_rate": 4.55e-05,
      "loss": 0.4802,
      "step": 92
    },
    {
      "epoch": 0.3339317773788151,
      "grad_norm": 0.1650390625,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5739,
      "step": 93
    },
    {
      "epoch": 0.3375224416517056,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.4902,
      "step": 94
    },
    {
      "epoch": 0.34111310592459604,
      "grad_norm": 0.146484375,
      "learning_rate": 4.7e-05,
      "loss": 0.4182,
      "step": 95
    },
    {
      "epoch": 0.34470377019748655,
      "grad_norm": 0.1845703125,
      "learning_rate": 4.75e-05,
      "loss": 0.5809,
      "step": 96
    },
    {
      "epoch": 0.348294434470377,
      "grad_norm": 0.173828125,
      "learning_rate": 4.8e-05,
      "loss": 0.3826,
      "step": 97
    },
    {
      "epoch": 0.3518850987432675,
      "grad_norm": 0.166015625,
      "learning_rate": 4.85e-05,
      "loss": 0.3945,
      "step": 98
    },
    {
      "epoch": 0.35547576301615796,
      "grad_norm": 0.1748046875,
      "learning_rate": 4.9e-05,
      "loss": 0.3954,
      "step": 99
    },
    {
      "epoch": 0.3590664272890485,
      "grad_norm": 0.1533203125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.4171,
      "step": 100
    },
    {
      "epoch": 0.362657091561939,
      "grad_norm": 0.220703125,
      "learning_rate": 5e-05,
      "loss": 1.0399,
      "step": 101
    },
    {
      "epoch": 0.36624775583482944,
      "grad_norm": 0.17578125,
      "learning_rate": 4.9981412639405204e-05,
      "loss": 0.4628,
      "step": 102
    },
    {
      "epoch": 0.36983842010771995,
      "grad_norm": 0.16015625,
      "learning_rate": 4.996282527881041e-05,
      "loss": 0.4232,
      "step": 103
    },
    {
      "epoch": 0.3734290843806104,
      "grad_norm": 0.18359375,
      "learning_rate": 4.9944237918215614e-05,
      "loss": 0.6964,
      "step": 104
    },
    {
      "epoch": 0.3770197486535009,
      "grad_norm": 0.1953125,
      "learning_rate": 4.9925650557620816e-05,
      "loss": 0.6994,
      "step": 105
    },
    {
      "epoch": 0.38061041292639136,
      "grad_norm": 0.16796875,
      "learning_rate": 4.9907063197026024e-05,
      "loss": 0.4462,
      "step": 106
    },
    {
      "epoch": 0.38420107719928187,
      "grad_norm": 0.1748046875,
      "learning_rate": 4.988847583643123e-05,
      "loss": 0.3738,
      "step": 107
    },
    {
      "epoch": 0.3877917414721723,
      "grad_norm": 0.1669921875,
      "learning_rate": 4.9869888475836434e-05,
      "loss": 0.3799,
      "step": 108
    },
    {
      "epoch": 0.39138240574506283,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.9851301115241635e-05,
      "loss": 0.6639,
      "step": 109
    },
    {
      "epoch": 0.39497307001795334,
      "grad_norm": 0.169921875,
      "learning_rate": 4.9832713754646844e-05,
      "loss": 0.4669,
      "step": 110
    },
    {
      "epoch": 0.3985637342908438,
      "grad_norm": 0.1865234375,
      "learning_rate": 4.9814126394052045e-05,
      "loss": 0.4504,
      "step": 111
    },
    {
      "epoch": 0.4021543985637343,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.979553903345725e-05,
      "loss": 0.5984,
      "step": 112
    },
    {
      "epoch": 0.40574506283662476,
      "grad_norm": 0.21484375,
      "learning_rate": 4.9776951672862455e-05,
      "loss": 0.451,
      "step": 113
    },
    {
      "epoch": 0.40933572710951527,
      "grad_norm": 0.1630859375,
      "learning_rate": 4.975836431226766e-05,
      "loss": 0.347,
      "step": 114
    },
    {
      "epoch": 0.4129263913824057,
      "grad_norm": 0.1640625,
      "learning_rate": 4.9739776951672865e-05,
      "loss": 0.5839,
      "step": 115
    },
    {
      "epoch": 0.41651705565529623,
      "grad_norm": 0.16015625,
      "learning_rate": 4.9721189591078074e-05,
      "loss": 0.4278,
      "step": 116
    },
    {
      "epoch": 0.42010771992818674,
      "grad_norm": 0.205078125,
      "learning_rate": 4.9702602230483275e-05,
      "loss": 0.5096,
      "step": 117
    },
    {
      "epoch": 0.4236983842010772,
      "grad_norm": 0.1396484375,
      "learning_rate": 4.968401486988848e-05,
      "loss": 0.5401,
      "step": 118
    },
    {
      "epoch": 0.4272890484739677,
      "grad_norm": 0.19140625,
      "learning_rate": 4.9665427509293685e-05,
      "loss": 0.5079,
      "step": 119
    },
    {
      "epoch": 0.43087971274685816,
      "grad_norm": 0.177734375,
      "learning_rate": 4.964684014869889e-05,
      "loss": 0.3585,
      "step": 120
    },
    {
      "epoch": 0.43447037701974867,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.962825278810409e-05,
      "loss": 0.9457,
      "step": 121
    },
    {
      "epoch": 0.4380610412926391,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.96096654275093e-05,
      "loss": 0.6209,
      "step": 122
    },
    {
      "epoch": 0.44165170556552963,
      "grad_norm": 0.162109375,
      "learning_rate": 4.95910780669145e-05,
      "loss": 0.5945,
      "step": 123
    },
    {
      "epoch": 0.4452423698384201,
      "grad_norm": 0.169921875,
      "learning_rate": 4.957249070631971e-05,
      "loss": 0.3722,
      "step": 124
    },
    {
      "epoch": 0.4488330341113106,
      "grad_norm": 0.1640625,
      "learning_rate": 4.955390334572491e-05,
      "loss": 0.4399,
      "step": 125
    },
    {
      "epoch": 0.4524236983842011,
      "grad_norm": 0.19921875,
      "learning_rate": 4.953531598513012e-05,
      "loss": 0.6232,
      "step": 126
    },
    {
      "epoch": 0.45601436265709155,
      "grad_norm": 0.2470703125,
      "learning_rate": 4.951672862453532e-05,
      "loss": 0.6242,
      "step": 127
    },
    {
      "epoch": 0.45960502692998206,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.949814126394052e-05,
      "loss": 0.6271,
      "step": 128
    },
    {
      "epoch": 0.4631956912028725,
      "grad_norm": 0.1884765625,
      "learning_rate": 4.947955390334573e-05,
      "loss": 0.4935,
      "step": 129
    },
    {
      "epoch": 0.466786355475763,
      "grad_norm": 0.18359375,
      "learning_rate": 4.946096654275093e-05,
      "loss": 0.7235,
      "step": 130
    },
    {
      "epoch": 0.4703770197486535,
      "grad_norm": 0.193359375,
      "learning_rate": 4.944237918215613e-05,
      "loss": 0.3453,
      "step": 131
    },
    {
      "epoch": 0.473967684021544,
      "grad_norm": 0.169921875,
      "learning_rate": 4.942379182156134e-05,
      "loss": 0.5143,
      "step": 132
    },
    {
      "epoch": 0.47755834829443444,
      "grad_norm": 0.1748046875,
      "learning_rate": 4.940520446096655e-05,
      "loss": 0.7716,
      "step": 133
    },
    {
      "epoch": 0.48114901256732495,
      "grad_norm": 0.1953125,
      "learning_rate": 4.938661710037175e-05,
      "loss": 0.3434,
      "step": 134
    },
    {
      "epoch": 0.48473967684021546,
      "grad_norm": 0.162109375,
      "learning_rate": 4.936802973977696e-05,
      "loss": 0.3586,
      "step": 135
    },
    {
      "epoch": 0.4883303411131059,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.934944237918216e-05,
      "loss": 0.5625,
      "step": 136
    },
    {
      "epoch": 0.4919210053859964,
      "grad_norm": 0.19921875,
      "learning_rate": 4.933085501858736e-05,
      "loss": 0.4331,
      "step": 137
    },
    {
      "epoch": 0.4955116696588869,
      "grad_norm": 0.24609375,
      "learning_rate": 4.931226765799257e-05,
      "loss": 0.5312,
      "step": 138
    },
    {
      "epoch": 0.4991023339317774,
      "grad_norm": 0.17578125,
      "learning_rate": 4.929368029739777e-05,
      "loss": 0.5553,
      "step": 139
    },
    {
      "epoch": 0.5026929982046678,
      "grad_norm": 0.263671875,
      "learning_rate": 4.927509293680297e-05,
      "loss": 0.4315,
      "step": 140
    },
    {
      "epoch": 0.5062836624775583,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.925650557620818e-05,
      "loss": 0.6043,
      "step": 141
    },
    {
      "epoch": 0.5098743267504489,
      "grad_norm": 0.19140625,
      "learning_rate": 4.923791821561339e-05,
      "loss": 0.4513,
      "step": 142
    },
    {
      "epoch": 0.5134649910233393,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.921933085501859e-05,
      "loss": 0.3575,
      "step": 143
    },
    {
      "epoch": 0.5170556552962298,
      "grad_norm": 0.23828125,
      "learning_rate": 4.920074349442379e-05,
      "loss": 0.5501,
      "step": 144
    },
    {
      "epoch": 0.5206463195691203,
      "grad_norm": 0.1884765625,
      "learning_rate": 4.9182156133829e-05,
      "loss": 0.5022,
      "step": 145
    },
    {
      "epoch": 0.5242369838420108,
      "grad_norm": 0.197265625,
      "learning_rate": 4.91635687732342e-05,
      "loss": 0.7161,
      "step": 146
    },
    {
      "epoch": 0.5278276481149012,
      "grad_norm": 0.1494140625,
      "learning_rate": 4.9144981412639404e-05,
      "loss": 0.4487,
      "step": 147
    },
    {
      "epoch": 0.5314183123877917,
      "grad_norm": 0.18359375,
      "learning_rate": 4.912639405204461e-05,
      "loss": 0.5485,
      "step": 148
    },
    {
      "epoch": 0.5350089766606823,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.9107806691449814e-05,
      "loss": 0.5361,
      "step": 149
    },
    {
      "epoch": 0.5385996409335727,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.9089219330855016e-05,
      "loss": 0.441,
      "step": 150
    },
    {
      "epoch": 0.5421903052064632,
      "grad_norm": 0.1865234375,
      "learning_rate": 4.907063197026023e-05,
      "loss": 0.3284,
      "step": 151
    },
    {
      "epoch": 0.5457809694793537,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.905204460966543e-05,
      "loss": 0.5081,
      "step": 152
    },
    {
      "epoch": 0.5493716337522442,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.9033457249070634e-05,
      "loss": 0.6263,
      "step": 153
    },
    {
      "epoch": 0.5529622980251346,
      "grad_norm": 0.1552734375,
      "learning_rate": 4.901486988847584e-05,
      "loss": 0.4535,
      "step": 154
    },
    {
      "epoch": 0.5565529622980251,
      "grad_norm": 0.20703125,
      "learning_rate": 4.8996282527881044e-05,
      "loss": 0.3371,
      "step": 155
    },
    {
      "epoch": 0.5601436265709157,
      "grad_norm": 0.193359375,
      "learning_rate": 4.8977695167286246e-05,
      "loss": 0.5316,
      "step": 156
    },
    {
      "epoch": 0.5637342908438061,
      "grad_norm": 0.205078125,
      "learning_rate": 4.8959107806691454e-05,
      "loss": 0.2859,
      "step": 157
    },
    {
      "epoch": 0.5673249551166966,
      "grad_norm": 0.1611328125,
      "learning_rate": 4.8940520446096656e-05,
      "loss": 0.5021,
      "step": 158
    },
    {
      "epoch": 0.5709156193895871,
      "grad_norm": 0.203125,
      "learning_rate": 4.892193308550186e-05,
      "loss": 0.5018,
      "step": 159
    },
    {
      "epoch": 0.5745062836624776,
      "grad_norm": 0.212890625,
      "learning_rate": 4.8903345724907066e-05,
      "loss": 0.3316,
      "step": 160
    },
    {
      "epoch": 0.578096947935368,
      "grad_norm": 0.1884765625,
      "learning_rate": 4.8884758364312274e-05,
      "loss": 0.415,
      "step": 161
    },
    {
      "epoch": 0.5816876122082585,
      "grad_norm": 0.17578125,
      "learning_rate": 4.8866171003717476e-05,
      "loss": 0.5045,
      "step": 162
    },
    {
      "epoch": 0.585278276481149,
      "grad_norm": 0.205078125,
      "learning_rate": 4.884758364312268e-05,
      "loss": 0.6017,
      "step": 163
    },
    {
      "epoch": 0.5888689407540395,
      "grad_norm": 0.208984375,
      "learning_rate": 4.8828996282527886e-05,
      "loss": 0.3341,
      "step": 164
    },
    {
      "epoch": 0.59245960502693,
      "grad_norm": 0.18359375,
      "learning_rate": 4.881040892193309e-05,
      "loss": 0.4014,
      "step": 165
    },
    {
      "epoch": 0.5960502692998204,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.879182156133829e-05,
      "loss": 0.5091,
      "step": 166
    },
    {
      "epoch": 0.599640933572711,
      "grad_norm": 0.15625,
      "learning_rate": 4.87732342007435e-05,
      "loss": 0.5766,
      "step": 167
    },
    {
      "epoch": 0.6032315978456014,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.87546468401487e-05,
      "loss": 0.3147,
      "step": 168
    },
    {
      "epoch": 0.6068222621184919,
      "grad_norm": 0.25390625,
      "learning_rate": 4.873605947955391e-05,
      "loss": 0.5908,
      "step": 169
    },
    {
      "epoch": 0.6104129263913824,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.8717472118959115e-05,
      "loss": 0.4263,
      "step": 170
    },
    {
      "epoch": 0.6140035906642729,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.869888475836432e-05,
      "loss": 0.5972,
      "step": 171
    },
    {
      "epoch": 0.6175942549371634,
      "grad_norm": 0.1875,
      "learning_rate": 4.868029739776952e-05,
      "loss": 0.2917,
      "step": 172
    },
    {
      "epoch": 0.6211849192100538,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.866171003717473e-05,
      "loss": 0.3631,
      "step": 173
    },
    {
      "epoch": 0.6247755834829444,
      "grad_norm": 0.1787109375,
      "learning_rate": 4.864312267657993e-05,
      "loss": 0.3489,
      "step": 174
    },
    {
      "epoch": 0.6283662477558348,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.862453531598513e-05,
      "loss": 0.4056,
      "step": 175
    },
    {
      "epoch": 0.6319569120287253,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.860594795539034e-05,
      "loss": 0.3715,
      "step": 176
    },
    {
      "epoch": 0.6355475763016158,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.858736059479554e-05,
      "loss": 0.5893,
      "step": 177
    },
    {
      "epoch": 0.6391382405745063,
      "grad_norm": 0.189453125,
      "learning_rate": 4.856877323420075e-05,
      "loss": 0.3379,
      "step": 178
    },
    {
      "epoch": 0.6427289048473968,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.855018587360595e-05,
      "loss": 0.4707,
      "step": 179
    },
    {
      "epoch": 0.6463195691202872,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.853159851301116e-05,
      "loss": 0.2922,
      "step": 180
    },
    {
      "epoch": 0.6499102333931778,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.851301115241636e-05,
      "loss": 0.5792,
      "step": 181
    },
    {
      "epoch": 0.6535008976660682,
      "grad_norm": 0.2109375,
      "learning_rate": 4.849442379182156e-05,
      "loss": 0.5753,
      "step": 182
    },
    {
      "epoch": 0.6570915619389587,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.847583643122677e-05,
      "loss": 0.5272,
      "step": 183
    },
    {
      "epoch": 0.6606822262118492,
      "grad_norm": 0.197265625,
      "learning_rate": 4.845724907063197e-05,
      "loss": 0.6929,
      "step": 184
    },
    {
      "epoch": 0.6642728904847397,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.843866171003717e-05,
      "loss": 0.5956,
      "step": 185
    },
    {
      "epoch": 0.6678635547576302,
      "grad_norm": 0.201171875,
      "learning_rate": 4.842007434944238e-05,
      "loss": 0.3031,
      "step": 186
    },
    {
      "epoch": 0.6714542190305206,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.840148698884759e-05,
      "loss": 0.4583,
      "step": 187
    },
    {
      "epoch": 0.6750448833034112,
      "grad_norm": 0.1953125,
      "learning_rate": 4.838289962825279e-05,
      "loss": 0.6521,
      "step": 188
    },
    {
      "epoch": 0.6786355475763016,
      "grad_norm": 0.24609375,
      "learning_rate": 4.836431226765799e-05,
      "loss": 0.357,
      "step": 189
    },
    {
      "epoch": 0.6822262118491921,
      "grad_norm": 0.1806640625,
      "learning_rate": 4.83457249070632e-05,
      "loss": 0.3689,
      "step": 190
    },
    {
      "epoch": 0.6858168761220825,
      "grad_norm": 0.212890625,
      "learning_rate": 4.83271375464684e-05,
      "loss": 0.6121,
      "step": 191
    },
    {
      "epoch": 0.6894075403949731,
      "grad_norm": 0.1865234375,
      "learning_rate": 4.8308550185873605e-05,
      "loss": 0.728,
      "step": 192
    },
    {
      "epoch": 0.6929982046678635,
      "grad_norm": 0.2451171875,
      "learning_rate": 4.828996282527881e-05,
      "loss": 0.515,
      "step": 193
    },
    {
      "epoch": 0.696588868940754,
      "grad_norm": 0.1845703125,
      "learning_rate": 4.8271375464684015e-05,
      "loss": 0.5763,
      "step": 194
    },
    {
      "epoch": 0.7001795332136446,
      "grad_norm": 0.20703125,
      "learning_rate": 4.8252788104089216e-05,
      "loss": 0.341,
      "step": 195
    },
    {
      "epoch": 0.703770197486535,
      "grad_norm": 0.1640625,
      "learning_rate": 4.823420074349443e-05,
      "loss": 0.5045,
      "step": 196
    },
    {
      "epoch": 0.7073608617594255,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.821561338289963e-05,
      "loss": 0.642,
      "step": 197
    },
    {
      "epoch": 0.7109515260323159,
      "grad_norm": 0.2265625,
      "learning_rate": 4.8197026022304835e-05,
      "loss": 0.625,
      "step": 198
    },
    {
      "epoch": 0.7145421903052065,
      "grad_norm": 0.18359375,
      "learning_rate": 4.817843866171004e-05,
      "loss": 0.7704,
      "step": 199
    },
    {
      "epoch": 0.718132854578097,
      "grad_norm": 0.2578125,
      "learning_rate": 4.8159851301115244e-05,
      "loss": 0.4697,
      "step": 200
    },
    {
      "epoch": 0.7217235188509874,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.8141263940520446e-05,
      "loss": 0.5859,
      "step": 201
    },
    {
      "epoch": 0.725314183123878,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.8122676579925654e-05,
      "loss": 0.3702,
      "step": 202
    },
    {
      "epoch": 0.7289048473967684,
      "grad_norm": 0.1796875,
      "learning_rate": 4.8104089219330856e-05,
      "loss": 0.35,
      "step": 203
    },
    {
      "epoch": 0.7324955116696589,
      "grad_norm": 0.18359375,
      "learning_rate": 4.808550185873606e-05,
      "loss": 0.5314,
      "step": 204
    },
    {
      "epoch": 0.7360861759425493,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.8066914498141266e-05,
      "loss": 0.3365,
      "step": 205
    },
    {
      "epoch": 0.7396768402154399,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.8048327137546474e-05,
      "loss": 0.2958,
      "step": 206
    },
    {
      "epoch": 0.7432675044883303,
      "grad_norm": 0.2578125,
      "learning_rate": 4.8029739776951676e-05,
      "loss": 0.4026,
      "step": 207
    },
    {
      "epoch": 0.7468581687612208,
      "grad_norm": 0.205078125,
      "learning_rate": 4.801115241635688e-05,
      "loss": 0.3244,
      "step": 208
    },
    {
      "epoch": 0.7504488330341114,
      "grad_norm": 0.16015625,
      "learning_rate": 4.7992565055762086e-05,
      "loss": 0.4842,
      "step": 209
    },
    {
      "epoch": 0.7540394973070018,
      "grad_norm": 0.220703125,
      "learning_rate": 4.797397769516729e-05,
      "loss": 0.4208,
      "step": 210
    },
    {
      "epoch": 0.7576301615798923,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.795539033457249e-05,
      "loss": 0.3167,
      "step": 211
    },
    {
      "epoch": 0.7612208258527827,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.79368029739777e-05,
      "loss": 0.4286,
      "step": 212
    },
    {
      "epoch": 0.7648114901256733,
      "grad_norm": 0.19140625,
      "learning_rate": 4.79182156133829e-05,
      "loss": 0.3382,
      "step": 213
    },
    {
      "epoch": 0.7684021543985637,
      "grad_norm": 0.1611328125,
      "learning_rate": 4.789962825278811e-05,
      "loss": 0.5665,
      "step": 214
    },
    {
      "epoch": 0.7719928186714542,
      "grad_norm": 0.197265625,
      "learning_rate": 4.7881040892193316e-05,
      "loss": 0.3384,
      "step": 215
    },
    {
      "epoch": 0.7755834829443446,
      "grad_norm": 0.1953125,
      "learning_rate": 4.786245353159852e-05,
      "loss": 0.2773,
      "step": 216
    },
    {
      "epoch": 0.7791741472172352,
      "grad_norm": 0.208984375,
      "learning_rate": 4.784386617100372e-05,
      "loss": 0.2804,
      "step": 217
    },
    {
      "epoch": 0.7827648114901257,
      "grad_norm": 0.2109375,
      "learning_rate": 4.782527881040893e-05,
      "loss": 0.4151,
      "step": 218
    },
    {
      "epoch": 0.7863554757630161,
      "grad_norm": 0.2109375,
      "learning_rate": 4.780669144981413e-05,
      "loss": 0.5222,
      "step": 219
    },
    {
      "epoch": 0.7899461400359067,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.778810408921933e-05,
      "loss": 0.4517,
      "step": 220
    },
    {
      "epoch": 0.7935368043087971,
      "grad_norm": 0.21875,
      "learning_rate": 4.776951672862454e-05,
      "loss": 0.4731,
      "step": 221
    },
    {
      "epoch": 0.7971274685816876,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.775092936802974e-05,
      "loss": 0.2941,
      "step": 222
    },
    {
      "epoch": 0.800718132854578,
      "grad_norm": 0.224609375,
      "learning_rate": 4.773234200743495e-05,
      "loss": 0.6241,
      "step": 223
    },
    {
      "epoch": 0.8043087971274686,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.771375464684015e-05,
      "loss": 0.3415,
      "step": 224
    },
    {
      "epoch": 0.8078994614003591,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.769516728624536e-05,
      "loss": 0.3359,
      "step": 225
    },
    {
      "epoch": 0.8114901256732495,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.767657992565056e-05,
      "loss": 0.2953,
      "step": 226
    },
    {
      "epoch": 0.8150807899461401,
      "grad_norm": 0.1435546875,
      "learning_rate": 4.765799256505576e-05,
      "loss": 0.4333,
      "step": 227
    },
    {
      "epoch": 0.8186714542190305,
      "grad_norm": 0.212890625,
      "learning_rate": 4.763940520446097e-05,
      "loss": 0.3488,
      "step": 228
    },
    {
      "epoch": 0.822262118491921,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.762081784386617e-05,
      "loss": 0.4038,
      "step": 229
    },
    {
      "epoch": 0.8258527827648114,
      "grad_norm": 0.1630859375,
      "learning_rate": 4.7602230483271374e-05,
      "loss": 0.3945,
      "step": 230
    },
    {
      "epoch": 0.829443447037702,
      "grad_norm": 0.19921875,
      "learning_rate": 4.758364312267658e-05,
      "loss": 0.3497,
      "step": 231
    },
    {
      "epoch": 0.8330341113105925,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.756505576208179e-05,
      "loss": 0.5027,
      "step": 232
    },
    {
      "epoch": 0.8366247755834829,
      "grad_norm": 0.244140625,
      "learning_rate": 4.754646840148699e-05,
      "loss": 0.4217,
      "step": 233
    },
    {
      "epoch": 0.8402154398563735,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.75278810408922e-05,
      "loss": 0.5247,
      "step": 234
    },
    {
      "epoch": 0.8438061041292639,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.75092936802974e-05,
      "loss": 0.3392,
      "step": 235
    },
    {
      "epoch": 0.8473967684021544,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.7490706319702603e-05,
      "loss": 0.3338,
      "step": 236
    },
    {
      "epoch": 0.8509874326750448,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.747211895910781e-05,
      "loss": 0.5743,
      "step": 237
    },
    {
      "epoch": 0.8545780969479354,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.745353159851301e-05,
      "loss": 0.458,
      "step": 238
    },
    {
      "epoch": 0.8581687612208259,
      "grad_norm": 0.20703125,
      "learning_rate": 4.7434944237918215e-05,
      "loss": 0.3706,
      "step": 239
    },
    {
      "epoch": 0.8617594254937163,
      "grad_norm": 0.21875,
      "learning_rate": 4.741635687732342e-05,
      "loss": 0.2846,
      "step": 240
    },
    {
      "epoch": 0.8653500897666068,
      "grad_norm": 0.193359375,
      "learning_rate": 4.739776951672863e-05,
      "loss": 0.3266,
      "step": 241
    },
    {
      "epoch": 0.8689407540394973,
      "grad_norm": 0.1787109375,
      "learning_rate": 4.737918215613383e-05,
      "loss": 0.6369,
      "step": 242
    },
    {
      "epoch": 0.8725314183123878,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.7360594795539035e-05,
      "loss": 0.5717,
      "step": 243
    },
    {
      "epoch": 0.8761220825852782,
      "grad_norm": 0.205078125,
      "learning_rate": 4.734200743494424e-05,
      "loss": 0.3249,
      "step": 244
    },
    {
      "epoch": 0.8797127468581688,
      "grad_norm": 0.20703125,
      "learning_rate": 4.7323420074349445e-05,
      "loss": 0.4156,
      "step": 245
    },
    {
      "epoch": 0.8833034111310593,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.7304832713754646e-05,
      "loss": 0.8429,
      "step": 246
    },
    {
      "epoch": 0.8868940754039497,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.7286245353159855e-05,
      "loss": 0.3741,
      "step": 247
    },
    {
      "epoch": 0.8904847396768402,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.7267657992565056e-05,
      "loss": 0.5796,
      "step": 248
    },
    {
      "epoch": 0.8940754039497307,
      "grad_norm": 0.1875,
      "learning_rate": 4.724907063197026e-05,
      "loss": 0.3287,
      "step": 249
    },
    {
      "epoch": 0.8976660682226212,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.7230483271375466e-05,
      "loss": 0.3639,
      "step": 250
    },
    {
      "epoch": 0.9012567324955116,
      "grad_norm": 0.2109375,
      "learning_rate": 4.7211895910780675e-05,
      "loss": 0.6186,
      "step": 251
    },
    {
      "epoch": 0.9048473967684022,
      "grad_norm": 0.189453125,
      "learning_rate": 4.7193308550185876e-05,
      "loss": 0.3296,
      "step": 252
    },
    {
      "epoch": 0.9084380610412927,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.717472118959108e-05,
      "loss": 0.3391,
      "step": 253
    },
    {
      "epoch": 0.9120287253141831,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.7156133828996286e-05,
      "loss": 0.3714,
      "step": 254
    },
    {
      "epoch": 0.9156193895870736,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.713754646840149e-05,
      "loss": 0.4494,
      "step": 255
    },
    {
      "epoch": 0.9192100538599641,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.711895910780669e-05,
      "loss": 0.3782,
      "step": 256
    },
    {
      "epoch": 0.9228007181328546,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.71003717472119e-05,
      "loss": 0.3204,
      "step": 257
    },
    {
      "epoch": 0.926391382405745,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.70817843866171e-05,
      "loss": 0.2866,
      "step": 258
    },
    {
      "epoch": 0.9299820466786356,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.706319702602231e-05,
      "loss": 0.3192,
      "step": 259
    },
    {
      "epoch": 0.933572710951526,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.7044609665427516e-05,
      "loss": 0.3509,
      "step": 260
    },
    {
      "epoch": 0.9371633752244165,
      "grad_norm": 0.197265625,
      "learning_rate": 4.702602230483272e-05,
      "loss": 0.3819,
      "step": 261
    },
    {
      "epoch": 0.940754039497307,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.700743494423792e-05,
      "loss": 0.3506,
      "step": 262
    },
    {
      "epoch": 0.9443447037701975,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.698884758364313e-05,
      "loss": 0.4074,
      "step": 263
    },
    {
      "epoch": 0.947935368043088,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.697026022304833e-05,
      "loss": 0.3106,
      "step": 264
    },
    {
      "epoch": 0.9515260323159784,
      "grad_norm": 0.203125,
      "learning_rate": 4.695167286245353e-05,
      "loss": 0.4296,
      "step": 265
    },
    {
      "epoch": 0.9551166965888689,
      "grad_norm": 0.23046875,
      "learning_rate": 4.693308550185874e-05,
      "loss": 0.7067,
      "step": 266
    },
    {
      "epoch": 0.9587073608617595,
      "grad_norm": 0.189453125,
      "learning_rate": 4.691449814126394e-05,
      "loss": 0.3292,
      "step": 267
    },
    {
      "epoch": 0.9622980251346499,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.689591078066915e-05,
      "loss": 0.4203,
      "step": 268
    },
    {
      "epoch": 0.9658886894075404,
      "grad_norm": 0.287109375,
      "learning_rate": 4.687732342007435e-05,
      "loss": 0.6782,
      "step": 269
    },
    {
      "epoch": 0.9694793536804309,
      "grad_norm": 0.203125,
      "learning_rate": 4.685873605947956e-05,
      "loss": 0.414,
      "step": 270
    },
    {
      "epoch": 0.9730700179533214,
      "grad_norm": 0.248046875,
      "learning_rate": 4.684014869888476e-05,
      "loss": 0.7328,
      "step": 271
    },
    {
      "epoch": 0.9766606822262118,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.682156133828996e-05,
      "loss": 0.5914,
      "step": 272
    },
    {
      "epoch": 0.9802513464991023,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.680297397769517e-05,
      "loss": 0.3164,
      "step": 273
    },
    {
      "epoch": 0.9838420107719928,
      "grad_norm": 0.208984375,
      "learning_rate": 4.678438661710037e-05,
      "loss": 0.6842,
      "step": 274
    },
    {
      "epoch": 0.9874326750448833,
      "grad_norm": 0.232421875,
      "learning_rate": 4.6765799256505574e-05,
      "loss": 0.3047,
      "step": 275
    },
    {
      "epoch": 0.9910233393177738,
      "grad_norm": 0.1572265625,
      "learning_rate": 4.674721189591078e-05,
      "loss": 0.4245,
      "step": 276
    },
    {
      "epoch": 0.9946140035906643,
      "grad_norm": 0.2109375,
      "learning_rate": 4.6728624535315984e-05,
      "loss": 0.5554,
      "step": 277
    },
    {
      "epoch": 0.9982046678635548,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.671003717472119e-05,
      "loss": 0.2936,
      "step": 278
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.33203125,
      "learning_rate": 4.66914498141264e-05,
      "loss": 0.232,
      "step": 279
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.28634944558143616,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 3.9434,
      "eval_samples_per_second": 1.268,
      "eval_steps_per_second": 1.268,
      "step": 279
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.6696981191635132,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 3.0537,
      "eval_samples_per_second": 1.637,
      "eval_steps_per_second": 1.637,
      "step": 279
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.38031893968582153,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 4.5144,
      "eval_samples_per_second": 1.108,
      "eval_steps_per_second": 1.108,
      "step": 279
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.007208228111267,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 4.7521,
      "eval_samples_per_second": 1.052,
      "eval_steps_per_second": 1.052,
      "step": 279
    },
    {
      "epoch": 1.0035906642728905,
      "grad_norm": 0.154296875,
      "learning_rate": 4.66728624535316e-05,
      "loss": 0.3818,
      "step": 280
    },
    {
      "epoch": 1.007181328545781,
      "grad_norm": 0.2265625,
      "learning_rate": 4.6654275092936804e-05,
      "loss": 0.5046,
      "step": 281
    },
    {
      "epoch": 1.0107719928186714,
      "grad_norm": 0.23046875,
      "learning_rate": 4.663568773234201e-05,
      "loss": 0.3584,
      "step": 282
    },
    {
      "epoch": 1.014362657091562,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.6617100371747214e-05,
      "loss": 0.339,
      "step": 283
    },
    {
      "epoch": 1.0179533213644525,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.6598513011152415e-05,
      "loss": 0.3078,
      "step": 284
    },
    {
      "epoch": 1.021543985637343,
      "grad_norm": 0.208984375,
      "learning_rate": 4.6579925650557624e-05,
      "loss": 0.6365,
      "step": 285
    },
    {
      "epoch": 1.0251346499102334,
      "grad_norm": 0.18359375,
      "learning_rate": 4.6561338289962825e-05,
      "loss": 0.4359,
      "step": 286
    },
    {
      "epoch": 1.0287253141831239,
      "grad_norm": 0.23046875,
      "learning_rate": 4.6542750929368034e-05,
      "loss": 0.3607,
      "step": 287
    },
    {
      "epoch": 1.0323159784560143,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.6524163568773235e-05,
      "loss": 0.2898,
      "step": 288
    },
    {
      "epoch": 1.0359066427289048,
      "grad_norm": 0.166015625,
      "learning_rate": 4.6505576208178444e-05,
      "loss": 0.3892,
      "step": 289
    },
    {
      "epoch": 1.0394973070017954,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.6486988847583645e-05,
      "loss": 0.4383,
      "step": 290
    },
    {
      "epoch": 1.0430879712746859,
      "grad_norm": 0.205078125,
      "learning_rate": 4.646840148698885e-05,
      "loss": 0.4544,
      "step": 291
    },
    {
      "epoch": 1.0466786355475763,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.6449814126394055e-05,
      "loss": 0.3161,
      "step": 292
    },
    {
      "epoch": 1.0502692998204668,
      "grad_norm": 0.18359375,
      "learning_rate": 4.643122676579926e-05,
      "loss": 0.5709,
      "step": 293
    },
    {
      "epoch": 1.0538599640933572,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.641263940520446e-05,
      "loss": 0.831,
      "step": 294
    },
    {
      "epoch": 1.0574506283662477,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.639405204460967e-05,
      "loss": 0.3443,
      "step": 295
    },
    {
      "epoch": 1.0610412926391382,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.6375464684014875e-05,
      "loss": 0.3068,
      "step": 296
    },
    {
      "epoch": 1.0646319569120286,
      "grad_norm": 0.19140625,
      "learning_rate": 4.635687732342008e-05,
      "loss": 0.3391,
      "step": 297
    },
    {
      "epoch": 1.0682226211849193,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.6338289962825285e-05,
      "loss": 0.3352,
      "step": 298
    },
    {
      "epoch": 1.0718132854578097,
      "grad_norm": 0.212890625,
      "learning_rate": 4.6319702602230487e-05,
      "loss": 0.3385,
      "step": 299
    },
    {
      "epoch": 1.0754039497307002,
      "grad_norm": 0.2353515625,
      "learning_rate": 4.630111524163569e-05,
      "loss": 0.7581,
      "step": 300
    },
    {
      "epoch": 1.0789946140035906,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.6282527881040897e-05,
      "loss": 0.4061,
      "step": 301
    },
    {
      "epoch": 1.082585278276481,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.62639405204461e-05,
      "loss": 0.4838,
      "step": 302
    },
    {
      "epoch": 1.0861759425493716,
      "grad_norm": 0.1533203125,
      "learning_rate": 4.62453531598513e-05,
      "loss": 0.4344,
      "step": 303
    },
    {
      "epoch": 1.0897666068222622,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.622676579925651e-05,
      "loss": 0.3896,
      "step": 304
    },
    {
      "epoch": 1.0933572710951527,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.6208178438661716e-05,
      "loss": 0.5859,
      "step": 305
    },
    {
      "epoch": 1.0969479353680431,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.618959107806692e-05,
      "loss": 0.4996,
      "step": 306
    },
    {
      "epoch": 1.1005385996409336,
      "grad_norm": 0.224609375,
      "learning_rate": 4.617100371747212e-05,
      "loss": 0.52,
      "step": 307
    },
    {
      "epoch": 1.104129263913824,
      "grad_norm": 0.203125,
      "learning_rate": 4.615241635687733e-05,
      "loss": 0.4457,
      "step": 308
    },
    {
      "epoch": 1.1077199281867145,
      "grad_norm": 0.2109375,
      "learning_rate": 4.613382899628253e-05,
      "loss": 0.8472,
      "step": 309
    },
    {
      "epoch": 1.111310592459605,
      "grad_norm": 0.244140625,
      "learning_rate": 4.611524163568773e-05,
      "loss": 0.4308,
      "step": 310
    },
    {
      "epoch": 1.1149012567324954,
      "grad_norm": 0.208984375,
      "learning_rate": 4.609665427509294e-05,
      "loss": 0.3639,
      "step": 311
    },
    {
      "epoch": 1.118491921005386,
      "grad_norm": 0.19140625,
      "learning_rate": 4.607806691449814e-05,
      "loss": 0.2779,
      "step": 312
    },
    {
      "epoch": 1.1220825852782765,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.605947955390334e-05,
      "loss": 0.3102,
      "step": 313
    },
    {
      "epoch": 1.125673249551167,
      "grad_norm": 0.20703125,
      "learning_rate": 4.604089219330856e-05,
      "loss": 0.2545,
      "step": 314
    },
    {
      "epoch": 1.1292639138240574,
      "grad_norm": 0.197265625,
      "learning_rate": 4.602230483271376e-05,
      "loss": 0.4191,
      "step": 315
    },
    {
      "epoch": 1.132854578096948,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.600371747211896e-05,
      "loss": 0.2813,
      "step": 316
    },
    {
      "epoch": 1.1364452423698383,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.598513011152417e-05,
      "loss": 0.4369,
      "step": 317
    },
    {
      "epoch": 1.140035906642729,
      "grad_norm": 0.20703125,
      "learning_rate": 4.596654275092937e-05,
      "loss": 0.3236,
      "step": 318
    },
    {
      "epoch": 1.1436265709156195,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.594795539033457e-05,
      "loss": 0.2906,
      "step": 319
    },
    {
      "epoch": 1.14721723518851,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.592936802973978e-05,
      "loss": 0.369,
      "step": 320
    },
    {
      "epoch": 1.1508078994614004,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.591078066914498e-05,
      "loss": 0.4485,
      "step": 321
    },
    {
      "epoch": 1.1543985637342908,
      "grad_norm": 0.23046875,
      "learning_rate": 4.5892193308550184e-05,
      "loss": 0.6548,
      "step": 322
    },
    {
      "epoch": 1.1579892280071813,
      "grad_norm": 0.224609375,
      "learning_rate": 4.587360594795539e-05,
      "loss": 0.2867,
      "step": 323
    },
    {
      "epoch": 1.1615798922800717,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.58550185873606e-05,
      "loss": 0.3632,
      "step": 324
    },
    {
      "epoch": 1.1651705565529622,
      "grad_norm": 0.232421875,
      "learning_rate": 4.58364312267658e-05,
      "loss": 0.4043,
      "step": 325
    },
    {
      "epoch": 1.1687612208258529,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.5817843866171004e-05,
      "loss": 0.4212,
      "step": 326
    },
    {
      "epoch": 1.1723518850987433,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.579925650557621e-05,
      "loss": 0.3223,
      "step": 327
    },
    {
      "epoch": 1.1759425493716338,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.5780669144981414e-05,
      "loss": 0.5652,
      "step": 328
    },
    {
      "epoch": 1.1795332136445242,
      "grad_norm": 0.2109375,
      "learning_rate": 4.5762081784386616e-05,
      "loss": 0.3001,
      "step": 329
    },
    {
      "epoch": 1.1831238779174147,
      "grad_norm": 0.1953125,
      "learning_rate": 4.5743494423791824e-05,
      "loss": 0.3211,
      "step": 330
    },
    {
      "epoch": 1.1867145421903051,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.5724907063197026e-05,
      "loss": 0.3574,
      "step": 331
    },
    {
      "epoch": 1.1903052064631956,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.5706319702602234e-05,
      "loss": 0.6303,
      "step": 332
    },
    {
      "epoch": 1.1938958707360863,
      "grad_norm": 0.1875,
      "learning_rate": 4.5687732342007436e-05,
      "loss": 0.3337,
      "step": 333
    },
    {
      "epoch": 1.1974865350089767,
      "grad_norm": 0.203125,
      "learning_rate": 4.5669144981412644e-05,
      "loss": 0.2527,
      "step": 334
    },
    {
      "epoch": 1.2010771992818672,
      "grad_norm": 0.2470703125,
      "learning_rate": 4.5650557620817846e-05,
      "loss": 0.604,
      "step": 335
    },
    {
      "epoch": 1.2046678635547576,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.563197026022305e-05,
      "loss": 0.4975,
      "step": 336
    },
    {
      "epoch": 1.208258527827648,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.5613382899628255e-05,
      "loss": 0.3092,
      "step": 337
    },
    {
      "epoch": 1.2118491921005385,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.559479553903346e-05,
      "loss": 0.4621,
      "step": 338
    },
    {
      "epoch": 1.215439856373429,
      "grad_norm": 0.220703125,
      "learning_rate": 4.557620817843866e-05,
      "loss": 0.3229,
      "step": 339
    },
    {
      "epoch": 1.2190305206463194,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.555762081784387e-05,
      "loss": 0.4324,
      "step": 340
    },
    {
      "epoch": 1.2226211849192101,
      "grad_norm": 0.20703125,
      "learning_rate": 4.5539033457249075e-05,
      "loss": 0.2997,
      "step": 341
    },
    {
      "epoch": 1.2262118491921006,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.552044609665428e-05,
      "loss": 0.6602,
      "step": 342
    },
    {
      "epoch": 1.229802513464991,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.5501858736059485e-05,
      "loss": 0.4265,
      "step": 343
    },
    {
      "epoch": 1.2333931777378815,
      "grad_norm": 0.234375,
      "learning_rate": 4.548327137546469e-05,
      "loss": 0.3953,
      "step": 344
    },
    {
      "epoch": 1.236983842010772,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.546468401486989e-05,
      "loss": 0.4705,
      "step": 345
    },
    {
      "epoch": 1.2405745062836624,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.54460966542751e-05,
      "loss": 0.397,
      "step": 346
    },
    {
      "epoch": 1.244165170556553,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.54275092936803e-05,
      "loss": 0.3164,
      "step": 347
    },
    {
      "epoch": 1.2477558348294435,
      "grad_norm": 0.2373046875,
      "learning_rate": 4.54089219330855e-05,
      "loss": 0.4066,
      "step": 348
    },
    {
      "epoch": 1.251346499102334,
      "grad_norm": 0.1640625,
      "learning_rate": 4.539033457249071e-05,
      "loss": 0.4298,
      "step": 349
    },
    {
      "epoch": 1.2549371633752244,
      "grad_norm": 0.216796875,
      "learning_rate": 4.537174721189592e-05,
      "loss": 0.2826,
      "step": 350
    },
    {
      "epoch": 1.2585278276481149,
      "grad_norm": 0.212890625,
      "learning_rate": 4.535315985130112e-05,
      "loss": 0.504,
      "step": 351
    },
    {
      "epoch": 1.2621184919210053,
      "grad_norm": 0.244140625,
      "learning_rate": 4.533457249070632e-05,
      "loss": 0.5487,
      "step": 352
    },
    {
      "epoch": 1.2657091561938958,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.531598513011153e-05,
      "loss": 0.3046,
      "step": 353
    },
    {
      "epoch": 1.2692998204667862,
      "grad_norm": 0.173828125,
      "learning_rate": 4.529739776951673e-05,
      "loss": 0.4365,
      "step": 354
    },
    {
      "epoch": 1.272890484739677,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.527881040892193e-05,
      "loss": 0.3541,
      "step": 355
    },
    {
      "epoch": 1.2764811490125674,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.526022304832714e-05,
      "loss": 0.4532,
      "step": 356
    },
    {
      "epoch": 1.2800718132854578,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.524163568773234e-05,
      "loss": 0.331,
      "step": 357
    },
    {
      "epoch": 1.2836624775583483,
      "grad_norm": 0.228515625,
      "learning_rate": 4.522304832713754e-05,
      "loss": 0.5541,
      "step": 358
    },
    {
      "epoch": 1.2872531418312387,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.520446096654276e-05,
      "loss": 0.2724,
      "step": 359
    },
    {
      "epoch": 1.2908438061041292,
      "grad_norm": 0.2109375,
      "learning_rate": 4.518587360594796e-05,
      "loss": 0.3383,
      "step": 360
    },
    {
      "epoch": 1.2944344703770199,
      "grad_norm": 0.1484375,
      "learning_rate": 4.516728624535316e-05,
      "loss": 0.3726,
      "step": 361
    },
    {
      "epoch": 1.2980251346499103,
      "grad_norm": 0.22265625,
      "learning_rate": 4.514869888475837e-05,
      "loss": 0.4091,
      "step": 362
    },
    {
      "epoch": 1.3016157989228008,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.513011152416357e-05,
      "loss": 0.2856,
      "step": 363
    },
    {
      "epoch": 1.3052064631956912,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.511152416356877e-05,
      "loss": 0.3424,
      "step": 364
    },
    {
      "epoch": 1.3087971274685817,
      "grad_norm": 0.166015625,
      "learning_rate": 4.509293680297398e-05,
      "loss": 0.3896,
      "step": 365
    },
    {
      "epoch": 1.3123877917414721,
      "grad_norm": 0.240234375,
      "learning_rate": 4.507434944237918e-05,
      "loss": 0.2506,
      "step": 366
    },
    {
      "epoch": 1.3159784560143626,
      "grad_norm": 0.228515625,
      "learning_rate": 4.5055762081784385e-05,
      "loss": 0.5162,
      "step": 367
    },
    {
      "epoch": 1.319569120287253,
      "grad_norm": 0.21875,
      "learning_rate": 4.503717472118959e-05,
      "loss": 0.2821,
      "step": 368
    },
    {
      "epoch": 1.3231597845601435,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.50185873605948e-05,
      "loss": 0.2855,
      "step": 369
    },
    {
      "epoch": 1.3267504488330342,
      "grad_norm": 0.18359375,
      "learning_rate": 4.5e-05,
      "loss": 0.435,
      "step": 370
    },
    {
      "epoch": 1.3303411131059246,
      "grad_norm": 0.205078125,
      "learning_rate": 4.4981412639405204e-05,
      "loss": 0.4025,
      "step": 371
    },
    {
      "epoch": 1.333931777378815,
      "grad_norm": 0.3671875,
      "learning_rate": 4.496282527881041e-05,
      "loss": 0.2664,
      "step": 372
    },
    {
      "epoch": 1.3375224416517055,
      "grad_norm": 0.1748046875,
      "learning_rate": 4.4944237918215614e-05,
      "loss": 0.4863,
      "step": 373
    },
    {
      "epoch": 1.341113105924596,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.4925650557620816e-05,
      "loss": 0.3111,
      "step": 374
    },
    {
      "epoch": 1.3447037701974867,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.4907063197026024e-05,
      "loss": 0.584,
      "step": 375
    },
    {
      "epoch": 1.3482944344703771,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.4888475836431226e-05,
      "loss": 0.3512,
      "step": 376
    },
    {
      "epoch": 1.3518850987432676,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.4869888475836434e-05,
      "loss": 0.2592,
      "step": 377
    },
    {
      "epoch": 1.355475763016158,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.485130111524164e-05,
      "loss": 0.5288,
      "step": 378
    },
    {
      "epoch": 1.3590664272890485,
      "grad_norm": 0.244140625,
      "learning_rate": 4.4832713754646844e-05,
      "loss": 0.643,
      "step": 379
    },
    {
      "epoch": 1.362657091561939,
      "grad_norm": 0.2265625,
      "learning_rate": 4.4814126394052046e-05,
      "loss": 0.3794,
      "step": 380
    },
    {
      "epoch": 1.3662477558348294,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.4795539033457254e-05,
      "loss": 0.3572,
      "step": 381
    },
    {
      "epoch": 1.3698384201077198,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.4776951672862456e-05,
      "loss": 0.5296,
      "step": 382
    },
    {
      "epoch": 1.3734290843806103,
      "grad_norm": 0.228515625,
      "learning_rate": 4.475836431226766e-05,
      "loss": 0.5691,
      "step": 383
    },
    {
      "epoch": 1.377019748653501,
      "grad_norm": 0.19140625,
      "learning_rate": 4.4739776951672866e-05,
      "loss": 0.3676,
      "step": 384
    },
    {
      "epoch": 1.3806104129263914,
      "grad_norm": 0.197265625,
      "learning_rate": 4.472118959107807e-05,
      "loss": 0.3141,
      "step": 385
    },
    {
      "epoch": 1.3842010771992819,
      "grad_norm": 0.166015625,
      "learning_rate": 4.4702602230483276e-05,
      "loss": 0.3311,
      "step": 386
    },
    {
      "epoch": 1.3877917414721723,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.468401486988848e-05,
      "loss": 0.7626,
      "step": 387
    },
    {
      "epoch": 1.3913824057450628,
      "grad_norm": 0.203125,
      "learning_rate": 4.4665427509293686e-05,
      "loss": 0.2882,
      "step": 388
    },
    {
      "epoch": 1.3949730700179535,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.464684014869889e-05,
      "loss": 0.4822,
      "step": 389
    },
    {
      "epoch": 1.398563734290844,
      "grad_norm": 0.1796875,
      "learning_rate": 4.462825278810409e-05,
      "loss": 0.4623,
      "step": 390
    },
    {
      "epoch": 1.4021543985637344,
      "grad_norm": 0.2265625,
      "learning_rate": 4.46096654275093e-05,
      "loss": 0.4934,
      "step": 391
    },
    {
      "epoch": 1.4057450628366248,
      "grad_norm": 0.2109375,
      "learning_rate": 4.45910780669145e-05,
      "loss": 0.259,
      "step": 392
    },
    {
      "epoch": 1.4093357271095153,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.45724907063197e-05,
      "loss": 0.5305,
      "step": 393
    },
    {
      "epoch": 1.4129263913824057,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.455390334572491e-05,
      "loss": 0.7658,
      "step": 394
    },
    {
      "epoch": 1.4165170556552962,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.453531598513012e-05,
      "loss": 0.3,
      "step": 395
    },
    {
      "epoch": 1.4201077199281866,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.451672862453532e-05,
      "loss": 0.3021,
      "step": 396
    },
    {
      "epoch": 1.423698384201077,
      "grad_norm": 0.220703125,
      "learning_rate": 4.449814126394053e-05,
      "loss": 0.2684,
      "step": 397
    },
    {
      "epoch": 1.4272890484739678,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.447955390334573e-05,
      "loss": 0.3448,
      "step": 398
    },
    {
      "epoch": 1.4308797127468582,
      "grad_norm": 0.216796875,
      "learning_rate": 4.446096654275093e-05,
      "loss": 0.5892,
      "step": 399
    },
    {
      "epoch": 1.4344703770197487,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.444237918215614e-05,
      "loss": 0.3782,
      "step": 400
    },
    {
      "epoch": 1.4380610412926391,
      "grad_norm": 0.236328125,
      "learning_rate": 4.442379182156134e-05,
      "loss": 0.4197,
      "step": 401
    },
    {
      "epoch": 1.4416517055655296,
      "grad_norm": 0.201171875,
      "learning_rate": 4.440520446096654e-05,
      "loss": 0.2854,
      "step": 402
    },
    {
      "epoch": 1.44524236983842,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.4386617100371743e-05,
      "loss": 0.5199,
      "step": 403
    },
    {
      "epoch": 1.4488330341113107,
      "grad_norm": 0.220703125,
      "learning_rate": 4.436802973977696e-05,
      "loss": 0.3352,
      "step": 404
    },
    {
      "epoch": 1.4524236983842012,
      "grad_norm": 0.1640625,
      "learning_rate": 4.434944237918216e-05,
      "loss": 0.4998,
      "step": 405
    },
    {
      "epoch": 1.4560143626570916,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.433085501858736e-05,
      "loss": 0.2566,
      "step": 406
    },
    {
      "epoch": 1.459605026929982,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.431226765799257e-05,
      "loss": 0.517,
      "step": 407
    },
    {
      "epoch": 1.4631956912028725,
      "grad_norm": 0.22265625,
      "learning_rate": 4.429368029739777e-05,
      "loss": 0.5892,
      "step": 408
    },
    {
      "epoch": 1.466786355475763,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.427509293680297e-05,
      "loss": 0.5199,
      "step": 409
    },
    {
      "epoch": 1.4703770197486534,
      "grad_norm": 0.2109375,
      "learning_rate": 4.425650557620818e-05,
      "loss": 0.6657,
      "step": 410
    },
    {
      "epoch": 1.4739676840215439,
      "grad_norm": 0.20703125,
      "learning_rate": 4.423791821561338e-05,
      "loss": 0.5219,
      "step": 411
    },
    {
      "epoch": 1.4775583482944343,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.4219330855018585e-05,
      "loss": 0.3106,
      "step": 412
    },
    {
      "epoch": 1.481149012567325,
      "grad_norm": 0.224609375,
      "learning_rate": 4.420074349442379e-05,
      "loss": 0.248,
      "step": 413
    },
    {
      "epoch": 1.4847396768402155,
      "grad_norm": 0.26171875,
      "learning_rate": 4.4182156133829e-05,
      "loss": 0.656,
      "step": 414
    },
    {
      "epoch": 1.488330341113106,
      "grad_norm": 0.197265625,
      "learning_rate": 4.41635687732342e-05,
      "loss": 0.2623,
      "step": 415
    },
    {
      "epoch": 1.4919210053859964,
      "grad_norm": 0.19140625,
      "learning_rate": 4.4144981412639405e-05,
      "loss": 0.2468,
      "step": 416
    },
    {
      "epoch": 1.4955116696588868,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.412639405204461e-05,
      "loss": 0.3161,
      "step": 417
    },
    {
      "epoch": 1.4991023339317775,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.4107806691449815e-05,
      "loss": 0.5057,
      "step": 418
    },
    {
      "epoch": 1.502692998204668,
      "grad_norm": 0.228515625,
      "learning_rate": 4.4089219330855016e-05,
      "loss": 0.4957,
      "step": 419
    },
    {
      "epoch": 1.5062836624775584,
      "grad_norm": 0.19921875,
      "learning_rate": 4.4070631970260225e-05,
      "loss": 0.3055,
      "step": 420
    },
    {
      "epoch": 1.5098743267504489,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.4052044609665426e-05,
      "loss": 0.3634,
      "step": 421
    },
    {
      "epoch": 1.5134649910233393,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.4033457249070635e-05,
      "loss": 0.559,
      "step": 422
    },
    {
      "epoch": 1.5170556552962298,
      "grad_norm": 0.208984375,
      "learning_rate": 4.401486988847584e-05,
      "loss": 0.3386,
      "step": 423
    },
    {
      "epoch": 1.5206463195691202,
      "grad_norm": 0.224609375,
      "learning_rate": 4.3996282527881045e-05,
      "loss": 0.7261,
      "step": 424
    },
    {
      "epoch": 1.5242369838420107,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.3977695167286246e-05,
      "loss": 0.3919,
      "step": 425
    },
    {
      "epoch": 1.5278276481149011,
      "grad_norm": 0.19921875,
      "learning_rate": 4.3959107806691455e-05,
      "loss": 0.4133,
      "step": 426
    },
    {
      "epoch": 1.5314183123877916,
      "grad_norm": 0.21484375,
      "learning_rate": 4.3940520446096656e-05,
      "loss": 0.2488,
      "step": 427
    },
    {
      "epoch": 1.5350089766606823,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.392193308550186e-05,
      "loss": 0.2981,
      "step": 428
    },
    {
      "epoch": 1.5385996409335727,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.3903345724907066e-05,
      "loss": 0.4748,
      "step": 429
    },
    {
      "epoch": 1.5421903052064632,
      "grad_norm": 0.220703125,
      "learning_rate": 4.388475836431227e-05,
      "loss": 0.601,
      "step": 430
    },
    {
      "epoch": 1.5457809694793538,
      "grad_norm": 0.1953125,
      "learning_rate": 4.3866171003717476e-05,
      "loss": 0.2806,
      "step": 431
    },
    {
      "epoch": 1.5493716337522443,
      "grad_norm": 0.21875,
      "learning_rate": 4.384758364312268e-05,
      "loss": 0.3031,
      "step": 432
    },
    {
      "epoch": 1.5529622980251347,
      "grad_norm": 0.21875,
      "learning_rate": 4.3828996282527886e-05,
      "loss": 0.4431,
      "step": 433
    },
    {
      "epoch": 1.5565529622980252,
      "grad_norm": 0.21484375,
      "learning_rate": 4.381040892193309e-05,
      "loss": 0.3391,
      "step": 434
    },
    {
      "epoch": 1.5601436265709157,
      "grad_norm": 0.220703125,
      "learning_rate": 4.379182156133829e-05,
      "loss": 0.6222,
      "step": 435
    },
    {
      "epoch": 1.563734290843806,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.37732342007435e-05,
      "loss": 0.3848,
      "step": 436
    },
    {
      "epoch": 1.5673249551166966,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.37546468401487e-05,
      "loss": 0.5241,
      "step": 437
    },
    {
      "epoch": 1.570915619389587,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.37360594795539e-05,
      "loss": 0.316,
      "step": 438
    },
    {
      "epoch": 1.5745062836624775,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.371747211895911e-05,
      "loss": 0.2983,
      "step": 439
    },
    {
      "epoch": 1.578096947935368,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.369888475836432e-05,
      "loss": 0.3054,
      "step": 440
    },
    {
      "epoch": 1.5816876122082584,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.368029739776952e-05,
      "loss": 0.4121,
      "step": 441
    },
    {
      "epoch": 1.585278276481149,
      "grad_norm": 0.18359375,
      "learning_rate": 4.366171003717473e-05,
      "loss": 0.3914,
      "step": 442
    },
    {
      "epoch": 1.5888689407540395,
      "grad_norm": 0.216796875,
      "learning_rate": 4.364312267657993e-05,
      "loss": 0.2949,
      "step": 443
    },
    {
      "epoch": 1.59245960502693,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.362453531598513e-05,
      "loss": 0.4656,
      "step": 444
    },
    {
      "epoch": 1.5960502692998204,
      "grad_norm": 0.197265625,
      "learning_rate": 4.360594795539034e-05,
      "loss": 0.5817,
      "step": 445
    },
    {
      "epoch": 1.599640933572711,
      "grad_norm": 0.201171875,
      "learning_rate": 4.358736059479554e-05,
      "loss": 0.2628,
      "step": 446
    },
    {
      "epoch": 1.6032315978456015,
      "grad_norm": 0.19921875,
      "learning_rate": 4.356877323420074e-05,
      "loss": 0.4424,
      "step": 447
    },
    {
      "epoch": 1.606822262118492,
      "grad_norm": 0.205078125,
      "learning_rate": 4.355018587360595e-05,
      "loss": 0.5851,
      "step": 448
    },
    {
      "epoch": 1.6104129263913824,
      "grad_norm": 0.1650390625,
      "learning_rate": 4.353159851301116e-05,
      "loss": 0.3117,
      "step": 449
    },
    {
      "epoch": 1.614003590664273,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.351301115241636e-05,
      "loss": 0.4864,
      "step": 450
    },
    {
      "epoch": 1.6175942549371634,
      "grad_norm": 0.23046875,
      "learning_rate": 4.349442379182156e-05,
      "loss": 0.3515,
      "step": 451
    },
    {
      "epoch": 1.6211849192100538,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.347583643122677e-05,
      "loss": 0.3482,
      "step": 452
    },
    {
      "epoch": 1.6247755834829443,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.345724907063197e-05,
      "loss": 0.4294,
      "step": 453
    },
    {
      "epoch": 1.6283662477558347,
      "grad_norm": 0.232421875,
      "learning_rate": 4.3438661710037174e-05,
      "loss": 0.4227,
      "step": 454
    },
    {
      "epoch": 1.6319569120287252,
      "grad_norm": 0.216796875,
      "learning_rate": 4.342007434944238e-05,
      "loss": 0.3211,
      "step": 455
    },
    {
      "epoch": 1.6355475763016158,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.3401486988847584e-05,
      "loss": 0.3965,
      "step": 456
    },
    {
      "epoch": 1.6391382405745063,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.3382899628252785e-05,
      "loss": 0.2978,
      "step": 457
    },
    {
      "epoch": 1.6427289048473968,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.3364312267658e-05,
      "loss": 0.2893,
      "step": 458
    },
    {
      "epoch": 1.6463195691202872,
      "grad_norm": 0.220703125,
      "learning_rate": 4.33457249070632e-05,
      "loss": 0.7826,
      "step": 459
    },
    {
      "epoch": 1.6499102333931779,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.3327137546468404e-05,
      "loss": 0.412,
      "step": 460
    },
    {
      "epoch": 1.6535008976660683,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.330855018587361e-05,
      "loss": 0.2426,
      "step": 461
    },
    {
      "epoch": 1.6570915619389588,
      "grad_norm": 0.220703125,
      "learning_rate": 4.3289962825278813e-05,
      "loss": 0.2443,
      "step": 462
    },
    {
      "epoch": 1.6606822262118492,
      "grad_norm": 0.259765625,
      "learning_rate": 4.3271375464684015e-05,
      "loss": 0.3942,
      "step": 463
    },
    {
      "epoch": 1.6642728904847397,
      "grad_norm": 0.232421875,
      "learning_rate": 4.3252788104089223e-05,
      "loss": 0.5769,
      "step": 464
    },
    {
      "epoch": 1.6678635547576302,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.3234200743494425e-05,
      "loss": 0.6283,
      "step": 465
    },
    {
      "epoch": 1.6714542190305206,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.3215613382899627e-05,
      "loss": 0.5611,
      "step": 466
    },
    {
      "epoch": 1.675044883303411,
      "grad_norm": 0.189453125,
      "learning_rate": 4.3197026022304835e-05,
      "loss": 0.5029,
      "step": 467
    },
    {
      "epoch": 1.6786355475763015,
      "grad_norm": 0.189453125,
      "learning_rate": 4.317843866171004e-05,
      "loss": 0.4081,
      "step": 468
    },
    {
      "epoch": 1.682226211849192,
      "grad_norm": 0.1591796875,
      "learning_rate": 4.3159851301115245e-05,
      "loss": 0.3425,
      "step": 469
    },
    {
      "epoch": 1.6858168761220824,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.3141263940520447e-05,
      "loss": 0.408,
      "step": 470
    },
    {
      "epoch": 1.689407540394973,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.3122676579925655e-05,
      "loss": 0.2425,
      "step": 471
    },
    {
      "epoch": 1.6929982046678635,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.3104089219330856e-05,
      "loss": 0.5171,
      "step": 472
    },
    {
      "epoch": 1.696588868940754,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.308550185873606e-05,
      "loss": 0.4631,
      "step": 473
    },
    {
      "epoch": 1.7001795332136447,
      "grad_norm": 0.19921875,
      "learning_rate": 4.3066914498141266e-05,
      "loss": 0.581,
      "step": 474
    },
    {
      "epoch": 1.7037701974865351,
      "grad_norm": 0.21875,
      "learning_rate": 4.304832713754647e-05,
      "loss": 0.281,
      "step": 475
    },
    {
      "epoch": 1.7073608617594256,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.3029739776951676e-05,
      "loss": 0.4728,
      "step": 476
    },
    {
      "epoch": 1.710951526032316,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.301115241635688e-05,
      "loss": 0.2792,
      "step": 477
    },
    {
      "epoch": 1.7145421903052065,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.2992565055762086e-05,
      "loss": 0.2616,
      "step": 478
    },
    {
      "epoch": 1.718132854578097,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.297397769516729e-05,
      "loss": 0.328,
      "step": 479
    },
    {
      "epoch": 1.7217235188509874,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.295539033457249e-05,
      "loss": 0.2634,
      "step": 480
    },
    {
      "epoch": 1.7253141831238779,
      "grad_norm": 0.265625,
      "learning_rate": 4.29368029739777e-05,
      "loss": 0.4253,
      "step": 481
    },
    {
      "epoch": 1.7289048473967683,
      "grad_norm": 0.310546875,
      "learning_rate": 4.29182156133829e-05,
      "loss": 0.384,
      "step": 482
    },
    {
      "epoch": 1.7324955116696588,
      "grad_norm": 0.201171875,
      "learning_rate": 4.28996282527881e-05,
      "loss": 0.4535,
      "step": 483
    },
    {
      "epoch": 1.7360861759425492,
      "grad_norm": 0.16796875,
      "learning_rate": 4.288104089219331e-05,
      "loss": 0.3941,
      "step": 484
    },
    {
      "epoch": 1.73967684021544,
      "grad_norm": 0.146484375,
      "learning_rate": 4.286245353159852e-05,
      "loss": 0.4452,
      "step": 485
    },
    {
      "epoch": 1.7432675044883303,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.284386617100372e-05,
      "loss": 0.2863,
      "step": 486
    },
    {
      "epoch": 1.7468581687612208,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.282527881040893e-05,
      "loss": 0.4329,
      "step": 487
    },
    {
      "epoch": 1.7504488330341115,
      "grad_norm": 0.212890625,
      "learning_rate": 4.280669144981413e-05,
      "loss": 0.465,
      "step": 488
    },
    {
      "epoch": 1.754039497307002,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.278810408921933e-05,
      "loss": 0.3176,
      "step": 489
    },
    {
      "epoch": 1.7576301615798924,
      "grad_norm": 0.2314453125,
      "learning_rate": 4.276951672862454e-05,
      "loss": 0.4032,
      "step": 490
    },
    {
      "epoch": 1.7612208258527828,
      "grad_norm": 0.240234375,
      "learning_rate": 4.275092936802974e-05,
      "loss": 0.3661,
      "step": 491
    },
    {
      "epoch": 1.7648114901256733,
      "grad_norm": 0.21875,
      "learning_rate": 4.273234200743494e-05,
      "loss": 0.3006,
      "step": 492
    },
    {
      "epoch": 1.7684021543985637,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.271375464684015e-05,
      "loss": 0.3067,
      "step": 493
    },
    {
      "epoch": 1.7719928186714542,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.269516728624536e-05,
      "loss": 0.3193,
      "step": 494
    },
    {
      "epoch": 1.7755834829443446,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.267657992565056e-05,
      "loss": 0.2902,
      "step": 495
    },
    {
      "epoch": 1.779174147217235,
      "grad_norm": 0.25,
      "learning_rate": 4.265799256505576e-05,
      "loss": 0.4452,
      "step": 496
    },
    {
      "epoch": 1.7827648114901256,
      "grad_norm": 0.203125,
      "learning_rate": 4.263940520446097e-05,
      "loss": 0.2502,
      "step": 497
    },
    {
      "epoch": 1.786355475763016,
      "grad_norm": 0.259765625,
      "learning_rate": 4.262081784386617e-05,
      "loss": 0.6095,
      "step": 498
    },
    {
      "epoch": 1.7899461400359067,
      "grad_norm": 0.224609375,
      "learning_rate": 4.2602230483271374e-05,
      "loss": 0.3303,
      "step": 499
    },
    {
      "epoch": 1.7935368043087971,
      "grad_norm": 0.240234375,
      "learning_rate": 4.258364312267658e-05,
      "loss": 0.5164,
      "step": 500
    },
    {
      "epoch": 1.7971274685816876,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.2565055762081784e-05,
      "loss": 0.4519,
      "step": 501
    },
    {
      "epoch": 1.800718132854578,
      "grad_norm": 0.22265625,
      "learning_rate": 4.2546468401486986e-05,
      "loss": 0.2106,
      "step": 502
    },
    {
      "epoch": 1.8043087971274687,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.25278810408922e-05,
      "loss": 0.3426,
      "step": 503
    },
    {
      "epoch": 1.8078994614003592,
      "grad_norm": 0.26171875,
      "learning_rate": 4.25092936802974e-05,
      "loss": 0.357,
      "step": 504
    },
    {
      "epoch": 1.8114901256732496,
      "grad_norm": 0.208984375,
      "learning_rate": 4.2490706319702604e-05,
      "loss": 0.5211,
      "step": 505
    },
    {
      "epoch": 1.81508078994614,
      "grad_norm": 0.2421875,
      "learning_rate": 4.247211895910781e-05,
      "loss": 0.4362,
      "step": 506
    },
    {
      "epoch": 1.8186714542190305,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.2453531598513014e-05,
      "loss": 0.2616,
      "step": 507
    },
    {
      "epoch": 1.822262118491921,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.2434944237918215e-05,
      "loss": 0.4464,
      "step": 508
    },
    {
      "epoch": 1.8258527827648114,
      "grad_norm": 0.216796875,
      "learning_rate": 4.2416356877323424e-05,
      "loss": 0.3297,
      "step": 509
    },
    {
      "epoch": 1.829443447037702,
      "grad_norm": 0.228515625,
      "learning_rate": 4.2397769516728625e-05,
      "loss": 0.8014,
      "step": 510
    },
    {
      "epoch": 1.8330341113105924,
      "grad_norm": 0.220703125,
      "learning_rate": 4.237918215613383e-05,
      "loss": 0.4611,
      "step": 511
    },
    {
      "epoch": 1.8366247755834828,
      "grad_norm": 0.166015625,
      "learning_rate": 4.2360594795539035e-05,
      "loss": 0.3836,
      "step": 512
    },
    {
      "epoch": 1.8402154398563735,
      "grad_norm": 0.138671875,
      "learning_rate": 4.2342007434944244e-05,
      "loss": 0.4204,
      "step": 513
    },
    {
      "epoch": 1.843806104129264,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.2323420074349445e-05,
      "loss": 0.2773,
      "step": 514
    },
    {
      "epoch": 1.8473967684021544,
      "grad_norm": 0.21875,
      "learning_rate": 4.230483271375465e-05,
      "loss": 0.4238,
      "step": 515
    },
    {
      "epoch": 1.8509874326750448,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.2286245353159855e-05,
      "loss": 0.3391,
      "step": 516
    },
    {
      "epoch": 1.8545780969479355,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.226765799256506e-05,
      "loss": 0.3631,
      "step": 517
    },
    {
      "epoch": 1.858168761220826,
      "grad_norm": 0.1953125,
      "learning_rate": 4.224907063197026e-05,
      "loss": 0.2885,
      "step": 518
    },
    {
      "epoch": 1.8617594254937164,
      "grad_norm": 0.228515625,
      "learning_rate": 4.223048327137547e-05,
      "loss": 0.3836,
      "step": 519
    },
    {
      "epoch": 1.8653500897666069,
      "grad_norm": 0.201171875,
      "learning_rate": 4.221189591078067e-05,
      "loss": 0.2848,
      "step": 520
    },
    {
      "epoch": 1.8689407540394973,
      "grad_norm": 0.22265625,
      "learning_rate": 4.219330855018588e-05,
      "loss": 0.4967,
      "step": 521
    },
    {
      "epoch": 1.8725314183123878,
      "grad_norm": 0.173828125,
      "learning_rate": 4.2174721189591085e-05,
      "loss": 0.3995,
      "step": 522
    },
    {
      "epoch": 1.8761220825852782,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.215613382899629e-05,
      "loss": 0.2886,
      "step": 523
    },
    {
      "epoch": 1.8797127468581687,
      "grad_norm": 0.2001953125,
      "learning_rate": 4.213754646840149e-05,
      "loss": 0.5412,
      "step": 524
    },
    {
      "epoch": 1.8833034111310591,
      "grad_norm": 0.2236328125,
      "learning_rate": 4.21189591078067e-05,
      "loss": 0.2263,
      "step": 525
    },
    {
      "epoch": 1.8868940754039496,
      "grad_norm": 0.2197265625,
      "learning_rate": 4.21003717472119e-05,
      "loss": 0.4597,
      "step": 526
    },
    {
      "epoch": 1.89048473967684,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.20817843866171e-05,
      "loss": 0.2892,
      "step": 527
    },
    {
      "epoch": 1.8940754039497307,
      "grad_norm": 0.26171875,
      "learning_rate": 4.206319702602231e-05,
      "loss": 0.4892,
      "step": 528
    },
    {
      "epoch": 1.8976660682226212,
      "grad_norm": 0.20703125,
      "learning_rate": 4.204460966542751e-05,
      "loss": 0.3135,
      "step": 529
    },
    {
      "epoch": 1.9012567324955116,
      "grad_norm": 0.23046875,
      "learning_rate": 4.202602230483272e-05,
      "loss": 0.3643,
      "step": 530
    },
    {
      "epoch": 1.9048473967684023,
      "grad_norm": 0.22265625,
      "learning_rate": 4.200743494423792e-05,
      "loss": 0.2984,
      "step": 531
    },
    {
      "epoch": 1.9084380610412928,
      "grad_norm": 0.228515625,
      "learning_rate": 4.198884758364313e-05,
      "loss": 0.3109,
      "step": 532
    },
    {
      "epoch": 1.9120287253141832,
      "grad_norm": 0.2314453125,
      "learning_rate": 4.197026022304833e-05,
      "loss": 0.4211,
      "step": 533
    },
    {
      "epoch": 1.9156193895870737,
      "grad_norm": 0.2265625,
      "learning_rate": 4.195167286245353e-05,
      "loss": 0.4605,
      "step": 534
    },
    {
      "epoch": 1.9192100538599641,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.193308550185874e-05,
      "loss": 0.5381,
      "step": 535
    },
    {
      "epoch": 1.9228007181328546,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.191449814126394e-05,
      "loss": 0.3614,
      "step": 536
    },
    {
      "epoch": 1.926391382405745,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.189591078066914e-05,
      "loss": 0.555,
      "step": 537
    },
    {
      "epoch": 1.9299820466786355,
      "grad_norm": 0.25,
      "learning_rate": 4.187732342007435e-05,
      "loss": 0.4886,
      "step": 538
    },
    {
      "epoch": 1.933572710951526,
      "grad_norm": 0.201171875,
      "learning_rate": 4.185873605947956e-05,
      "loss": 0.2972,
      "step": 539
    },
    {
      "epoch": 1.9371633752244164,
      "grad_norm": 0.224609375,
      "learning_rate": 4.184014869888476e-05,
      "loss": 0.3546,
      "step": 540
    },
    {
      "epoch": 1.9407540394973068,
      "grad_norm": 0.150390625,
      "learning_rate": 4.182156133828997e-05,
      "loss": 0.4356,
      "step": 541
    },
    {
      "epoch": 1.9443447037701975,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.180297397769517e-05,
      "loss": 0.4029,
      "step": 542
    },
    {
      "epoch": 1.947935368043088,
      "grad_norm": 0.265625,
      "learning_rate": 4.178438661710037e-05,
      "loss": 0.3333,
      "step": 543
    },
    {
      "epoch": 1.9515260323159784,
      "grad_norm": 0.2353515625,
      "learning_rate": 4.176579925650558e-05,
      "loss": 0.5598,
      "step": 544
    },
    {
      "epoch": 1.9551166965888689,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.174721189591078e-05,
      "loss": 0.2177,
      "step": 545
    },
    {
      "epoch": 1.9587073608617596,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.1728624535315984e-05,
      "loss": 0.6807,
      "step": 546
    },
    {
      "epoch": 1.96229802513465,
      "grad_norm": 0.146484375,
      "learning_rate": 4.171003717472119e-05,
      "loss": 0.4046,
      "step": 547
    },
    {
      "epoch": 1.9658886894075405,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.16914498141264e-05,
      "loss": 0.5034,
      "step": 548
    },
    {
      "epoch": 1.969479353680431,
      "grad_norm": 0.20703125,
      "learning_rate": 4.16728624535316e-05,
      "loss": 0.2981,
      "step": 549
    },
    {
      "epoch": 1.9730700179533214,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.1654275092936804e-05,
      "loss": 0.3522,
      "step": 550
    },
    {
      "epoch": 1.9766606822262118,
      "grad_norm": 0.24609375,
      "learning_rate": 4.163568773234201e-05,
      "loss": 0.2776,
      "step": 551
    },
    {
      "epoch": 1.9802513464991023,
      "grad_norm": 0.224609375,
      "learning_rate": 4.1617100371747214e-05,
      "loss": 0.4453,
      "step": 552
    },
    {
      "epoch": 1.9838420107719927,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.1598513011152416e-05,
      "loss": 0.5823,
      "step": 553
    },
    {
      "epoch": 1.9874326750448832,
      "grad_norm": 0.236328125,
      "learning_rate": 4.1579925650557624e-05,
      "loss": 0.3883,
      "step": 554
    },
    {
      "epoch": 1.9910233393177736,
      "grad_norm": 0.298828125,
      "learning_rate": 4.1561338289962826e-05,
      "loss": 0.545,
      "step": 555
    },
    {
      "epoch": 1.9946140035906643,
      "grad_norm": 0.2119140625,
      "learning_rate": 4.154275092936803e-05,
      "loss": 0.5936,
      "step": 556
    },
    {
      "epoch": 1.9982046678635548,
      "grad_norm": 0.21875,
      "learning_rate": 4.1524163568773236e-05,
      "loss": 0.3757,
      "step": 557
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.390625,
      "learning_rate": 4.1505576208178444e-05,
      "loss": 0.2596,
      "step": 558
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2662683129310608,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 3.9842,
      "eval_samples_per_second": 1.255,
      "eval_steps_per_second": 1.255,
      "step": 558
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6189020872116089,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 3.0561,
      "eval_samples_per_second": 1.636,
      "eval_steps_per_second": 1.636,
      "step": 558
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.35473424196243286,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 4.5153,
      "eval_samples_per_second": 1.107,
      "eval_steps_per_second": 1.107,
      "step": 558
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.9680498242378235,
      "eval_model_preparation_time": 0.0083,
      "eval_runtime": 4.7566,
      "eval_samples_per_second": 1.051,
      "eval_steps_per_second": 1.051,
      "step": 558
    },
    {
      "epoch": 3.998204667863555,
      "grad_norm": 0.736983597278595,
      "learning_rate": 4.1486988847583646e-05,
      "loss": 0.45,
      "step": 559
    },
    {
      "epoch": 3.998204667863555,
      "eval_loss": 0.32840269804000854,
      "eval_runtime": 5.0221,
      "eval_samples_per_second": 0.996,
      "eval_steps_per_second": 0.996,
      "step": 559
    },
    {
      "epoch": 3.998204667863555,
      "eval_loss": 0.6879445314407349,
      "eval_runtime": 3.2056,
      "eval_samples_per_second": 1.56,
      "eval_steps_per_second": 1.56,
      "step": 559
    },
    {
      "epoch": 3.998204667863555,
      "eval_loss": 0.40759116411209106,
      "eval_runtime": 4.7013,
      "eval_samples_per_second": 1.064,
      "eval_steps_per_second": 1.064,
      "step": 559
    },
    {
      "epoch": 3.998204667863555,
      "eval_loss": 1.0050503015518188,
      "eval_runtime": 4.9378,
      "eval_samples_per_second": 1.013,
      "eval_steps_per_second": 1.013,
      "step": 559
    },
    {
      "epoch": 2.007181328545781,
      "grad_norm": 0.29202306270599365,
      "learning_rate": 3.2346153846153846e-05,
      "loss": 0.499,
      "step": 560
    },
    {
      "epoch": 2.0107719928186714,
      "grad_norm": 0.3116172254085541,
      "learning_rate": 4.1449814126394056e-05,
      "loss": 0.3858,
      "step": 561
    },
    {
      "epoch": 2.014362657091562,
      "grad_norm": 0.4639987349510193,
      "learning_rate": 4.143122676579926e-05,
      "loss": 0.2768,
      "step": 562
    },
    {
      "epoch": 2.0179533213644523,
      "grad_norm": 0.24592472612857819,
      "learning_rate": 4.141263940520446e-05,
      "loss": 0.2924,
      "step": 563
    },
    {
      "epoch": 2.0215439856373427,
      "grad_norm": 0.24846947193145752,
      "learning_rate": 4.139405204460967e-05,
      "loss": 0.2778,
      "step": 564
    },
    {
      "epoch": 2.025134649910233,
      "grad_norm": 0.24283242225646973,
      "learning_rate": 4.137546468401487e-05,
      "loss": 0.4572,
      "step": 565
    },
    {
      "epoch": 2.028725314183124,
      "grad_norm": 0.2674539089202881,
      "learning_rate": 4.135687732342008e-05,
      "loss": 0.6989,
      "step": 566
    },
    {
      "epoch": 2.0323159784560145,
      "grad_norm": 0.2617143988609314,
      "learning_rate": 4.1338289962825285e-05,
      "loss": 0.4977,
      "step": 567
    },
    {
      "epoch": 2.035906642728905,
      "grad_norm": 0.283547967672348,
      "learning_rate": 4.131970260223049e-05,
      "loss": 0.5724,
      "step": 568
    },
    {
      "epoch": 2.0394973070017954,
      "grad_norm": 0.19893530011177063,
      "learning_rate": 4.130111524163569e-05,
      "loss": 0.3121,
      "step": 569
    },
    {
      "epoch": 2.043087971274686,
      "grad_norm": 0.2420407235622406,
      "learning_rate": 4.12825278810409e-05,
      "loss": 0.4878,
      "step": 570
    },
    {
      "epoch": 2.0466786355475763,
      "grad_norm": 0.21978935599327087,
      "learning_rate": 4.12639405204461e-05,
      "loss": 0.2606,
      "step": 571
    },
    {
      "epoch": 2.050269299820467,
      "grad_norm": 0.2838546335697174,
      "learning_rate": 4.12453531598513e-05,
      "loss": 0.486,
      "step": 572
    },
    {
      "epoch": 2.0538599640933572,
      "grad_norm": 0.252580851316452,
      "learning_rate": 4.122676579925651e-05,
      "loss": 0.3731,
      "step": 573
    },
    {
      "epoch": 2.0574506283662477,
      "grad_norm": 0.20983606576919556,
      "learning_rate": 4.120817843866171e-05,
      "loss": 0.3743,
      "step": 574
    },
    {
      "epoch": 2.061041292639138,
      "grad_norm": 0.24320125579833984,
      "learning_rate": 4.118959107806692e-05,
      "loss": 0.5766,
      "step": 575
    },
    {
      "epoch": 2.0646319569120286,
      "grad_norm": 0.2148170918226242,
      "learning_rate": 4.117100371747212e-05,
      "loss": 0.2474,
      "step": 576
    },
    {
      "epoch": 2.068222621184919,
      "grad_norm": 0.22428511083126068,
      "learning_rate": 4.115241635687733e-05,
      "loss": 0.2942,
      "step": 577
    },
    {
      "epoch": 2.0718132854578095,
      "grad_norm": 0.2216494083404541,
      "learning_rate": 4.113382899628253e-05,
      "loss": 0.243,
      "step": 578
    },
    {
      "epoch": 2.0754039497307,
      "grad_norm": 0.1910577118396759,
      "learning_rate": 4.111524163568773e-05,
      "loss": 0.2868,
      "step": 579
    },
    {
      "epoch": 2.078994614003591,
      "grad_norm": 0.23045727610588074,
      "learning_rate": 4.109665427509294e-05,
      "loss": 0.4197,
      "step": 580
    },
    {
      "epoch": 2.0825852782764813,
      "grad_norm": 0.29146772623062134,
      "learning_rate": 4.107806691449814e-05,
      "loss": 0.7042,
      "step": 581
    },
    {
      "epoch": 2.0861759425493718,
      "grad_norm": 0.26387372612953186,
      "learning_rate": 4.105947955390334e-05,
      "loss": 0.3576,
      "step": 582
    },
    {
      "epoch": 2.0897666068222622,
      "grad_norm": 0.22849120199680328,
      "learning_rate": 4.104089219330855e-05,
      "loss": 0.4363,
      "step": 583
    },
    {
      "epoch": 2.0933572710951527,
      "grad_norm": 0.24442975223064423,
      "learning_rate": 4.102230483271376e-05,
      "loss": 0.7325,
      "step": 584
    },
    {
      "epoch": 2.096947935368043,
      "grad_norm": 0.21099072694778442,
      "learning_rate": 4.100371747211896e-05,
      "loss": 0.4149,
      "step": 585
    },
    {
      "epoch": 2.1005385996409336,
      "grad_norm": 0.16995461285114288,
      "learning_rate": 4.098513011152417e-05,
      "loss": 0.4343,
      "step": 586
    },
    {
      "epoch": 2.104129263913824,
      "grad_norm": 0.2476228028535843,
      "learning_rate": 4.096654275092937e-05,
      "loss": 0.3433,
      "step": 587
    },
    {
      "epoch": 2.1077199281867145,
      "grad_norm": 0.2269037514925003,
      "learning_rate": 4.094795539033457e-05,
      "loss": 0.3239,
      "step": 588
    },
    {
      "epoch": 2.111310592459605,
      "grad_norm": 0.21938857436180115,
      "learning_rate": 4.092936802973978e-05,
      "loss": 0.3537,
      "step": 589
    },
    {
      "epoch": 2.1149012567324954,
      "grad_norm": 0.20075447857379913,
      "learning_rate": 4.091078066914498e-05,
      "loss": 0.3655,
      "step": 590
    },
    {
      "epoch": 2.118491921005386,
      "grad_norm": 0.22805887460708618,
      "learning_rate": 4.0892193308550185e-05,
      "loss": 0.6233,
      "step": 591
    },
    {
      "epoch": 2.1220825852782763,
      "grad_norm": 0.2195020467042923,
      "learning_rate": 4.087360594795539e-05,
      "loss": 0.2648,
      "step": 592
    },
    {
      "epoch": 2.1256732495511668,
      "grad_norm": 0.21565400063991547,
      "learning_rate": 4.08550185873606e-05,
      "loss": 0.2611,
      "step": 593
    },
    {
      "epoch": 2.129263913824057,
      "grad_norm": 0.21944652497768402,
      "learning_rate": 4.08364312267658e-05,
      "loss": 0.3149,
      "step": 594
    },
    {
      "epoch": 2.132854578096948,
      "grad_norm": 0.21269197762012482,
      "learning_rate": 4.0817843866171005e-05,
      "loss": 0.2187,
      "step": 595
    },
    {
      "epoch": 2.1364452423698386,
      "grad_norm": 0.20605073869228363,
      "learning_rate": 4.079925650557621e-05,
      "loss": 0.2831,
      "step": 596
    },
    {
      "epoch": 2.140035906642729,
      "grad_norm": 0.2229422777891159,
      "learning_rate": 4.0780669144981414e-05,
      "loss": 0.4586,
      "step": 597
    },
    {
      "epoch": 2.1436265709156195,
      "grad_norm": 0.23093971610069275,
      "learning_rate": 4.0762081784386616e-05,
      "loss": 0.2458,
      "step": 598
    },
    {
      "epoch": 2.14721723518851,
      "grad_norm": 0.2051890790462494,
      "learning_rate": 4.0743494423791824e-05,
      "loss": 0.3812,
      "step": 599
    },
    {
      "epoch": 2.1508078994614004,
      "grad_norm": 0.20690622925758362,
      "learning_rate": 4.0724907063197026e-05,
      "loss": 0.4564,
      "step": 600
    },
    {
      "epoch": 2.154398563734291,
      "grad_norm": 0.25249406695365906,
      "learning_rate": 4.070631970260223e-05,
      "loss": 0.3528,
      "step": 601
    },
    {
      "epoch": 2.1579892280071813,
      "grad_norm": 0.20594564080238342,
      "learning_rate": 4.068773234200744e-05,
      "loss": 0.244,
      "step": 602
    },
    {
      "epoch": 2.1615798922800717,
      "grad_norm": 0.24631986021995544,
      "learning_rate": 4.0669144981412644e-05,
      "loss": 0.5382,
      "step": 603
    },
    {
      "epoch": 2.165170556552962,
      "grad_norm": 0.24446341395378113,
      "learning_rate": 4.0650557620817846e-05,
      "loss": 0.3298,
      "step": 604
    },
    {
      "epoch": 2.1687612208258527,
      "grad_norm": 0.2270420640707016,
      "learning_rate": 4.0631970260223054e-05,
      "loss": 0.2607,
      "step": 605
    },
    {
      "epoch": 2.172351885098743,
      "grad_norm": 0.17458461225032806,
      "learning_rate": 4.0613382899628256e-05,
      "loss": 0.2785,
      "step": 606
    },
    {
      "epoch": 2.1759425493716336,
      "grad_norm": 0.2115759700536728,
      "learning_rate": 4.059479553903346e-05,
      "loss": 0.3708,
      "step": 607
    },
    {
      "epoch": 2.1795332136445245,
      "grad_norm": 0.19865788519382477,
      "learning_rate": 4.0576208178438666e-05,
      "loss": 0.2405,
      "step": 608
    },
    {
      "epoch": 2.183123877917415,
      "grad_norm": 0.223898246884346,
      "learning_rate": 4.055762081784387e-05,
      "loss": 0.5367,
      "step": 609
    },
    {
      "epoch": 2.1867145421903054,
      "grad_norm": 0.24209162592887878,
      "learning_rate": 4.053903345724907e-05,
      "loss": 0.3687,
      "step": 610
    },
    {
      "epoch": 2.190305206463196,
      "grad_norm": 0.21878036856651306,
      "learning_rate": 4.052044609665428e-05,
      "loss": 0.2585,
      "step": 611
    },
    {
      "epoch": 2.1938958707360863,
      "grad_norm": 0.20252569019794464,
      "learning_rate": 4.0501858736059486e-05,
      "loss": 0.2658,
      "step": 612
    },
    {
      "epoch": 2.1974865350089767,
      "grad_norm": 0.22461776435375214,
      "learning_rate": 4.048327137546469e-05,
      "loss": 0.5568,
      "step": 613
    },
    {
      "epoch": 2.201077199281867,
      "grad_norm": 0.25522252917289734,
      "learning_rate": 4.046468401486989e-05,
      "loss": 0.3982,
      "step": 614
    },
    {
      "epoch": 2.2046678635547576,
      "grad_norm": 0.22976653277873993,
      "learning_rate": 4.04460966542751e-05,
      "loss": 0.2486,
      "step": 615
    },
    {
      "epoch": 2.208258527827648,
      "grad_norm": 0.18421605229377747,
      "learning_rate": 4.04275092936803e-05,
      "loss": 0.4511,
      "step": 616
    },
    {
      "epoch": 2.2118491921005385,
      "grad_norm": 0.1958160400390625,
      "learning_rate": 4.04089219330855e-05,
      "loss": 0.4177,
      "step": 617
    },
    {
      "epoch": 2.215439856373429,
      "grad_norm": 0.2223958522081375,
      "learning_rate": 4.039033457249071e-05,
      "loss": 0.4789,
      "step": 618
    },
    {
      "epoch": 2.2190305206463194,
      "grad_norm": 0.22048398852348328,
      "learning_rate": 4.037174721189591e-05,
      "loss": 0.255,
      "step": 619
    },
    {
      "epoch": 2.22262118491921,
      "grad_norm": 0.26030489802360535,
      "learning_rate": 4.035315985130111e-05,
      "loss": 0.3346,
      "step": 620
    },
    {
      "epoch": 2.2262118491921004,
      "grad_norm": 0.22482699155807495,
      "learning_rate": 4.033457249070633e-05,
      "loss": 0.5221,
      "step": 621
    },
    {
      "epoch": 2.229802513464991,
      "grad_norm": 0.19205592572689056,
      "learning_rate": 4.031598513011153e-05,
      "loss": 0.3583,
      "step": 622
    },
    {
      "epoch": 2.2333931777378817,
      "grad_norm": 0.24615982174873352,
      "learning_rate": 4.029739776951673e-05,
      "loss": 0.401,
      "step": 623
    },
    {
      "epoch": 2.236983842010772,
      "grad_norm": 0.22189082205295563,
      "learning_rate": 4.027881040892193e-05,
      "loss": 0.3307,
      "step": 624
    },
    {
      "epoch": 2.2405745062836626,
      "grad_norm": 0.21865728497505188,
      "learning_rate": 4.026022304832714e-05,
      "loss": 0.3422,
      "step": 625
    },
    {
      "epoch": 2.244165170556553,
      "grad_norm": 0.20585933327674866,
      "learning_rate": 4.024163568773234e-05,
      "loss": 0.2795,
      "step": 626
    },
    {
      "epoch": 2.2477558348294435,
      "grad_norm": 0.22439399361610413,
      "learning_rate": 4.0223048327137544e-05,
      "loss": 0.2484,
      "step": 627
    },
    {
      "epoch": 2.251346499102334,
      "grad_norm": 0.24238793551921844,
      "learning_rate": 4.020446096654275e-05,
      "loss": 0.491,
      "step": 628
    },
    {
      "epoch": 2.2549371633752244,
      "grad_norm": 0.1820775717496872,
      "learning_rate": 4.0185873605947953e-05,
      "loss": 0.3169,
      "step": 629
    },
    {
      "epoch": 2.258527827648115,
      "grad_norm": 0.2727394104003906,
      "learning_rate": 4.016728624535316e-05,
      "loss": 0.3524,
      "step": 630
    },
    {
      "epoch": 2.2621184919210053,
      "grad_norm": 0.236244335770607,
      "learning_rate": 4.014869888475837e-05,
      "loss": 0.552,
      "step": 631
    },
    {
      "epoch": 2.265709156193896,
      "grad_norm": 0.20718300342559814,
      "learning_rate": 4.013011152416357e-05,
      "loss": 0.2669,
      "step": 632
    },
    {
      "epoch": 2.2692998204667862,
      "grad_norm": 0.21164847910404205,
      "learning_rate": 4.0111524163568773e-05,
      "loss": 0.3569,
      "step": 633
    },
    {
      "epoch": 2.2728904847396767,
      "grad_norm": 0.19459955394268036,
      "learning_rate": 4.009293680297398e-05,
      "loss": 0.2745,
      "step": 634
    },
    {
      "epoch": 2.276481149012567,
      "grad_norm": 0.2551638185977936,
      "learning_rate": 4.007434944237918e-05,
      "loss": 0.5704,
      "step": 635
    },
    {
      "epoch": 2.280071813285458,
      "grad_norm": 0.2391808182001114,
      "learning_rate": 4.0055762081784385e-05,
      "loss": 0.419,
      "step": 636
    },
    {
      "epoch": 2.283662477558348,
      "grad_norm": 0.18660658597946167,
      "learning_rate": 4.003717472118959e-05,
      "loss": 0.3959,
      "step": 637
    },
    {
      "epoch": 2.287253141831239,
      "grad_norm": 0.20906654000282288,
      "learning_rate": 4.0018587360594795e-05,
      "loss": 0.3812,
      "step": 638
    },
    {
      "epoch": 2.2908438061041294,
      "grad_norm": 0.21249833703041077,
      "learning_rate": 4e-05,
      "loss": 0.2628,
      "step": 639
    },
    {
      "epoch": 2.29443447037702,
      "grad_norm": 0.26300564408302307,
      "learning_rate": 3.9981412639405205e-05,
      "loss": 0.5022,
      "step": 640
    },
    {
      "epoch": 2.2980251346499103,
      "grad_norm": 0.20901352167129517,
      "learning_rate": 3.996282527881041e-05,
      "loss": 0.2433,
      "step": 641
    },
    {
      "epoch": 2.3016157989228008,
      "grad_norm": 0.22321009635925293,
      "learning_rate": 3.9944237918215615e-05,
      "loss": 0.6564,
      "step": 642
    },
    {
      "epoch": 2.3052064631956912,
      "grad_norm": 0.22152264416217804,
      "learning_rate": 3.9925650557620816e-05,
      "loss": 0.5379,
      "step": 643
    },
    {
      "epoch": 2.3087971274685817,
      "grad_norm": 0.2184627652168274,
      "learning_rate": 3.9907063197026025e-05,
      "loss": 0.5102,
      "step": 644
    },
    {
      "epoch": 2.312387791741472,
      "grad_norm": 0.21871109306812286,
      "learning_rate": 3.9888475836431226e-05,
      "loss": 0.2985,
      "step": 645
    },
    {
      "epoch": 2.3159784560143626,
      "grad_norm": 0.19524914026260376,
      "learning_rate": 3.986988847583643e-05,
      "loss": 0.2771,
      "step": 646
    },
    {
      "epoch": 2.319569120287253,
      "grad_norm": 0.2243078500032425,
      "learning_rate": 3.9851301115241636e-05,
      "loss": 0.4704,
      "step": 647
    },
    {
      "epoch": 2.3231597845601435,
      "grad_norm": 0.2101546823978424,
      "learning_rate": 3.9832713754646845e-05,
      "loss": 0.2632,
      "step": 648
    },
    {
      "epoch": 2.326750448833034,
      "grad_norm": 0.1943230777978897,
      "learning_rate": 3.9814126394052046e-05,
      "loss": 0.3155,
      "step": 649
    },
    {
      "epoch": 2.3303411131059244,
      "grad_norm": 0.18585434556007385,
      "learning_rate": 3.9795539033457255e-05,
      "loss": 0.3567,
      "step": 650
    },
    {
      "epoch": 2.3339317773788153,
      "grad_norm": 0.20737795531749725,
      "learning_rate": 3.9776951672862456e-05,
      "loss": 0.364,
      "step": 651
    },
    {
      "epoch": 2.3375224416517058,
      "grad_norm": 0.24219688773155212,
      "learning_rate": 3.975836431226766e-05,
      "loss": 0.2834,
      "step": 652
    },
    {
      "epoch": 2.341113105924596,
      "grad_norm": 0.21932369470596313,
      "learning_rate": 3.9739776951672866e-05,
      "loss": 0.2462,
      "step": 653
    },
    {
      "epoch": 2.3447037701974867,
      "grad_norm": 0.2433534562587738,
      "learning_rate": 3.972118959107807e-05,
      "loss": 0.4063,
      "step": 654
    },
    {
      "epoch": 2.348294434470377,
      "grad_norm": 0.21031802892684937,
      "learning_rate": 3.970260223048327e-05,
      "loss": 0.2081,
      "step": 655
    },
    {
      "epoch": 2.3518850987432676,
      "grad_norm": 0.20939530432224274,
      "learning_rate": 3.968401486988848e-05,
      "loss": 0.2243,
      "step": 656
    },
    {
      "epoch": 2.355475763016158,
      "grad_norm": 0.20537997782230377,
      "learning_rate": 3.9665427509293686e-05,
      "loss": 0.2575,
      "step": 657
    },
    {
      "epoch": 2.3590664272890485,
      "grad_norm": 0.2059287130832672,
      "learning_rate": 3.964684014869889e-05,
      "loss": 0.2585,
      "step": 658
    },
    {
      "epoch": 2.362657091561939,
      "grad_norm": 0.21840043365955353,
      "learning_rate": 3.962825278810409e-05,
      "loss": 0.8477,
      "step": 659
    },
    {
      "epoch": 2.3662477558348294,
      "grad_norm": 0.22509434819221497,
      "learning_rate": 3.96096654275093e-05,
      "loss": 0.2888,
      "step": 660
    },
    {
      "epoch": 2.36983842010772,
      "grad_norm": 0.21274316310882568,
      "learning_rate": 3.95910780669145e-05,
      "loss": 0.279,
      "step": 661
    },
    {
      "epoch": 2.3734290843806103,
      "grad_norm": 0.21214140951633453,
      "learning_rate": 3.95724907063197e-05,
      "loss": 0.5373,
      "step": 662
    },
    {
      "epoch": 2.3770197486535007,
      "grad_norm": 0.25335779786109924,
      "learning_rate": 3.955390334572491e-05,
      "loss": 0.5205,
      "step": 663
    },
    {
      "epoch": 2.380610412926391,
      "grad_norm": 0.2235039621591568,
      "learning_rate": 3.953531598513011e-05,
      "loss": 0.2726,
      "step": 664
    },
    {
      "epoch": 2.3842010771992816,
      "grad_norm": 0.2175634503364563,
      "learning_rate": 3.951672862453531e-05,
      "loss": 0.2353,
      "step": 665
    },
    {
      "epoch": 2.3877917414721725,
      "grad_norm": 0.21217671036720276,
      "learning_rate": 3.949814126394053e-05,
      "loss": 0.235,
      "step": 666
    },
    {
      "epoch": 2.391382405745063,
      "grad_norm": 0.2158106416463852,
      "learning_rate": 3.947955390334573e-05,
      "loss": 0.4968,
      "step": 667
    },
    {
      "epoch": 2.3949730700179535,
      "grad_norm": 0.2487468272447586,
      "learning_rate": 3.946096654275093e-05,
      "loss": 0.32,
      "step": 668
    },
    {
      "epoch": 2.398563734290844,
      "grad_norm": 0.18173646926879883,
      "learning_rate": 3.944237918215614e-05,
      "loss": 0.3526,
      "step": 669
    },
    {
      "epoch": 2.4021543985637344,
      "grad_norm": 0.22278617322444916,
      "learning_rate": 3.942379182156134e-05,
      "loss": 0.4308,
      "step": 670
    },
    {
      "epoch": 2.405745062836625,
      "grad_norm": 0.22270919382572174,
      "learning_rate": 3.940520446096654e-05,
      "loss": 0.2957,
      "step": 671
    },
    {
      "epoch": 2.4093357271095153,
      "grad_norm": 0.2146954983472824,
      "learning_rate": 3.938661710037175e-05,
      "loss": 0.2097,
      "step": 672
    },
    {
      "epoch": 2.4129263913824057,
      "grad_norm": 0.2354154884815216,
      "learning_rate": 3.936802973977695e-05,
      "loss": 0.434,
      "step": 673
    },
    {
      "epoch": 2.416517055655296,
      "grad_norm": 0.21656747162342072,
      "learning_rate": 3.9349442379182154e-05,
      "loss": 0.2936,
      "step": 674
    },
    {
      "epoch": 2.4201077199281866,
      "grad_norm": 0.2229207158088684,
      "learning_rate": 3.933085501858736e-05,
      "loss": 0.3566,
      "step": 675
    },
    {
      "epoch": 2.423698384201077,
      "grad_norm": 0.1864607185125351,
      "learning_rate": 3.931226765799257e-05,
      "loss": 0.4394,
      "step": 676
    },
    {
      "epoch": 2.4272890484739675,
      "grad_norm": 0.323684424161911,
      "learning_rate": 3.929368029739777e-05,
      "loss": 0.3537,
      "step": 677
    },
    {
      "epoch": 2.430879712746858,
      "grad_norm": 0.21473781764507294,
      "learning_rate": 3.9275092936802974e-05,
      "loss": 0.2355,
      "step": 678
    },
    {
      "epoch": 2.434470377019749,
      "grad_norm": 0.2334701269865036,
      "learning_rate": 3.925650557620818e-05,
      "loss": 0.7503,
      "step": 679
    },
    {
      "epoch": 2.438061041292639,
      "grad_norm": 0.23550528287887573,
      "learning_rate": 3.9237918215613384e-05,
      "loss": 0.4698,
      "step": 680
    },
    {
      "epoch": 2.44165170556553,
      "grad_norm": 0.21039387583732605,
      "learning_rate": 3.9219330855018585e-05,
      "loss": 0.4537,
      "step": 681
    },
    {
      "epoch": 2.4452423698384202,
      "grad_norm": 0.20627757906913757,
      "learning_rate": 3.9200743494423794e-05,
      "loss": 0.26,
      "step": 682
    },
    {
      "epoch": 2.4488330341113107,
      "grad_norm": 0.18321868777275085,
      "learning_rate": 3.9182156133828995e-05,
      "loss": 0.3202,
      "step": 683
    },
    {
      "epoch": 2.452423698384201,
      "grad_norm": 0.22849880158901215,
      "learning_rate": 3.9163568773234204e-05,
      "loss": 0.4398,
      "step": 684
    },
    {
      "epoch": 2.4560143626570916,
      "grad_norm": 0.21748821437358856,
      "learning_rate": 3.914498141263941e-05,
      "loss": 0.4723,
      "step": 685
    },
    {
      "epoch": 2.459605026929982,
      "grad_norm": 0.2234264612197876,
      "learning_rate": 3.9126394052044614e-05,
      "loss": 0.4674,
      "step": 686
    },
    {
      "epoch": 2.4631956912028725,
      "grad_norm": 0.23748402297496796,
      "learning_rate": 3.9107806691449815e-05,
      "loss": 0.3356,
      "step": 687
    },
    {
      "epoch": 2.466786355475763,
      "grad_norm": 0.22270852327346802,
      "learning_rate": 3.9089219330855024e-05,
      "loss": 0.5805,
      "step": 688
    },
    {
      "epoch": 2.4703770197486534,
      "grad_norm": 0.22018538415431976,
      "learning_rate": 3.9070631970260225e-05,
      "loss": 0.2271,
      "step": 689
    },
    {
      "epoch": 2.473967684021544,
      "grad_norm": 0.1985296756029129,
      "learning_rate": 3.905204460966543e-05,
      "loss": 0.3506,
      "step": 690
    },
    {
      "epoch": 2.4775583482944343,
      "grad_norm": 0.19198134541511536,
      "learning_rate": 3.9033457249070635e-05,
      "loss": 0.6122,
      "step": 691
    },
    {
      "epoch": 2.481149012567325,
      "grad_norm": 0.21730239689350128,
      "learning_rate": 3.901486988847584e-05,
      "loss": 0.2284,
      "step": 692
    },
    {
      "epoch": 2.4847396768402152,
      "grad_norm": 0.2051417976617813,
      "learning_rate": 3.8996282527881045e-05,
      "loss": 0.254,
      "step": 693
    },
    {
      "epoch": 2.488330341113106,
      "grad_norm": 0.2547988295555115,
      "learning_rate": 3.897769516728625e-05,
      "loss": 0.4145,
      "step": 694
    },
    {
      "epoch": 2.4919210053859966,
      "grad_norm": 0.21223850548267365,
      "learning_rate": 3.8959107806691455e-05,
      "loss": 0.3143,
      "step": 695
    },
    {
      "epoch": 2.495511669658887,
      "grad_norm": 0.2998548150062561,
      "learning_rate": 3.8940520446096657e-05,
      "loss": 0.3546,
      "step": 696
    },
    {
      "epoch": 2.4991023339317775,
      "grad_norm": 0.2185979038476944,
      "learning_rate": 3.892193308550186e-05,
      "loss": 0.4213,
      "step": 697
    },
    {
      "epoch": 2.502692998204668,
      "grad_norm": 0.2466382533311844,
      "learning_rate": 3.8903345724907067e-05,
      "loss": 0.3208,
      "step": 698
    },
    {
      "epoch": 2.5062836624775584,
      "grad_norm": 0.22230884432792664,
      "learning_rate": 3.888475836431227e-05,
      "loss": 0.485,
      "step": 699
    },
    {
      "epoch": 2.509874326750449,
      "grad_norm": 0.21683518588542938,
      "learning_rate": 3.886617100371747e-05,
      "loss": 0.3325,
      "step": 700
    },
    {
      "epoch": 2.5134649910233393,
      "grad_norm": 0.2718752324581146,
      "learning_rate": 3.884758364312268e-05,
      "loss": 0.2491,
      "step": 701
    },
    {
      "epoch": 2.5170556552962298,
      "grad_norm": 0.22807037830352783,
      "learning_rate": 3.8828996282527886e-05,
      "loss": 0.4282,
      "step": 702
    },
    {
      "epoch": 2.52064631956912,
      "grad_norm": 0.20825286209583282,
      "learning_rate": 3.881040892193309e-05,
      "loss": 0.3282,
      "step": 703
    },
    {
      "epoch": 2.5242369838420107,
      "grad_norm": 0.2053377628326416,
      "learning_rate": 3.879182156133829e-05,
      "loss": 0.581,
      "step": 704
    },
    {
      "epoch": 2.527827648114901,
      "grad_norm": 0.19462762773036957,
      "learning_rate": 3.87732342007435e-05,
      "loss": 0.3073,
      "step": 705
    },
    {
      "epoch": 2.5314183123877916,
      "grad_norm": 0.23764720559120178,
      "learning_rate": 3.87546468401487e-05,
      "loss": 0.4073,
      "step": 706
    },
    {
      "epoch": 2.5350089766606825,
      "grad_norm": 0.24111855030059814,
      "learning_rate": 3.87360594795539e-05,
      "loss": 0.4178,
      "step": 707
    },
    {
      "epoch": 2.5385996409335725,
      "grad_norm": 0.23524948954582214,
      "learning_rate": 3.871747211895911e-05,
      "loss": 0.3151,
      "step": 708
    },
    {
      "epoch": 2.5421903052064634,
      "grad_norm": 0.2086200714111328,
      "learning_rate": 3.869888475836431e-05,
      "loss": 0.2353,
      "step": 709
    },
    {
      "epoch": 2.545780969479354,
      "grad_norm": 0.24298080801963806,
      "learning_rate": 3.868029739776951e-05,
      "loss": 0.3905,
      "step": 710
    },
    {
      "epoch": 2.5493716337522443,
      "grad_norm": 0.26241743564605713,
      "learning_rate": 3.866171003717473e-05,
      "loss": 0.5066,
      "step": 711
    },
    {
      "epoch": 2.5529622980251347,
      "grad_norm": 0.20287124812602997,
      "learning_rate": 3.864312267657993e-05,
      "loss": 0.2977,
      "step": 712
    },
    {
      "epoch": 2.556552962298025,
      "grad_norm": 0.22816255688667297,
      "learning_rate": 3.862453531598513e-05,
      "loss": 0.2395,
      "step": 713
    },
    {
      "epoch": 2.5601436265709157,
      "grad_norm": 0.23237672448158264,
      "learning_rate": 3.860594795539034e-05,
      "loss": 0.4149,
      "step": 714
    },
    {
      "epoch": 2.563734290843806,
      "grad_norm": 0.23388557136058807,
      "learning_rate": 3.858736059479554e-05,
      "loss": 0.2012,
      "step": 715
    },
    {
      "epoch": 2.5673249551166966,
      "grad_norm": 0.2215479612350464,
      "learning_rate": 3.856877323420074e-05,
      "loss": 0.3667,
      "step": 716
    },
    {
      "epoch": 2.570915619389587,
      "grad_norm": 0.23165841400623322,
      "learning_rate": 3.855018587360595e-05,
      "loss": 0.3723,
      "step": 717
    },
    {
      "epoch": 2.5745062836624775,
      "grad_norm": 0.21912583708763123,
      "learning_rate": 3.853159851301115e-05,
      "loss": 0.2335,
      "step": 718
    },
    {
      "epoch": 2.578096947935368,
      "grad_norm": 0.22727367281913757,
      "learning_rate": 3.8513011152416354e-05,
      "loss": 0.303,
      "step": 719
    },
    {
      "epoch": 2.5816876122082584,
      "grad_norm": 0.19240792095661163,
      "learning_rate": 3.849442379182156e-05,
      "loss": 0.3469,
      "step": 720
    },
    {
      "epoch": 2.585278276481149,
      "grad_norm": 0.28860506415367126,
      "learning_rate": 3.847583643122677e-05,
      "loss": 0.4635,
      "step": 721
    },
    {
      "epoch": 2.5888689407540397,
      "grad_norm": 0.2124456763267517,
      "learning_rate": 3.845724907063197e-05,
      "loss": 0.2359,
      "step": 722
    },
    {
      "epoch": 2.5924596050269297,
      "grad_norm": 0.20722822844982147,
      "learning_rate": 3.8438661710037174e-05,
      "loss": 0.3024,
      "step": 723
    },
    {
      "epoch": 2.5960502692998206,
      "grad_norm": 0.20976626873016357,
      "learning_rate": 3.842007434944238e-05,
      "loss": 0.4132,
      "step": 724
    },
    {
      "epoch": 2.599640933572711,
      "grad_norm": 0.1718696802854538,
      "learning_rate": 3.8401486988847584e-05,
      "loss": 0.441,
      "step": 725
    },
    {
      "epoch": 2.6032315978456015,
      "grad_norm": 0.2038293480873108,
      "learning_rate": 3.8382899628252786e-05,
      "loss": 0.2315,
      "step": 726
    },
    {
      "epoch": 2.606822262118492,
      "grad_norm": 0.24002283811569214,
      "learning_rate": 3.8364312267657994e-05,
      "loss": 0.4538,
      "step": 727
    },
    {
      "epoch": 2.6104129263913824,
      "grad_norm": 0.24887163937091827,
      "learning_rate": 3.8345724907063196e-05,
      "loss": 0.3057,
      "step": 728
    },
    {
      "epoch": 2.614003590664273,
      "grad_norm": 0.24764491617679596,
      "learning_rate": 3.8327137546468404e-05,
      "loss": 0.4351,
      "step": 729
    },
    {
      "epoch": 2.6175942549371634,
      "grad_norm": 0.1958940178155899,
      "learning_rate": 3.830855018587361e-05,
      "loss": 0.2081,
      "step": 730
    },
    {
      "epoch": 2.621184919210054,
      "grad_norm": 0.2178664356470108,
      "learning_rate": 3.8289962825278814e-05,
      "loss": 0.2682,
      "step": 731
    },
    {
      "epoch": 2.6247755834829443,
      "grad_norm": 0.2063489556312561,
      "learning_rate": 3.8271375464684015e-05,
      "loss": 0.2633,
      "step": 732
    },
    {
      "epoch": 2.6283662477558347,
      "grad_norm": 0.23657168447971344,
      "learning_rate": 3.8252788104089224e-05,
      "loss": 0.2851,
      "step": 733
    },
    {
      "epoch": 2.631956912028725,
      "grad_norm": 0.16955550014972687,
      "learning_rate": 3.8234200743494425e-05,
      "loss": 0.2771,
      "step": 734
    },
    {
      "epoch": 2.635547576301616,
      "grad_norm": 0.23422382771968842,
      "learning_rate": 3.821561338289963e-05,
      "loss": 0.4701,
      "step": 735
    },
    {
      "epoch": 2.639138240574506,
      "grad_norm": 0.20796895027160645,
      "learning_rate": 3.8197026022304835e-05,
      "loss": 0.2433,
      "step": 736
    },
    {
      "epoch": 2.642728904847397,
      "grad_norm": 0.22373691201210022,
      "learning_rate": 3.817843866171004e-05,
      "loss": 0.3588,
      "step": 737
    },
    {
      "epoch": 2.646319569120287,
      "grad_norm": 0.22111985087394714,
      "learning_rate": 3.8159851301115245e-05,
      "loss": 0.2078,
      "step": 738
    },
    {
      "epoch": 2.649910233393178,
      "grad_norm": 0.2490144520998001,
      "learning_rate": 3.814126394052045e-05,
      "loss": 0.4707,
      "step": 739
    },
    {
      "epoch": 2.6535008976660683,
      "grad_norm": 0.2452942579984665,
      "learning_rate": 3.8122676579925655e-05,
      "loss": 0.4552,
      "step": 740
    },
    {
      "epoch": 2.657091561938959,
      "grad_norm": 0.172909677028656,
      "learning_rate": 3.810408921933086e-05,
      "loss": 0.3624,
      "step": 741
    },
    {
      "epoch": 2.6606822262118492,
      "grad_norm": 0.22822527587413788,
      "learning_rate": 3.808550185873606e-05,
      "loss": 0.5644,
      "step": 742
    },
    {
      "epoch": 2.6642728904847397,
      "grad_norm": 0.4786962568759918,
      "learning_rate": 3.806691449814127e-05,
      "loss": 0.4816,
      "step": 743
    },
    {
      "epoch": 2.66786355475763,
      "grad_norm": 0.21632438898086548,
      "learning_rate": 3.804832713754647e-05,
      "loss": 0.2191,
      "step": 744
    },
    {
      "epoch": 2.6714542190305206,
      "grad_norm": 0.23707488179206848,
      "learning_rate": 3.802973977695167e-05,
      "loss": 0.3573,
      "step": 745
    },
    {
      "epoch": 2.675044883303411,
      "grad_norm": 0.2525188624858856,
      "learning_rate": 3.801115241635688e-05,
      "loss": 0.5263,
      "step": 746
    },
    {
      "epoch": 2.6786355475763015,
      "grad_norm": 0.21515002846717834,
      "learning_rate": 3.799256505576209e-05,
      "loss": 0.2525,
      "step": 747
    },
    {
      "epoch": 2.682226211849192,
      "grad_norm": 0.20307545363903046,
      "learning_rate": 3.797397769516729e-05,
      "loss": 0.2833,
      "step": 748
    },
    {
      "epoch": 2.6858168761220824,
      "grad_norm": 0.22167539596557617,
      "learning_rate": 3.79553903345725e-05,
      "loss": 0.5054,
      "step": 749
    },
    {
      "epoch": 2.6894075403949733,
      "grad_norm": 0.21899880468845367,
      "learning_rate": 3.79368029739777e-05,
      "loss": 0.6181,
      "step": 750
    },
    {
      "epoch": 2.6929982046678633,
      "grad_norm": 0.2619737386703491,
      "learning_rate": 3.79182156133829e-05,
      "loss": 0.3779,
      "step": 751
    },
    {
      "epoch": 2.6965888689407542,
      "grad_norm": 0.2214362472295761,
      "learning_rate": 3.789962825278811e-05,
      "loss": 0.4346,
      "step": 752
    },
    {
      "epoch": 2.7001795332136447,
      "grad_norm": 0.22990389168262482,
      "learning_rate": 3.788104089219331e-05,
      "loss": 0.2542,
      "step": 753
    },
    {
      "epoch": 2.703770197486535,
      "grad_norm": 0.22517742216587067,
      "learning_rate": 3.786245353159851e-05,
      "loss": 0.4326,
      "step": 754
    },
    {
      "epoch": 2.7073608617594256,
      "grad_norm": 0.24093058705329895,
      "learning_rate": 3.784386617100372e-05,
      "loss": 0.5197,
      "step": 755
    },
    {
      "epoch": 2.710951526032316,
      "grad_norm": 0.24075867235660553,
      "learning_rate": 3.782527881040893e-05,
      "loss": 0.4971,
      "step": 756
    },
    {
      "epoch": 2.7145421903052065,
      "grad_norm": 0.23287354409694672,
      "learning_rate": 3.780669144981413e-05,
      "loss": 0.6513,
      "step": 757
    },
    {
      "epoch": 2.718132854578097,
      "grad_norm": 0.24915672838687897,
      "learning_rate": 3.778810408921933e-05,
      "loss": 0.3626,
      "step": 758
    },
    {
      "epoch": 2.7217235188509874,
      "grad_norm": 0.2462567239999771,
      "learning_rate": 3.776951672862454e-05,
      "loss": 0.4646,
      "step": 759
    },
    {
      "epoch": 2.725314183123878,
      "grad_norm": 0.22583819925785065,
      "learning_rate": 3.775092936802974e-05,
      "loss": 0.2841,
      "step": 760
    },
    {
      "epoch": 2.7289048473967683,
      "grad_norm": 0.20874974131584167,
      "learning_rate": 3.773234200743494e-05,
      "loss": 0.2622,
      "step": 761
    },
    {
      "epoch": 2.7324955116696588,
      "grad_norm": 0.22346436977386475,
      "learning_rate": 3.771375464684015e-05,
      "loss": 0.4403,
      "step": 762
    },
    {
      "epoch": 2.736086175942549,
      "grad_norm": 0.2198016345500946,
      "learning_rate": 3.769516728624535e-05,
      "loss": 0.2464,
      "step": 763
    },
    {
      "epoch": 2.7396768402154397,
      "grad_norm": 0.20353373885154724,
      "learning_rate": 3.7676579925650554e-05,
      "loss": 0.2212,
      "step": 764
    },
    {
      "epoch": 2.7432675044883306,
      "grad_norm": 0.2260318547487259,
      "learning_rate": 3.765799256505577e-05,
      "loss": 0.3031,
      "step": 765
    },
    {
      "epoch": 2.7468581687612206,
      "grad_norm": 0.22028394043445587,
      "learning_rate": 3.763940520446097e-05,
      "loss": 0.2432,
      "step": 766
    },
    {
      "epoch": 2.7504488330341115,
      "grad_norm": 0.19746582210063934,
      "learning_rate": 3.762081784386617e-05,
      "loss": 0.4437,
      "step": 767
    },
    {
      "epoch": 2.754039497307002,
      "grad_norm": 0.24079349637031555,
      "learning_rate": 3.760223048327138e-05,
      "loss": 0.3072,
      "step": 768
    },
    {
      "epoch": 2.7576301615798924,
      "grad_norm": 0.2821216583251953,
      "learning_rate": 3.758364312267658e-05,
      "loss": 0.2393,
      "step": 769
    },
    {
      "epoch": 2.761220825852783,
      "grad_norm": 0.24254126846790314,
      "learning_rate": 3.7565055762081784e-05,
      "loss": 0.3153,
      "step": 770
    },
    {
      "epoch": 2.7648114901256733,
      "grad_norm": 0.21517518162727356,
      "learning_rate": 3.754646840148699e-05,
      "loss": 0.2553,
      "step": 771
    },
    {
      "epoch": 2.7684021543985637,
      "grad_norm": 0.18402713537216187,
      "learning_rate": 3.7527881040892194e-05,
      "loss": 0.4268,
      "step": 772
    },
    {
      "epoch": 2.771992818671454,
      "grad_norm": 0.22168104350566864,
      "learning_rate": 3.7509293680297396e-05,
      "loss": 0.2504,
      "step": 773
    },
    {
      "epoch": 2.7755834829443446,
      "grad_norm": 0.2077265828847885,
      "learning_rate": 3.7490706319702604e-05,
      "loss": 0.1996,
      "step": 774
    },
    {
      "epoch": 2.779174147217235,
      "grad_norm": 0.19974707067012787,
      "learning_rate": 3.747211895910781e-05,
      "loss": 0.2095,
      "step": 775
    },
    {
      "epoch": 2.7827648114901256,
      "grad_norm": 0.23894919455051422,
      "learning_rate": 3.7453531598513014e-05,
      "loss": 0.3112,
      "step": 776
    },
    {
      "epoch": 2.786355475763016,
      "grad_norm": 0.23281501233577728,
      "learning_rate": 3.7434944237918216e-05,
      "loss": 0.4275,
      "step": 777
    },
    {
      "epoch": 2.789946140035907,
      "grad_norm": 0.25066229701042175,
      "learning_rate": 3.7416356877323424e-05,
      "loss": 0.3429,
      "step": 778
    },
    {
      "epoch": 2.793536804308797,
      "grad_norm": 0.25102099776268005,
      "learning_rate": 3.7397769516728626e-05,
      "loss": 0.3572,
      "step": 779
    },
    {
      "epoch": 2.797127468581688,
      "grad_norm": 0.2247728705406189,
      "learning_rate": 3.737918215613383e-05,
      "loss": 0.2162,
      "step": 780
    },
    {
      "epoch": 2.800718132854578,
      "grad_norm": 0.2391573190689087,
      "learning_rate": 3.7360594795539036e-05,
      "loss": 0.5201,
      "step": 781
    },
    {
      "epoch": 2.8043087971274687,
      "grad_norm": 0.20773687958717346,
      "learning_rate": 3.734200743494424e-05,
      "loss": 0.2641,
      "step": 782
    },
    {
      "epoch": 2.807899461400359,
      "grad_norm": 0.21373656392097473,
      "learning_rate": 3.7323420074349446e-05,
      "loss": 0.2529,
      "step": 783
    },
    {
      "epoch": 2.8114901256732496,
      "grad_norm": 0.22377777099609375,
      "learning_rate": 3.730483271375465e-05,
      "loss": 0.2234,
      "step": 784
    },
    {
      "epoch": 2.81508078994614,
      "grad_norm": 0.196843683719635,
      "learning_rate": 3.7286245353159856e-05,
      "loss": 0.3517,
      "step": 785
    },
    {
      "epoch": 2.8186714542190305,
      "grad_norm": 0.21378426253795624,
      "learning_rate": 3.726765799256506e-05,
      "loss": 0.2646,
      "step": 786
    },
    {
      "epoch": 2.822262118491921,
      "grad_norm": 0.2156754434108734,
      "learning_rate": 3.724907063197026e-05,
      "loss": 0.3193,
      "step": 787
    },
    {
      "epoch": 2.8258527827648114,
      "grad_norm": 0.1853809803724289,
      "learning_rate": 3.723048327137547e-05,
      "loss": 0.2371,
      "step": 788
    },
    {
      "epoch": 2.829443447037702,
      "grad_norm": 0.21774469316005707,
      "learning_rate": 3.721189591078067e-05,
      "loss": 0.2631,
      "step": 789
    },
    {
      "epoch": 2.8330341113105924,
      "grad_norm": 0.24335286021232605,
      "learning_rate": 3.719330855018587e-05,
      "loss": 0.4058,
      "step": 790
    },
    {
      "epoch": 2.836624775583483,
      "grad_norm": 0.26099446415901184,
      "learning_rate": 3.717472118959108e-05,
      "loss": 0.3063,
      "step": 791
    },
    {
      "epoch": 2.8402154398563733,
      "grad_norm": 0.23308567702770233,
      "learning_rate": 3.715613382899629e-05,
      "loss": 0.4334,
      "step": 792
    },
    {
      "epoch": 2.843806104129264,
      "grad_norm": 0.22924552857875824,
      "learning_rate": 3.713754646840149e-05,
      "loss": 0.2618,
      "step": 793
    },
    {
      "epoch": 2.847396768402154,
      "grad_norm": 0.2257472425699234,
      "learning_rate": 3.71189591078067e-05,
      "loss": 0.2551,
      "step": 794
    },
    {
      "epoch": 2.850987432675045,
      "grad_norm": 0.24115467071533203,
      "learning_rate": 3.71003717472119e-05,
      "loss": 0.4813,
      "step": 795
    },
    {
      "epoch": 2.8545780969479355,
      "grad_norm": 0.21469315886497498,
      "learning_rate": 3.70817843866171e-05,
      "loss": 0.3703,
      "step": 796
    },
    {
      "epoch": 2.858168761220826,
      "grad_norm": 0.22838495671749115,
      "learning_rate": 3.706319702602231e-05,
      "loss": 0.2822,
      "step": 797
    },
    {
      "epoch": 2.8617594254937164,
      "grad_norm": 0.22902695834636688,
      "learning_rate": 3.704460966542751e-05,
      "loss": 0.2106,
      "step": 798
    },
    {
      "epoch": 2.865350089766607,
      "grad_norm": 0.22726859152317047,
      "learning_rate": 3.702602230483271e-05,
      "loss": 0.2564,
      "step": 799
    },
    {
      "epoch": 2.8689407540394973,
      "grad_norm": 0.21750037372112274,
      "learning_rate": 3.700743494423792e-05,
      "loss": 0.5452,
      "step": 800
    },
    {
      "epoch": 2.872531418312388,
      "grad_norm": 0.2247849702835083,
      "learning_rate": 3.698884758364313e-05,
      "loss": 0.4754,
      "step": 801
    },
    {
      "epoch": 2.8761220825852782,
      "grad_norm": 0.2383289784193039,
      "learning_rate": 3.697026022304833e-05,
      "loss": 0.2444,
      "step": 802
    },
    {
      "epoch": 2.8797127468581687,
      "grad_norm": 0.23990707099437714,
      "learning_rate": 3.695167286245353e-05,
      "loss": 0.3302,
      "step": 803
    },
    {
      "epoch": 2.883303411131059,
      "grad_norm": 0.24475328624248505,
      "learning_rate": 3.693308550185874e-05,
      "loss": 0.7169,
      "step": 804
    },
    {
      "epoch": 2.8868940754039496,
      "grad_norm": 0.2279435396194458,
      "learning_rate": 3.691449814126394e-05,
      "loss": 0.2887,
      "step": 805
    },
    {
      "epoch": 2.89048473967684,
      "grad_norm": 0.24259644746780396,
      "learning_rate": 3.689591078066914e-05,
      "loss": 0.4699,
      "step": 806
    },
    {
      "epoch": 2.8940754039497305,
      "grad_norm": 0.21734336018562317,
      "learning_rate": 3.687732342007435e-05,
      "loss": 0.2462,
      "step": 807
    },
    {
      "epoch": 2.8976660682226214,
      "grad_norm": 0.21373748779296875,
      "learning_rate": 3.685873605947955e-05,
      "loss": 0.278,
      "step": 808
    },
    {
      "epoch": 2.9012567324955114,
      "grad_norm": 0.22459064424037933,
      "learning_rate": 3.6840148698884755e-05,
      "loss": 0.5252,
      "step": 809
    },
    {
      "epoch": 2.9048473967684023,
      "grad_norm": 0.21417276561260223,
      "learning_rate": 3.682156133828997e-05,
      "loss": 0.2604,
      "step": 810
    },
    {
      "epoch": 2.9084380610412928,
      "grad_norm": 0.23451541364192963,
      "learning_rate": 3.680297397769517e-05,
      "loss": 0.2681,
      "step": 811
    },
    {
      "epoch": 2.912028725314183,
      "grad_norm": 0.2309127151966095,
      "learning_rate": 3.678438661710037e-05,
      "loss": 0.2886,
      "step": 812
    },
    {
      "epoch": 2.9156193895870737,
      "grad_norm": 0.24367988109588623,
      "learning_rate": 3.676579925650558e-05,
      "loss": 0.3467,
      "step": 813
    },
    {
      "epoch": 2.919210053859964,
      "grad_norm": 0.23962272703647614,
      "learning_rate": 3.674721189591078e-05,
      "loss": 0.2872,
      "step": 814
    },
    {
      "epoch": 2.9228007181328546,
      "grad_norm": 0.21698503196239471,
      "learning_rate": 3.6728624535315985e-05,
      "loss": 0.2436,
      "step": 815
    },
    {
      "epoch": 2.926391382405745,
      "grad_norm": 0.21728599071502686,
      "learning_rate": 3.671003717472119e-05,
      "loss": 0.2175,
      "step": 816
    },
    {
      "epoch": 2.9299820466786355,
      "grad_norm": 0.22689121961593628,
      "learning_rate": 3.6691449814126395e-05,
      "loss": 0.2507,
      "step": 817
    },
    {
      "epoch": 2.933572710951526,
      "grad_norm": 0.24804234504699707,
      "learning_rate": 3.6672862453531596e-05,
      "loss": 0.2742,
      "step": 818
    },
    {
      "epoch": 2.9371633752244164,
      "grad_norm": 0.21934914588928223,
      "learning_rate": 3.6654275092936805e-05,
      "loss": 0.3031,
      "step": 819
    },
    {
      "epoch": 2.940754039497307,
      "grad_norm": 0.22417475283145905,
      "learning_rate": 3.663568773234201e-05,
      "loss": 0.2748,
      "step": 820
    },
    {
      "epoch": 2.9443447037701977,
      "grad_norm": 0.21038201451301575,
      "learning_rate": 3.6617100371747215e-05,
      "loss": 0.3384,
      "step": 821
    },
    {
      "epoch": 2.9479353680430878,
      "grad_norm": 0.23272216320037842,
      "learning_rate": 3.6598513011152416e-05,
      "loss": 0.2416,
      "step": 822
    },
    {
      "epoch": 2.9515260323159787,
      "grad_norm": 0.19926446676254272,
      "learning_rate": 3.6579925650557625e-05,
      "loss": 0.3208,
      "step": 823
    },
    {
      "epoch": 2.9551166965888687,
      "grad_norm": 0.30075350403785706,
      "learning_rate": 3.6561338289962826e-05,
      "loss": 0.5777,
      "step": 824
    },
    {
      "epoch": 2.9587073608617596,
      "grad_norm": 0.21599018573760986,
      "learning_rate": 3.654275092936803e-05,
      "loss": 0.2586,
      "step": 825
    },
    {
      "epoch": 2.96229802513465,
      "grad_norm": 0.24514907598495483,
      "learning_rate": 3.6524163568773236e-05,
      "loss": 0.3378,
      "step": 826
    },
    {
      "epoch": 2.9658886894075405,
      "grad_norm": 0.31676822900772095,
      "learning_rate": 3.650557620817844e-05,
      "loss": 0.5152,
      "step": 827
    },
    {
      "epoch": 2.969479353680431,
      "grad_norm": 0.20849023759365082,
      "learning_rate": 3.6486988847583646e-05,
      "loss": 0.3311,
      "step": 828
    },
    {
      "epoch": 2.9730700179533214,
      "grad_norm": 0.24945573508739471,
      "learning_rate": 3.6468401486988854e-05,
      "loss": 0.6275,
      "step": 829
    },
    {
      "epoch": 2.976660682226212,
      "grad_norm": 0.2051740288734436,
      "learning_rate": 3.6449814126394056e-05,
      "loss": 0.4788,
      "step": 830
    },
    {
      "epoch": 2.9802513464991023,
      "grad_norm": 0.22466179728507996,
      "learning_rate": 3.643122676579926e-05,
      "loss": 0.2386,
      "step": 831
    },
    {
      "epoch": 2.9838420107719927,
      "grad_norm": 0.2456570714712143,
      "learning_rate": 3.6412639405204466e-05,
      "loss": 0.5835,
      "step": 832
    },
    {
      "epoch": 2.987432675044883,
      "grad_norm": 0.22517696022987366,
      "learning_rate": 3.639405204460967e-05,
      "loss": 0.2328,
      "step": 833
    },
    {
      "epoch": 2.9910233393177736,
      "grad_norm": 0.1714985966682434,
      "learning_rate": 3.637546468401487e-05,
      "loss": 0.3487,
      "step": 834
    },
    {
      "epoch": 2.994614003590664,
      "grad_norm": 0.22309787571430206,
      "learning_rate": 3.635687732342008e-05,
      "loss": 0.4502,
      "step": 835
    },
    {
      "epoch": 2.998204667863555,
      "grad_norm": 0.21070626378059387,
      "learning_rate": 3.633828996282528e-05,
      "loss": 0.2134,
      "step": 836
    },
    {
      "epoch": 2.998204667863555,
      "eval_loss": 0.24989382922649384,
      "eval_runtime": 4.8915,
      "eval_samples_per_second": 1.022,
      "eval_steps_per_second": 1.022,
      "step": 836
    },
    {
      "epoch": 2.998204667863555,
      "eval_loss": 0.5793555378913879,
      "eval_runtime": 3.2163,
      "eval_samples_per_second": 1.555,
      "eval_steps_per_second": 1.555,
      "step": 836
    },
    {
      "epoch": 2.998204667863555,
      "eval_loss": 0.3387088477611542,
      "eval_runtime": 4.7248,
      "eval_samples_per_second": 1.058,
      "eval_steps_per_second": 1.058,
      "step": 836
    },
    {
      "epoch": 2.998204667863555,
      "eval_loss": 0.9338661432266235,
      "eval_runtime": 4.958,
      "eval_samples_per_second": 1.008,
      "eval_steps_per_second": 1.008,
      "step": 836
    },
    {
      "epoch": 3.0035906642728905,
      "grad_norm": 0.42102205753326416,
      "learning_rate": 3.631970260223049e-05,
      "loss": 0.4628,
      "step": 837
    },
    {
      "epoch": 3.007181328545781,
      "grad_norm": 0.2277543693780899,
      "learning_rate": 3.630111524163569e-05,
      "loss": 0.302,
      "step": 838
    },
    {
      "epoch": 3.0107719928186714,
      "grad_norm": 0.2387964427471161,
      "learning_rate": 3.62825278810409e-05,
      "loss": 0.2627,
      "step": 839
    },
    {
      "epoch": 3.014362657091562,
      "grad_norm": 0.2366485446691513,
      "learning_rate": 3.62639405204461e-05,
      "loss": 0.3437,
      "step": 840
    },
    {
      "epoch": 3.0179533213644523,
      "grad_norm": 0.231557235121727,
      "learning_rate": 3.62453531598513e-05,
      "loss": 0.209,
      "step": 841
    },
    {
      "epoch": 3.0215439856373427,
      "grad_norm": 0.20200546085834503,
      "learning_rate": 3.622676579925651e-05,
      "loss": 0.3719,
      "step": 842
    },
    {
      "epoch": 3.025134649910233,
      "grad_norm": 0.2607450485229492,
      "learning_rate": 3.620817843866171e-05,
      "loss": 0.263,
      "step": 843
    },
    {
      "epoch": 3.028725314183124,
      "grad_norm": 0.2630870044231415,
      "learning_rate": 3.618959107806691e-05,
      "loss": 0.3476,
      "step": 844
    },
    {
      "epoch": 3.0323159784560145,
      "grad_norm": 0.23129869997501373,
      "learning_rate": 3.617100371747212e-05,
      "loss": 0.2489,
      "step": 845
    },
    {
      "epoch": 3.035906642728905,
      "grad_norm": 0.2324124276638031,
      "learning_rate": 3.615241635687733e-05,
      "loss": 0.3842,
      "step": 846
    },
    {
      "epoch": 3.0394973070017954,
      "grad_norm": 0.22826291620731354,
      "learning_rate": 3.613382899628253e-05,
      "loss": 0.2092,
      "step": 847
    },
    {
      "epoch": 3.043087971274686,
      "grad_norm": 0.2717207074165344,
      "learning_rate": 3.611524163568773e-05,
      "loss": 0.4617,
      "step": 848
    },
    {
      "epoch": 3.0466786355475763,
      "grad_norm": 0.21705307066440582,
      "learning_rate": 3.609665427509294e-05,
      "loss": 0.2234,
      "step": 849
    },
    {
      "epoch": 3.050269299820467,
      "grad_norm": 0.2054000198841095,
      "learning_rate": 3.607806691449814e-05,
      "loss": 0.4793,
      "step": 850
    },
    {
      "epoch": 3.0538599640933572,
      "grad_norm": 0.2032655030488968,
      "learning_rate": 3.6059479553903344e-05,
      "loss": 0.3357,
      "step": 851
    },
    {
      "epoch": 3.0574506283662477,
      "grad_norm": 0.22404751181602478,
      "learning_rate": 3.604089219330855e-05,
      "loss": 0.235,
      "step": 852
    },
    {
      "epoch": 3.061041292639138,
      "grad_norm": 0.2352001667022705,
      "learning_rate": 3.6022304832713754e-05,
      "loss": 0.2158,
      "step": 853
    },
    {
      "epoch": 3.0646319569120286,
      "grad_norm": 0.26029929518699646,
      "learning_rate": 3.6003717472118955e-05,
      "loss": 0.3701,
      "step": 854
    },
    {
      "epoch": 3.068222621184919,
      "grad_norm": 0.22839951515197754,
      "learning_rate": 3.598513011152417e-05,
      "loss": 0.2117,
      "step": 855
    },
    {
      "epoch": 3.0718132854578095,
      "grad_norm": 0.24447956681251526,
      "learning_rate": 3.596654275092937e-05,
      "loss": 0.4173,
      "step": 856
    },
    {
      "epoch": 3.0754039497307,
      "grad_norm": 0.22876973450183868,
      "learning_rate": 3.5947955390334573e-05,
      "loss": 0.3585,
      "step": 857
    },
    {
      "epoch": 3.078994614003591,
      "grad_norm": 0.2286512851715088,
      "learning_rate": 3.592936802973978e-05,
      "loss": 0.6644,
      "step": 858
    },
    {
      "epoch": 3.0825852782764813,
      "grad_norm": 0.19107036292552948,
      "learning_rate": 3.5910780669144983e-05,
      "loss": 0.2985,
      "step": 859
    },
    {
      "epoch": 3.0861759425493718,
      "grad_norm": 0.23487545549869537,
      "learning_rate": 3.5892193308550185e-05,
      "loss": 0.2713,
      "step": 860
    },
    {
      "epoch": 3.0897666068222622,
      "grad_norm": 0.2414615899324417,
      "learning_rate": 3.5873605947955393e-05,
      "loss": 0.2293,
      "step": 861
    },
    {
      "epoch": 3.0933572710951527,
      "grad_norm": 0.22764723002910614,
      "learning_rate": 3.5855018587360595e-05,
      "loss": 0.2157,
      "step": 862
    },
    {
      "epoch": 3.096947935368043,
      "grad_norm": 0.23762504756450653,
      "learning_rate": 3.5836431226765797e-05,
      "loss": 0.2552,
      "step": 863
    },
    {
      "epoch": 3.1005385996409336,
      "grad_norm": 0.22676469385623932,
      "learning_rate": 3.5817843866171005e-05,
      "loss": 0.2258,
      "step": 864
    },
    {
      "epoch": 3.104129263913824,
      "grad_norm": 0.21756449341773987,
      "learning_rate": 3.579925650557621e-05,
      "loss": 0.2046,
      "step": 865
    },
    {
      "epoch": 3.1077199281867145,
      "grad_norm": 0.21746301651000977,
      "learning_rate": 3.5780669144981415e-05,
      "loss": 0.2156,
      "step": 866
    },
    {
      "epoch": 3.111310592459605,
      "grad_norm": 0.2278330773115158,
      "learning_rate": 3.5762081784386617e-05,
      "loss": 0.234,
      "step": 867
    },
    {
      "epoch": 3.1149012567324954,
      "grad_norm": 0.292064905166626,
      "learning_rate": 3.5743494423791825e-05,
      "loss": 0.3262,
      "step": 868
    },
    {
      "epoch": 3.118491921005386,
      "grad_norm": 0.22023943066596985,
      "learning_rate": 3.5724907063197026e-05,
      "loss": 0.2401,
      "step": 869
    },
    {
      "epoch": 3.1220825852782763,
      "grad_norm": 0.2602372467517853,
      "learning_rate": 3.570631970260223e-05,
      "loss": 0.2972,
      "step": 870
    },
    {
      "epoch": 3.1256732495511668,
      "grad_norm": 0.24960604310035706,
      "learning_rate": 3.5687732342007436e-05,
      "loss": 0.25,
      "step": 871
    },
    {
      "epoch": 3.129263913824057,
      "grad_norm": 0.25422757863998413,
      "learning_rate": 3.566914498141264e-05,
      "loss": 0.4306,
      "step": 872
    },
    {
      "epoch": 3.132854578096948,
      "grad_norm": 0.26995566487312317,
      "learning_rate": 3.5650557620817846e-05,
      "loss": 0.4975,
      "step": 873
    },
    {
      "epoch": 3.1364452423698386,
      "grad_norm": 0.21613450348377228,
      "learning_rate": 3.5631970260223055e-05,
      "loss": 0.3371,
      "step": 874
    },
    {
      "epoch": 3.140035906642729,
      "grad_norm": 0.22593232989311218,
      "learning_rate": 3.5613382899628256e-05,
      "loss": 0.375,
      "step": 875
    },
    {
      "epoch": 3.1436265709156195,
      "grad_norm": 0.24248242378234863,
      "learning_rate": 3.559479553903346e-05,
      "loss": 0.2516,
      "step": 876
    },
    {
      "epoch": 3.14721723518851,
      "grad_norm": 0.23180945217609406,
      "learning_rate": 3.5576208178438666e-05,
      "loss": 0.2105,
      "step": 877
    },
    {
      "epoch": 3.1508078994614004,
      "grad_norm": 0.25550076365470886,
      "learning_rate": 3.555762081784387e-05,
      "loss": 0.4488,
      "step": 878
    },
    {
      "epoch": 3.154398563734291,
      "grad_norm": 0.23841145634651184,
      "learning_rate": 3.553903345724907e-05,
      "loss": 0.356,
      "step": 879
    },
    {
      "epoch": 3.1579892280071813,
      "grad_norm": 0.23527619242668152,
      "learning_rate": 3.552044609665428e-05,
      "loss": 0.2135,
      "step": 880
    },
    {
      "epoch": 3.1615798922800717,
      "grad_norm": 0.20658589899539948,
      "learning_rate": 3.550185873605948e-05,
      "loss": 0.2655,
      "step": 881
    },
    {
      "epoch": 3.165170556552962,
      "grad_norm": 0.2505059242248535,
      "learning_rate": 3.548327137546469e-05,
      "loss": 0.3536,
      "step": 882
    },
    {
      "epoch": 3.1687612208258527,
      "grad_norm": 0.23881104588508606,
      "learning_rate": 3.546468401486989e-05,
      "loss": 0.6058,
      "step": 883
    },
    {
      "epoch": 3.172351885098743,
      "grad_norm": 0.24098846316337585,
      "learning_rate": 3.54460966542751e-05,
      "loss": 0.2765,
      "step": 884
    },
    {
      "epoch": 3.1759425493716336,
      "grad_norm": 0.22490918636322021,
      "learning_rate": 3.54275092936803e-05,
      "loss": 0.2028,
      "step": 885
    },
    {
      "epoch": 3.1795332136445245,
      "grad_norm": 0.2743310332298279,
      "learning_rate": 3.54089219330855e-05,
      "loss": 0.3877,
      "step": 886
    },
    {
      "epoch": 3.183123877917415,
      "grad_norm": 0.21716614067554474,
      "learning_rate": 3.539033457249071e-05,
      "loss": 0.3151,
      "step": 887
    },
    {
      "epoch": 3.1867145421903054,
      "grad_norm": 0.2769034504890442,
      "learning_rate": 3.537174721189591e-05,
      "loss": 0.3703,
      "step": 888
    },
    {
      "epoch": 3.190305206463196,
      "grad_norm": 0.23417754471302032,
      "learning_rate": 3.535315985130111e-05,
      "loss": 0.3027,
      "step": 889
    },
    {
      "epoch": 3.1938958707360863,
      "grad_norm": 0.2200378030538559,
      "learning_rate": 3.533457249070632e-05,
      "loss": 0.1821,
      "step": 890
    },
    {
      "epoch": 3.1974865350089767,
      "grad_norm": 0.19787663221359253,
      "learning_rate": 3.531598513011153e-05,
      "loss": 0.4399,
      "step": 891
    },
    {
      "epoch": 3.201077199281867,
      "grad_norm": 0.26004669070243835,
      "learning_rate": 3.529739776951673e-05,
      "loss": 0.5516,
      "step": 892
    },
    {
      "epoch": 3.2046678635547576,
      "grad_norm": 0.21507258713245392,
      "learning_rate": 3.527881040892194e-05,
      "loss": 0.2182,
      "step": 893
    },
    {
      "epoch": 3.208258527827648,
      "grad_norm": 0.20027105510234833,
      "learning_rate": 3.526022304832714e-05,
      "loss": 0.2201,
      "step": 894
    },
    {
      "epoch": 3.2118491921005385,
      "grad_norm": 0.2260197103023529,
      "learning_rate": 3.524163568773234e-05,
      "loss": 0.2175,
      "step": 895
    },
    {
      "epoch": 3.215439856373429,
      "grad_norm": 0.2894649803638458,
      "learning_rate": 3.522304832713755e-05,
      "loss": 0.3554,
      "step": 896
    },
    {
      "epoch": 3.2190305206463194,
      "grad_norm": 0.2269311547279358,
      "learning_rate": 3.520446096654275e-05,
      "loss": 0.5072,
      "step": 897
    },
    {
      "epoch": 3.22262118491921,
      "grad_norm": 0.2356909215450287,
      "learning_rate": 3.5185873605947954e-05,
      "loss": 0.2716,
      "step": 898
    },
    {
      "epoch": 3.2262118491921004,
      "grad_norm": 0.243980273604393,
      "learning_rate": 3.516728624535316e-05,
      "loss": 0.2578,
      "step": 899
    },
    {
      "epoch": 3.229802513464991,
      "grad_norm": 0.23752309381961823,
      "learning_rate": 3.514869888475837e-05,
      "loss": 0.4524,
      "step": 900
    },
    {
      "epoch": 3.2333931777378817,
      "grad_norm": 0.20363779366016388,
      "learning_rate": 3.513011152416357e-05,
      "loss": 0.3874,
      "step": 901
    },
    {
      "epoch": 3.236983842010772,
      "grad_norm": 0.258384644985199,
      "learning_rate": 3.5111524163568774e-05,
      "loss": 0.2863,
      "step": 902
    },
    {
      "epoch": 3.2405745062836626,
      "grad_norm": 0.23435814678668976,
      "learning_rate": 3.509293680297398e-05,
      "loss": 0.2048,
      "step": 903
    },
    {
      "epoch": 3.244165170556553,
      "grad_norm": 0.23492570221424103,
      "learning_rate": 3.5074349442379184e-05,
      "loss": 0.2184,
      "step": 904
    },
    {
      "epoch": 3.2477558348294435,
      "grad_norm": 0.2166668027639389,
      "learning_rate": 3.5055762081784385e-05,
      "loss": 0.3047,
      "step": 905
    },
    {
      "epoch": 3.251346499102334,
      "grad_norm": 0.26951858401298523,
      "learning_rate": 3.5037174721189594e-05,
      "loss": 0.2608,
      "step": 906
    },
    {
      "epoch": 3.2549371633752244,
      "grad_norm": 0.2568517029285431,
      "learning_rate": 3.5018587360594795e-05,
      "loss": 0.2702,
      "step": 907
    },
    {
      "epoch": 3.258527827648115,
      "grad_norm": 0.23653770983219147,
      "learning_rate": 3.5e-05,
      "loss": 0.3424,
      "step": 908
    },
    {
      "epoch": 3.2621184919210053,
      "grad_norm": 0.23119595646858215,
      "learning_rate": 3.498141263940521e-05,
      "loss": 0.2292,
      "step": 909
    },
    {
      "epoch": 3.265709156193896,
      "grad_norm": 0.2254658341407776,
      "learning_rate": 3.4962825278810414e-05,
      "loss": 0.2163,
      "step": 910
    },
    {
      "epoch": 3.2692998204667862,
      "grad_norm": 0.2580417990684509,
      "learning_rate": 3.4944237918215615e-05,
      "loss": 0.3022,
      "step": 911
    },
    {
      "epoch": 3.2728904847396767,
      "grad_norm": 0.24373739957809448,
      "learning_rate": 3.4925650557620824e-05,
      "loss": 0.4604,
      "step": 912
    },
    {
      "epoch": 3.276481149012567,
      "grad_norm": 0.22800825536251068,
      "learning_rate": 3.4907063197026025e-05,
      "loss": 0.2756,
      "step": 913
    },
    {
      "epoch": 3.280071813285458,
      "grad_norm": 0.24947123229503632,
      "learning_rate": 3.488847583643123e-05,
      "loss": 0.271,
      "step": 914
    },
    {
      "epoch": 3.283662477558348,
      "grad_norm": 0.2299010157585144,
      "learning_rate": 3.4869888475836435e-05,
      "loss": 0.5015,
      "step": 915
    },
    {
      "epoch": 3.287253141831239,
      "grad_norm": 0.22028698027133942,
      "learning_rate": 3.485130111524164e-05,
      "loss": 0.316,
      "step": 916
    },
    {
      "epoch": 3.2908438061041294,
      "grad_norm": 0.2375858575105667,
      "learning_rate": 3.483271375464684e-05,
      "loss": 0.2211,
      "step": 917
    },
    {
      "epoch": 3.29443447037702,
      "grad_norm": 0.23691894114017487,
      "learning_rate": 3.481412639405205e-05,
      "loss": 0.206,
      "step": 918
    },
    {
      "epoch": 3.2980251346499103,
      "grad_norm": 0.2448868453502655,
      "learning_rate": 3.4795539033457255e-05,
      "loss": 0.2557,
      "step": 919
    },
    {
      "epoch": 3.3016157989228008,
      "grad_norm": 0.25631511211395264,
      "learning_rate": 3.477695167286246e-05,
      "loss": 0.6096,
      "step": 920
    },
    {
      "epoch": 3.3052064631956912,
      "grad_norm": 0.23671364784240723,
      "learning_rate": 3.475836431226766e-05,
      "loss": 0.2372,
      "step": 921
    },
    {
      "epoch": 3.3087971274685817,
      "grad_norm": 0.17365451157093048,
      "learning_rate": 3.473977695167287e-05,
      "loss": 0.2717,
      "step": 922
    },
    {
      "epoch": 3.312387791741472,
      "grad_norm": 0.22741751372814178,
      "learning_rate": 3.472118959107807e-05,
      "loss": 0.386,
      "step": 923
    },
    {
      "epoch": 3.3159784560143626,
      "grad_norm": 0.19589616358280182,
      "learning_rate": 3.470260223048327e-05,
      "loss": 0.3074,
      "step": 924
    },
    {
      "epoch": 3.319569120287253,
      "grad_norm": 0.22598440945148468,
      "learning_rate": 3.468401486988848e-05,
      "loss": 0.3867,
      "step": 925
    },
    {
      "epoch": 3.3231597845601435,
      "grad_norm": 0.24053017795085907,
      "learning_rate": 3.466542750929368e-05,
      "loss": 0.2676,
      "step": 926
    },
    {
      "epoch": 3.326750448833034,
      "grad_norm": 0.23663635551929474,
      "learning_rate": 3.464684014869889e-05,
      "loss": 0.3467,
      "step": 927
    },
    {
      "epoch": 3.3303411131059244,
      "grad_norm": 0.2416193038225174,
      "learning_rate": 3.462825278810409e-05,
      "loss": 0.2998,
      "step": 928
    },
    {
      "epoch": 3.3339317773788153,
      "grad_norm": 0.255321204662323,
      "learning_rate": 3.46096654275093e-05,
      "loss": 0.4317,
      "step": 929
    },
    {
      "epoch": 3.3375224416517058,
      "grad_norm": 0.22693777084350586,
      "learning_rate": 3.45910780669145e-05,
      "loss": 0.4671,
      "step": 930
    },
    {
      "epoch": 3.341113105924596,
      "grad_norm": 0.22795173525810242,
      "learning_rate": 3.45724907063197e-05,
      "loss": 0.2429,
      "step": 931
    },
    {
      "epoch": 3.3447037701974867,
      "grad_norm": 0.23681585490703583,
      "learning_rate": 3.455390334572491e-05,
      "loss": 0.2568,
      "step": 932
    },
    {
      "epoch": 3.348294434470377,
      "grad_norm": 0.22826191782951355,
      "learning_rate": 3.453531598513011e-05,
      "loss": 0.2077,
      "step": 933
    },
    {
      "epoch": 3.3518850987432676,
      "grad_norm": 0.2275114804506302,
      "learning_rate": 3.451672862453531e-05,
      "loss": 0.4311,
      "step": 934
    },
    {
      "epoch": 3.355475763016158,
      "grad_norm": 0.21391969919204712,
      "learning_rate": 3.449814126394052e-05,
      "loss": 0.3278,
      "step": 935
    },
    {
      "epoch": 3.3590664272890485,
      "grad_norm": 0.23450447618961334,
      "learning_rate": 3.447955390334573e-05,
      "loss": 0.2262,
      "step": 936
    },
    {
      "epoch": 3.362657091561939,
      "grad_norm": 0.246677964925766,
      "learning_rate": 3.446096654275093e-05,
      "loss": 0.4341,
      "step": 937
    },
    {
      "epoch": 3.3662477558348294,
      "grad_norm": 0.22753877937793732,
      "learning_rate": 3.444237918215614e-05,
      "loss": 0.732,
      "step": 938
    },
    {
      "epoch": 3.36983842010772,
      "grad_norm": 0.2572675347328186,
      "learning_rate": 3.442379182156134e-05,
      "loss": 0.3436,
      "step": 939
    },
    {
      "epoch": 3.3734290843806103,
      "grad_norm": 0.24647191166877747,
      "learning_rate": 3.440520446096654e-05,
      "loss": 0.2256,
      "step": 940
    },
    {
      "epoch": 3.3770197486535007,
      "grad_norm": 0.21687743067741394,
      "learning_rate": 3.438661710037175e-05,
      "loss": 0.2743,
      "step": 941
    },
    {
      "epoch": 3.380610412926391,
      "grad_norm": 0.23255589604377747,
      "learning_rate": 3.436802973977695e-05,
      "loss": 0.2305,
      "step": 942
    },
    {
      "epoch": 3.3842010771992816,
      "grad_norm": 0.2580336630344391,
      "learning_rate": 3.4349442379182154e-05,
      "loss": 0.4032,
      "step": 943
    },
    {
      "epoch": 3.3877917414721725,
      "grad_norm": 0.2315930426120758,
      "learning_rate": 3.433085501858736e-05,
      "loss": 0.4615,
      "step": 944
    },
    {
      "epoch": 3.391382405745063,
      "grad_norm": 0.23232048749923706,
      "learning_rate": 3.4312267657992564e-05,
      "loss": 0.3705,
      "step": 945
    },
    {
      "epoch": 3.3949730700179535,
      "grad_norm": 0.21706070005893707,
      "learning_rate": 3.429368029739777e-05,
      "loss": 0.4655,
      "step": 946
    },
    {
      "epoch": 3.398563734290844,
      "grad_norm": 0.23911674320697784,
      "learning_rate": 3.4275092936802974e-05,
      "loss": 0.2492,
      "step": 947
    },
    {
      "epoch": 3.4021543985637344,
      "grad_norm": 0.23120921850204468,
      "learning_rate": 3.425650557620818e-05,
      "loss": 0.4147,
      "step": 948
    },
    {
      "epoch": 3.405745062836625,
      "grad_norm": 0.2566313147544861,
      "learning_rate": 3.4237918215613384e-05,
      "loss": 0.2027,
      "step": 949
    },
    {
      "epoch": 3.4093357271095153,
      "grad_norm": 0.24153225123882294,
      "learning_rate": 3.4219330855018586e-05,
      "loss": 0.3285,
      "step": 950
    },
    {
      "epoch": 3.4129263913824057,
      "grad_norm": 0.24654339253902435,
      "learning_rate": 3.4200743494423794e-05,
      "loss": 0.2748,
      "step": 951
    },
    {
      "epoch": 3.416517055655296,
      "grad_norm": 0.23755230009555817,
      "learning_rate": 3.4182156133828996e-05,
      "loss": 0.3967,
      "step": 952
    },
    {
      "epoch": 3.4201077199281866,
      "grad_norm": 0.2590872645378113,
      "learning_rate": 3.41635687732342e-05,
      "loss": 0.2604,
      "step": 953
    },
    {
      "epoch": 3.423698384201077,
      "grad_norm": 0.20745839178562164,
      "learning_rate": 3.4144981412639406e-05,
      "loss": 0.2448,
      "step": 954
    },
    {
      "epoch": 3.4272890484739675,
      "grad_norm": 0.2287701815366745,
      "learning_rate": 3.4126394052044614e-05,
      "loss": 0.2367,
      "step": 955
    },
    {
      "epoch": 3.430879712746858,
      "grad_norm": 0.2382553517818451,
      "learning_rate": 3.4107806691449816e-05,
      "loss": 0.4651,
      "step": 956
    },
    {
      "epoch": 3.434470377019749,
      "grad_norm": 0.24780866503715515,
      "learning_rate": 3.4089219330855024e-05,
      "loss": 0.5236,
      "step": 957
    },
    {
      "epoch": 3.438061041292639,
      "grad_norm": 0.21977148950099945,
      "learning_rate": 3.4070631970260226e-05,
      "loss": 0.2111,
      "step": 958
    },
    {
      "epoch": 3.44165170556553,
      "grad_norm": 0.2576964199542999,
      "learning_rate": 3.405204460966543e-05,
      "loss": 0.5021,
      "step": 959
    },
    {
      "epoch": 3.4452423698384202,
      "grad_norm": 0.21948488056659698,
      "learning_rate": 3.4033457249070636e-05,
      "loss": 0.203,
      "step": 960
    },
    {
      "epoch": 3.4488330341113107,
      "grad_norm": 0.22087602317333221,
      "learning_rate": 3.401486988847584e-05,
      "loss": 0.301,
      "step": 961
    },
    {
      "epoch": 3.452423698384201,
      "grad_norm": 0.2240106463432312,
      "learning_rate": 3.399628252788104e-05,
      "loss": 0.2134,
      "step": 962
    },
    {
      "epoch": 3.4560143626570916,
      "grad_norm": 0.2634586691856384,
      "learning_rate": 3.397769516728625e-05,
      "loss": 0.404,
      "step": 963
    },
    {
      "epoch": 3.459605026929982,
      "grad_norm": 0.27492567896842957,
      "learning_rate": 3.3959107806691455e-05,
      "loss": 0.5943,
      "step": 964
    },
    {
      "epoch": 3.4631956912028725,
      "grad_norm": 0.23831108212471008,
      "learning_rate": 3.394052044609666e-05,
      "loss": 0.3551,
      "step": 965
    },
    {
      "epoch": 3.466786355475763,
      "grad_norm": 0.2484758347272873,
      "learning_rate": 3.392193308550186e-05,
      "loss": 0.1918,
      "step": 966
    },
    {
      "epoch": 3.4703770197486534,
      "grad_norm": 0.2530819773674011,
      "learning_rate": 3.390334572490707e-05,
      "loss": 0.2201,
      "step": 967
    },
    {
      "epoch": 3.473967684021544,
      "grad_norm": 0.3028958737850189,
      "learning_rate": 3.388475836431227e-05,
      "loss": 0.5957,
      "step": 968
    },
    {
      "epoch": 3.4775583482944343,
      "grad_norm": 0.2843305468559265,
      "learning_rate": 3.386617100371747e-05,
      "loss": 0.3329,
      "step": 969
    },
    {
      "epoch": 3.481149012567325,
      "grad_norm": 0.24480485916137695,
      "learning_rate": 3.384758364312268e-05,
      "loss": 0.5034,
      "step": 970
    },
    {
      "epoch": 3.4847396768402152,
      "grad_norm": 0.23834842443466187,
      "learning_rate": 3.382899628252788e-05,
      "loss": 0.2247,
      "step": 971
    },
    {
      "epoch": 3.488330341113106,
      "grad_norm": 0.2566714584827423,
      "learning_rate": 3.381040892193308e-05,
      "loss": 0.3893,
      "step": 972
    },
    {
      "epoch": 3.4919210053859966,
      "grad_norm": 0.2754567861557007,
      "learning_rate": 3.37918215613383e-05,
      "loss": 0.2472,
      "step": 973
    },
    {
      "epoch": 3.495511669658887,
      "grad_norm": 0.24323149025440216,
      "learning_rate": 3.37732342007435e-05,
      "loss": 0.3101,
      "step": 974
    },
    {
      "epoch": 3.4991023339317775,
      "grad_norm": 0.22556358575820923,
      "learning_rate": 3.37546468401487e-05,
      "loss": 0.1926,
      "step": 975
    },
    {
      "epoch": 3.502692998204668,
      "grad_norm": 0.2004850208759308,
      "learning_rate": 3.373605947955391e-05,
      "loss": 0.2994,
      "step": 976
    },
    {
      "epoch": 3.5062836624775584,
      "grad_norm": 0.2524142563343048,
      "learning_rate": 3.371747211895911e-05,
      "loss": 0.247,
      "step": 977
    },
    {
      "epoch": 3.509874326750449,
      "grad_norm": 0.27899715304374695,
      "learning_rate": 3.369888475836431e-05,
      "loss": 0.5251,
      "step": 978
    },
    {
      "epoch": 3.5134649910233393,
      "grad_norm": 0.27834099531173706,
      "learning_rate": 3.368029739776952e-05,
      "loss": 0.2246,
      "step": 979
    },
    {
      "epoch": 3.5170556552962298,
      "grad_norm": 0.2637667953968048,
      "learning_rate": 3.366171003717472e-05,
      "loss": 0.6623,
      "step": 980
    },
    {
      "epoch": 3.52064631956912,
      "grad_norm": 0.28303778171539307,
      "learning_rate": 3.364312267657992e-05,
      "loss": 0.7971,
      "step": 981
    },
    {
      "epoch": 3.5242369838420107,
      "grad_norm": 0.22411303222179413,
      "learning_rate": 3.362453531598513e-05,
      "loss": 0.352,
      "step": 982
    },
    {
      "epoch": 3.527827648114901,
      "grad_norm": 0.25557249784469604,
      "learning_rate": 3.360594795539034e-05,
      "loss": 0.2443,
      "step": 983
    },
    {
      "epoch": 3.5314183123877916,
      "grad_norm": 0.25239795446395874,
      "learning_rate": 3.358736059479554e-05,
      "loss": 0.5371,
      "step": 984
    },
    {
      "epoch": 3.5350089766606825,
      "grad_norm": 0.228390634059906,
      "learning_rate": 3.356877323420074e-05,
      "loss": 0.2234,
      "step": 985
    },
    {
      "epoch": 3.5385996409335725,
      "grad_norm": 0.2815796136856079,
      "learning_rate": 3.355018587360595e-05,
      "loss": 0.2807,
      "step": 986
    },
    {
      "epoch": 3.5421903052064634,
      "grad_norm": 0.25439542531967163,
      "learning_rate": 3.353159851301115e-05,
      "loss": 0.3913,
      "step": 987
    },
    {
      "epoch": 3.545780969479354,
      "grad_norm": 0.26104992628097534,
      "learning_rate": 3.3513011152416355e-05,
      "loss": 0.3039,
      "step": 988
    },
    {
      "epoch": 3.5493716337522443,
      "grad_norm": 0.23205602169036865,
      "learning_rate": 3.349442379182156e-05,
      "loss": 0.2315,
      "step": 989
    },
    {
      "epoch": 3.5529622980251347,
      "grad_norm": 0.25087153911590576,
      "learning_rate": 3.3475836431226765e-05,
      "loss": 0.1839,
      "step": 990
    },
    {
      "epoch": 3.556552962298025,
      "grad_norm": 0.25209444761276245,
      "learning_rate": 3.345724907063197e-05,
      "loss": 0.2134,
      "step": 991
    },
    {
      "epoch": 3.5601436265709157,
      "grad_norm": 0.23739683628082275,
      "learning_rate": 3.343866171003718e-05,
      "loss": 0.2083,
      "step": 992
    },
    {
      "epoch": 3.563734290843806,
      "grad_norm": 0.21525926887989044,
      "learning_rate": 3.342007434944238e-05,
      "loss": 0.3765,
      "step": 993
    },
    {
      "epoch": 3.5673249551166966,
      "grad_norm": 0.25445452332496643,
      "learning_rate": 3.3401486988847584e-05,
      "loss": 0.4483,
      "step": 994
    },
    {
      "epoch": 3.570915619389587,
      "grad_norm": 0.23755180835723877,
      "learning_rate": 3.338289962825279e-05,
      "loss": 0.2349,
      "step": 995
    },
    {
      "epoch": 3.5745062836624775,
      "grad_norm": 0.255470335483551,
      "learning_rate": 3.3364312267657994e-05,
      "loss": 0.1988,
      "step": 996
    },
    {
      "epoch": 3.578096947935368,
      "grad_norm": 0.22067584097385406,
      "learning_rate": 3.3345724907063196e-05,
      "loss": 0.51,
      "step": 997
    },
    {
      "epoch": 3.5816876122082584,
      "grad_norm": 0.2467593103647232,
      "learning_rate": 3.3327137546468404e-05,
      "loss": 0.2701,
      "step": 998
    },
    {
      "epoch": 3.585278276481149,
      "grad_norm": 0.2329201102256775,
      "learning_rate": 3.3308550185873606e-05,
      "loss": 0.2606,
      "step": 999
    },
    {
      "epoch": 3.5888689407540397,
      "grad_norm": 0.24074877798557281,
      "learning_rate": 3.3289962825278814e-05,
      "loss": 0.2207,
      "step": 1000
    },
    {
      "epoch": 3.5924596050269297,
      "grad_norm": 0.25259992480278015,
      "learning_rate": 3.3271375464684016e-05,
      "loss": 0.2594,
      "step": 1001
    },
    {
      "epoch": 3.5960502692998206,
      "grad_norm": 0.2501515746116638,
      "learning_rate": 3.3252788104089224e-05,
      "loss": 0.5372,
      "step": 1002
    },
    {
      "epoch": 3.599640933572711,
      "grad_norm": 0.24933889508247375,
      "learning_rate": 3.3234200743494426e-05,
      "loss": 0.29,
      "step": 1003
    },
    {
      "epoch": 3.6032315978456015,
      "grad_norm": 0.27657684683799744,
      "learning_rate": 3.321561338289963e-05,
      "loss": 0.2718,
      "step": 1004
    },
    {
      "epoch": 3.606822262118492,
      "grad_norm": 0.24257688224315643,
      "learning_rate": 3.3197026022304836e-05,
      "loss": 0.3187,
      "step": 1005
    },
    {
      "epoch": 3.6104129263913824,
      "grad_norm": 0.23851336538791656,
      "learning_rate": 3.317843866171004e-05,
      "loss": 0.4939,
      "step": 1006
    },
    {
      "epoch": 3.614003590664273,
      "grad_norm": 0.24251188337802887,
      "learning_rate": 3.315985130111524e-05,
      "loss": 0.3638,
      "step": 1007
    },
    {
      "epoch": 3.6175942549371634,
      "grad_norm": 0.267450213432312,
      "learning_rate": 3.314126394052045e-05,
      "loss": 0.3122,
      "step": 1008
    },
    {
      "epoch": 3.621184919210054,
      "grad_norm": 0.23413099348545074,
      "learning_rate": 3.3122676579925656e-05,
      "loss": 0.2391,
      "step": 1009
    },
    {
      "epoch": 3.6247755834829443,
      "grad_norm": 0.25738203525543213,
      "learning_rate": 3.310408921933086e-05,
      "loss": 0.215,
      "step": 1010
    },
    {
      "epoch": 3.6283662477558347,
      "grad_norm": 0.2884233593940735,
      "learning_rate": 3.308550185873606e-05,
      "loss": 0.3707,
      "step": 1011
    },
    {
      "epoch": 3.631956912028725,
      "grad_norm": 0.2453780323266983,
      "learning_rate": 3.306691449814127e-05,
      "loss": 0.2271,
      "step": 1012
    },
    {
      "epoch": 3.635547576301616,
      "grad_norm": 0.24752557277679443,
      "learning_rate": 3.304832713754647e-05,
      "loss": 0.3123,
      "step": 1013
    },
    {
      "epoch": 3.639138240574506,
      "grad_norm": 0.245054692029953,
      "learning_rate": 3.302973977695167e-05,
      "loss": 0.2948,
      "step": 1014
    },
    {
      "epoch": 3.642728904847397,
      "grad_norm": 0.23093979060649872,
      "learning_rate": 3.301115241635688e-05,
      "loss": 0.3342,
      "step": 1015
    },
    {
      "epoch": 3.646319569120287,
      "grad_norm": 0.2407323122024536,
      "learning_rate": 3.299256505576208e-05,
      "loss": 0.3776,
      "step": 1016
    },
    {
      "epoch": 3.649910233393178,
      "grad_norm": 0.2845363914966583,
      "learning_rate": 3.297397769516728e-05,
      "loss": 0.5165,
      "step": 1017
    },
    {
      "epoch": 3.6535008976660683,
      "grad_norm": 0.23844368755817413,
      "learning_rate": 3.29553903345725e-05,
      "loss": 0.4028,
      "step": 1018
    },
    {
      "epoch": 3.657091561938959,
      "grad_norm": 0.24652452766895294,
      "learning_rate": 3.29368029739777e-05,
      "loss": 0.3741,
      "step": 1019
    },
    {
      "epoch": 3.6606822262118492,
      "grad_norm": 0.29566073417663574,
      "learning_rate": 3.29182156133829e-05,
      "loss": 0.3344,
      "step": 1020
    },
    {
      "epoch": 3.6642728904847397,
      "grad_norm": 0.18847578763961792,
      "learning_rate": 3.289962825278811e-05,
      "loss": 0.279,
      "step": 1021
    },
    {
      "epoch": 3.66786355475763,
      "grad_norm": 0.2534039616584778,
      "learning_rate": 3.288104089219331e-05,
      "loss": 0.4598,
      "step": 1022
    },
    {
      "epoch": 3.6714542190305206,
      "grad_norm": 0.23907436430454254,
      "learning_rate": 3.286245353159851e-05,
      "loss": 0.1843,
      "step": 1023
    },
    {
      "epoch": 3.675044883303411,
      "grad_norm": 0.2514495253562927,
      "learning_rate": 3.284386617100372e-05,
      "loss": 0.251,
      "step": 1024
    },
    {
      "epoch": 3.6786355475763015,
      "grad_norm": 0.25034528970718384,
      "learning_rate": 3.282527881040892e-05,
      "loss": 0.2927,
      "step": 1025
    },
    {
      "epoch": 3.682226211849192,
      "grad_norm": 0.2477673441171646,
      "learning_rate": 3.2806691449814123e-05,
      "loss": 0.4727,
      "step": 1026
    },
    {
      "epoch": 3.6858168761220824,
      "grad_norm": 0.2766311466693878,
      "learning_rate": 3.278810408921933e-05,
      "loss": 0.4761,
      "step": 1027
    },
    {
      "epoch": 3.6894075403949733,
      "grad_norm": 0.23413556814193726,
      "learning_rate": 3.276951672862454e-05,
      "loss": 0.2589,
      "step": 1028
    },
    {
      "epoch": 3.6929982046678633,
      "grad_norm": 0.24681393802165985,
      "learning_rate": 3.275092936802974e-05,
      "loss": 0.5658,
      "step": 1029
    },
    {
      "epoch": 3.6965888689407542,
      "grad_norm": 0.22542937099933624,
      "learning_rate": 3.273234200743494e-05,
      "loss": 0.2326,
      "step": 1030
    },
    {
      "epoch": 3.7001795332136447,
      "grad_norm": 0.2477526217699051,
      "learning_rate": 3.271375464684015e-05,
      "loss": 0.2883,
      "step": 1031
    },
    {
      "epoch": 3.703770197486535,
      "grad_norm": 0.23465923964977264,
      "learning_rate": 3.269516728624535e-05,
      "loss": 0.245,
      "step": 1032
    },
    {
      "epoch": 3.7073608617594256,
      "grad_norm": 0.23174822330474854,
      "learning_rate": 3.2676579925650555e-05,
      "loss": 0.1931,
      "step": 1033
    },
    {
      "epoch": 3.710951526032316,
      "grad_norm": 0.24177594482898712,
      "learning_rate": 3.265799256505576e-05,
      "loss": 0.2175,
      "step": 1034
    },
    {
      "epoch": 3.7145421903052065,
      "grad_norm": 0.2057540863752365,
      "learning_rate": 3.2639405204460965e-05,
      "loss": 0.2581,
      "step": 1035
    },
    {
      "epoch": 3.718132854578097,
      "grad_norm": 0.32663190364837646,
      "learning_rate": 3.262081784386617e-05,
      "loss": 0.4311,
      "step": 1036
    },
    {
      "epoch": 3.7217235188509874,
      "grad_norm": 0.22969475388526917,
      "learning_rate": 3.260223048327138e-05,
      "loss": 0.2089,
      "step": 1037
    },
    {
      "epoch": 3.725314183123878,
      "grad_norm": 0.24128836393356323,
      "learning_rate": 3.258364312267658e-05,
      "loss": 0.1886,
      "step": 1038
    },
    {
      "epoch": 3.7289048473967683,
      "grad_norm": 0.26187437772750854,
      "learning_rate": 3.2565055762081785e-05,
      "loss": 0.4489,
      "step": 1039
    },
    {
      "epoch": 3.7324955116696588,
      "grad_norm": 0.2626861035823822,
      "learning_rate": 3.254646840148699e-05,
      "loss": 0.5041,
      "step": 1040
    },
    {
      "epoch": 3.736086175942549,
      "grad_norm": 0.23693925142288208,
      "learning_rate": 3.2527881040892195e-05,
      "loss": 0.2502,
      "step": 1041
    },
    {
      "epoch": 3.7396768402154397,
      "grad_norm": 0.24844014644622803,
      "learning_rate": 3.2509293680297396e-05,
      "loss": 0.3667,
      "step": 1042
    },
    {
      "epoch": 3.7432675044883306,
      "grad_norm": 0.2781262993812561,
      "learning_rate": 3.2490706319702605e-05,
      "loss": 0.4423,
      "step": 1043
    },
    {
      "epoch": 3.7468581687612206,
      "grad_norm": 0.20927174389362335,
      "learning_rate": 3.2472118959107806e-05,
      "loss": 0.2852,
      "step": 1044
    },
    {
      "epoch": 3.7504488330341115,
      "grad_norm": 0.25317448377609253,
      "learning_rate": 3.2453531598513015e-05,
      "loss": 0.2652,
      "step": 1045
    },
    {
      "epoch": 3.754039497307002,
      "grad_norm": 0.23749756813049316,
      "learning_rate": 3.2434944237918216e-05,
      "loss": 0.2341,
      "step": 1046
    },
    {
      "epoch": 3.7576301615798924,
      "grad_norm": 0.2301895171403885,
      "learning_rate": 3.2416356877323425e-05,
      "loss": 0.1823,
      "step": 1047
    },
    {
      "epoch": 3.761220825852783,
      "grad_norm": 0.2600250244140625,
      "learning_rate": 3.2397769516728626e-05,
      "loss": 0.3152,
      "step": 1048
    },
    {
      "epoch": 3.7648114901256733,
      "grad_norm": 0.26873940229415894,
      "learning_rate": 3.237918215613383e-05,
      "loss": 0.2446,
      "step": 1049
    },
    {
      "epoch": 3.7684021543985637,
      "grad_norm": 0.2349143624305725,
      "learning_rate": 3.2360594795539036e-05,
      "loss": 0.4216,
      "step": 1050
    },
    {
      "epoch": 3.771992818671454,
      "grad_norm": 0.2713150978088379,
      "learning_rate": 3.234200743494424e-05,
      "loss": 0.2499,
      "step": 1051
    },
    {
      "epoch": 3.7755834829443446,
      "grad_norm": 0.2645110487937927,
      "learning_rate": 3.232342007434944e-05,
      "loss": 0.2679,
      "step": 1052
    },
    {
      "epoch": 3.779174147217235,
      "grad_norm": 0.24026186764240265,
      "learning_rate": 3.230483271375465e-05,
      "loss": 0.3406,
      "step": 1053
    },
    {
      "epoch": 3.7827648114901256,
      "grad_norm": 0.2962729036808014,
      "learning_rate": 3.2286245353159856e-05,
      "loss": 0.542,
      "step": 1054
    },
    {
      "epoch": 3.786355475763016,
      "grad_norm": 0.23875300586223602,
      "learning_rate": 3.226765799256506e-05,
      "loss": 0.2664,
      "step": 1055
    },
    {
      "epoch": 3.789946140035907,
      "grad_norm": 0.28504592180252075,
      "learning_rate": 3.2249070631970266e-05,
      "loss": 0.3049,
      "step": 1056
    },
    {
      "epoch": 3.793536804308797,
      "grad_norm": 0.2641824185848236,
      "learning_rate": 3.223048327137547e-05,
      "loss": 0.5037,
      "step": 1057
    },
    {
      "epoch": 3.797127468581688,
      "grad_norm": 0.25814446806907654,
      "learning_rate": 3.221189591078067e-05,
      "loss": 0.3161,
      "step": 1058
    },
    {
      "epoch": 3.800718132854578,
      "grad_norm": 0.2324041724205017,
      "learning_rate": 3.219330855018588e-05,
      "loss": 0.198,
      "step": 1059
    },
    {
      "epoch": 3.8043087971274687,
      "grad_norm": 0.24683985114097595,
      "learning_rate": 3.217472118959108e-05,
      "loss": 0.2922,
      "step": 1060
    },
    {
      "epoch": 3.807899461400359,
      "grad_norm": 0.25921210646629333,
      "learning_rate": 3.215613382899628e-05,
      "loss": 0.5951,
      "step": 1061
    },
    {
      "epoch": 3.8114901256732496,
      "grad_norm": 0.26796332001686096,
      "learning_rate": 3.213754646840149e-05,
      "loss": 0.2929,
      "step": 1062
    },
    {
      "epoch": 3.81508078994614,
      "grad_norm": 0.23859286308288574,
      "learning_rate": 3.21189591078067e-05,
      "loss": 0.241,
      "step": 1063
    },
    {
      "epoch": 3.8186714542190305,
      "grad_norm": 0.24994225800037384,
      "learning_rate": 3.21003717472119e-05,
      "loss": 0.5845,
      "step": 1064
    },
    {
      "epoch": 3.822262118491921,
      "grad_norm": 0.20379087328910828,
      "learning_rate": 3.20817843866171e-05,
      "loss": 0.2833,
      "step": 1065
    },
    {
      "epoch": 3.8258527827648114,
      "grad_norm": 0.22900806367397308,
      "learning_rate": 3.206319702602231e-05,
      "loss": 0.245,
      "step": 1066
    },
    {
      "epoch": 3.829443447037702,
      "grad_norm": 0.24416114389896393,
      "learning_rate": 3.204460966542751e-05,
      "loss": 0.2379,
      "step": 1067
    },
    {
      "epoch": 3.8330341113105924,
      "grad_norm": 0.25771522521972656,
      "learning_rate": 3.202602230483271e-05,
      "loss": 0.2555,
      "step": 1068
    },
    {
      "epoch": 3.836624775583483,
      "grad_norm": 0.1971089094877243,
      "learning_rate": 3.200743494423792e-05,
      "loss": 0.4228,
      "step": 1069
    },
    {
      "epoch": 3.8402154398563733,
      "grad_norm": 0.23408563435077667,
      "learning_rate": 3.198884758364312e-05,
      "loss": 0.2053,
      "step": 1070
    },
    {
      "epoch": 3.843806104129264,
      "grad_norm": 0.2520309388637543,
      "learning_rate": 3.1970260223048324e-05,
      "loss": 0.262,
      "step": 1071
    },
    {
      "epoch": 3.847396768402154,
      "grad_norm": 0.22477324306964874,
      "learning_rate": 3.195167286245353e-05,
      "loss": 0.3401,
      "step": 1072
    },
    {
      "epoch": 3.850987432675045,
      "grad_norm": 0.25791388750076294,
      "learning_rate": 3.193308550185874e-05,
      "loss": 0.5192,
      "step": 1073
    },
    {
      "epoch": 3.8545780969479355,
      "grad_norm": 0.2490009069442749,
      "learning_rate": 3.191449814126394e-05,
      "loss": 0.4479,
      "step": 1074
    },
    {
      "epoch": 3.858168761220826,
      "grad_norm": 0.25246310234069824,
      "learning_rate": 3.1895910780669144e-05,
      "loss": 0.4275,
      "step": 1075
    },
    {
      "epoch": 3.8617594254937164,
      "grad_norm": 0.18956060707569122,
      "learning_rate": 3.187732342007435e-05,
      "loss": 0.3412,
      "step": 1076
    },
    {
      "epoch": 3.865350089766607,
      "grad_norm": 0.24956631660461426,
      "learning_rate": 3.1858736059479554e-05,
      "loss": 0.2458,
      "step": 1077
    },
    {
      "epoch": 3.8689407540394973,
      "grad_norm": 0.23087207973003387,
      "learning_rate": 3.1840148698884755e-05,
      "loss": 0.3597,
      "step": 1078
    },
    {
      "epoch": 3.872531418312388,
      "grad_norm": 0.2217182219028473,
      "learning_rate": 3.1821561338289964e-05,
      "loss": 0.4418,
      "step": 1079
    },
    {
      "epoch": 3.8761220825852782,
      "grad_norm": 0.22124174237251282,
      "learning_rate": 3.1802973977695165e-05,
      "loss": 0.398,
      "step": 1080
    },
    {
      "epoch": 3.8797127468581687,
      "grad_norm": 0.22706274688243866,
      "learning_rate": 3.1784386617100374e-05,
      "loss": 0.4434,
      "step": 1081
    },
    {
      "epoch": 3.883303411131059,
      "grad_norm": 0.21429337561130524,
      "learning_rate": 3.176579925650558e-05,
      "loss": 0.3234,
      "step": 1082
    },
    {
      "epoch": 3.8868940754039496,
      "grad_norm": 0.19179020822048187,
      "learning_rate": 3.1747211895910784e-05,
      "loss": 0.2786,
      "step": 1083
    },
    {
      "epoch": 3.89048473967684,
      "grad_norm": 0.2352266013622284,
      "learning_rate": 3.1728624535315985e-05,
      "loss": 0.1927,
      "step": 1084
    },
    {
      "epoch": 3.8940754039497305,
      "grad_norm": 0.27387452125549316,
      "learning_rate": 3.1710037174721194e-05,
      "loss": 0.3132,
      "step": 1085
    },
    {
      "epoch": 3.8976660682226214,
      "grad_norm": 0.24403011798858643,
      "learning_rate": 3.1691449814126395e-05,
      "loss": 0.5861,
      "step": 1086
    },
    {
      "epoch": 3.9012567324955114,
      "grad_norm": 0.26065370440483093,
      "learning_rate": 3.16728624535316e-05,
      "loss": 0.3326,
      "step": 1087
    },
    {
      "epoch": 3.9048473967684023,
      "grad_norm": 0.2281435877084732,
      "learning_rate": 3.1654275092936805e-05,
      "loss": 0.458,
      "step": 1088
    },
    {
      "epoch": 3.9084380610412928,
      "grad_norm": 0.20710134506225586,
      "learning_rate": 3.163568773234201e-05,
      "loss": 0.302,
      "step": 1089
    },
    {
      "epoch": 3.912028725314183,
      "grad_norm": 0.24037006497383118,
      "learning_rate": 3.1617100371747215e-05,
      "loss": 0.2284,
      "step": 1090
    },
    {
      "epoch": 3.9156193895870737,
      "grad_norm": 0.24021536111831665,
      "learning_rate": 3.1598513011152417e-05,
      "loss": 0.2184,
      "step": 1091
    },
    {
      "epoch": 3.919210053859964,
      "grad_norm": 0.32284629344940186,
      "learning_rate": 3.1579925650557625e-05,
      "loss": 0.3859,
      "step": 1092
    },
    {
      "epoch": 3.9228007181328546,
      "grad_norm": 0.2614433467388153,
      "learning_rate": 3.1561338289962827e-05,
      "loss": 0.3873,
      "step": 1093
    },
    {
      "epoch": 3.926391382405745,
      "grad_norm": 0.21848826110363007,
      "learning_rate": 3.154275092936803e-05,
      "loss": 0.1962,
      "step": 1094
    },
    {
      "epoch": 3.9299820466786355,
      "grad_norm": 0.23821566998958588,
      "learning_rate": 3.1524163568773237e-05,
      "loss": 0.3515,
      "step": 1095
    },
    {
      "epoch": 3.933572710951526,
      "grad_norm": 0.23001305758953094,
      "learning_rate": 3.150557620817844e-05,
      "loss": 0.2479,
      "step": 1096
    },
    {
      "epoch": 3.9371633752244164,
      "grad_norm": 0.23379534482955933,
      "learning_rate": 3.148698884758364e-05,
      "loss": 0.2931,
      "step": 1097
    },
    {
      "epoch": 3.940754039497307,
      "grad_norm": 0.25249776244163513,
      "learning_rate": 3.146840148698885e-05,
      "loss": 0.3169,
      "step": 1098
    },
    {
      "epoch": 3.9443447037701977,
      "grad_norm": 0.2559717893600464,
      "learning_rate": 3.1449814126394056e-05,
      "loss": 0.5104,
      "step": 1099
    },
    {
      "epoch": 3.9479353680430878,
      "grad_norm": 0.29940029978752136,
      "learning_rate": 3.143122676579926e-05,
      "loss": 0.4945,
      "step": 1100
    },
    {
      "epoch": 3.9515260323159787,
      "grad_norm": 0.2559410631656647,
      "learning_rate": 3.1412639405204466e-05,
      "loss": 0.5018,
      "step": 1101
    },
    {
      "epoch": 3.9551166965888687,
      "grad_norm": 0.24963879585266113,
      "learning_rate": 3.139405204460967e-05,
      "loss": 0.4066,
      "step": 1102
    },
    {
      "epoch": 3.9587073608617596,
      "grad_norm": 0.2834954857826233,
      "learning_rate": 3.137546468401487e-05,
      "loss": 0.4732,
      "step": 1103
    },
    {
      "epoch": 3.96229802513465,
      "grad_norm": 0.23344196379184723,
      "learning_rate": 3.135687732342008e-05,
      "loss": 0.495,
      "step": 1104
    },
    {
      "epoch": 3.9658886894075405,
      "grad_norm": 0.1778457760810852,
      "learning_rate": 3.133828996282528e-05,
      "loss": 0.4511,
      "step": 1105
    },
    {
      "epoch": 3.969479353680431,
      "grad_norm": 0.19722320139408112,
      "learning_rate": 3.131970260223048e-05,
      "loss": 0.2799,
      "step": 1106
    },
    {
      "epoch": 3.9730700179533214,
      "grad_norm": 0.2674300968647003,
      "learning_rate": 3.130111524163569e-05,
      "loss": 0.2274,
      "step": 1107
    },
    {
      "epoch": 3.976660682226212,
      "grad_norm": 0.24930217862129211,
      "learning_rate": 3.12825278810409e-05,
      "loss": 0.2962,
      "step": 1108
    },
    {
      "epoch": 3.9802513464991023,
      "grad_norm": 0.21329018473625183,
      "learning_rate": 3.12639405204461e-05,
      "loss": 0.4447,
      "step": 1109
    },
    {
      "epoch": 3.9838420107719927,
      "grad_norm": 0.2399543821811676,
      "learning_rate": 3.12453531598513e-05,
      "loss": 0.4666,
      "step": 1110
    },
    {
      "epoch": 3.987432675044883,
      "grad_norm": 0.28229373693466187,
      "learning_rate": 3.122676579925651e-05,
      "loss": 0.3515,
      "step": 1111
    },
    {
      "epoch": 3.9910233393177736,
      "grad_norm": 0.19896718859672546,
      "learning_rate": 3.120817843866171e-05,
      "loss": 0.294,
      "step": 1112
    },
    {
      "epoch": 3.994614003590664,
      "grad_norm": 0.24525916576385498,
      "learning_rate": 3.118959107806691e-05,
      "loss": 0.2526,
      "step": 1113
    },
    {
      "epoch": 3.998204667863555,
      "grad_norm": 0.26347070932388306,
      "learning_rate": 3.117100371747212e-05,
      "loss": 0.3175,
      "step": 1114
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3968011736869812,
      "learning_rate": 3.115241635687732e-05,
      "loss": 0.1702,
      "step": 1115
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.24459858238697052,
      "eval_runtime": 4.162,
      "eval_samples_per_second": 1.201,
      "eval_steps_per_second": 1.201,
      "step": 1115
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.5691695213317871,
      "eval_runtime": 3.2192,
      "eval_samples_per_second": 1.553,
      "eval_steps_per_second": 1.553,
      "step": 1115
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.3336179852485657,
      "eval_runtime": 4.7271,
      "eval_samples_per_second": 1.058,
      "eval_steps_per_second": 1.058,
      "step": 1115
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.9284563064575195,
      "eval_runtime": 4.953,
      "eval_samples_per_second": 1.009,
      "eval_steps_per_second": 1.009,
      "step": 1115
    }
  ],
  "logging_steps": 1,
  "max_steps": 2790,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.979998459821656e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}