{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.333868378812198,
  "eval_steps": 500,
  "global_step": 1300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006420545746388443,
      "grad_norm": 2768.0,
      "learning_rate": 0.0,
      "loss": 67.2236,
      "step": 1
    },
    {
      "epoch": 0.012841091492776886,
      "grad_norm": 2720.0,
      "learning_rate": 1e-08,
      "loss": 66.1013,
      "step": 2
    },
    {
      "epoch": 0.019261637239165328,
      "grad_norm": 2384.0,
      "learning_rate": 2e-08,
      "loss": 63.5176,
      "step": 3
    },
    {
      "epoch": 0.025682182985553772,
      "grad_norm": 2544.0,
      "learning_rate": 3e-08,
      "loss": 66.3939,
      "step": 4
    },
    {
      "epoch": 0.03210272873194221,
      "grad_norm": 2992.0,
      "learning_rate": 4e-08,
      "loss": 67.2986,
      "step": 5
    },
    {
      "epoch": 0.038523274478330656,
      "grad_norm": 2752.0,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 68.0958,
      "step": 6
    },
    {
      "epoch": 0.0449438202247191,
      "grad_norm": 2720.0,
      "learning_rate": 6e-08,
      "loss": 66.5096,
      "step": 7
    },
    {
      "epoch": 0.051364365971107544,
      "grad_norm": 2640.0,
      "learning_rate": 7e-08,
      "loss": 68.3534,
      "step": 8
    },
    {
      "epoch": 0.05778491171749599,
      "grad_norm": 2608.0,
      "learning_rate": 8e-08,
      "loss": 63.9223,
      "step": 9
    },
    {
      "epoch": 0.06420545746388442,
      "grad_norm": 2688.0,
      "learning_rate": 9e-08,
      "loss": 69.6877,
      "step": 10
    },
    {
      "epoch": 0.07062600321027288,
      "grad_norm": 2736.0,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 71.1996,
      "step": 11
    },
    {
      "epoch": 0.07704654895666131,
      "grad_norm": 2560.0,
      "learning_rate": 1.1e-07,
      "loss": 65.4418,
      "step": 12
    },
    {
      "epoch": 0.08346709470304976,
      "grad_norm": 2624.0,
      "learning_rate": 1.2e-07,
      "loss": 67.7092,
      "step": 13
    },
    {
      "epoch": 0.0898876404494382,
      "grad_norm": 3568.0,
      "learning_rate": 1.3e-07,
      "loss": 62.1403,
      "step": 14
    },
    {
      "epoch": 0.09630818619582665,
      "grad_norm": 2576.0,
      "learning_rate": 1.4e-07,
      "loss": 69.297,
      "step": 15
    },
    {
      "epoch": 0.10272873194221509,
      "grad_norm": 2672.0,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 64.7075,
      "step": 16
    },
    {
      "epoch": 0.10914927768860354,
      "grad_norm": 2704.0,
      "learning_rate": 1.6e-07,
      "loss": 68.4324,
      "step": 17
    },
    {
      "epoch": 0.11556982343499198,
      "grad_norm": 2656.0,
      "learning_rate": 1.7e-07,
      "loss": 66.2092,
      "step": 18
    },
    {
      "epoch": 0.12199036918138041,
      "grad_norm": 2688.0,
      "learning_rate": 1.8e-07,
      "loss": 69.1647,
      "step": 19
    },
    {
      "epoch": 0.12841091492776885,
      "grad_norm": 2832.0,
      "learning_rate": 1.9e-07,
      "loss": 69.9635,
      "step": 20
    },
    {
      "epoch": 0.1348314606741573,
      "grad_norm": 2640.0,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 66.5122,
      "step": 21
    },
    {
      "epoch": 0.14125200642054575,
      "grad_norm": 2688.0,
      "learning_rate": 2.1e-07,
      "loss": 67.5501,
      "step": 22
    },
    {
      "epoch": 0.1476725521669342,
      "grad_norm": 2672.0,
      "learning_rate": 2.2e-07,
      "loss": 68.4469,
      "step": 23
    },
    {
      "epoch": 0.15409309791332262,
      "grad_norm": 2784.0,
      "learning_rate": 2.3e-07,
      "loss": 68.6615,
      "step": 24
    },
    {
      "epoch": 0.16051364365971107,
      "grad_norm": 2624.0,
      "learning_rate": 2.4e-07,
      "loss": 66.0181,
      "step": 25
    },
    {
      "epoch": 0.16693418940609953,
      "grad_norm": 2336.0,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 60.2096,
      "step": 26
    },
    {
      "epoch": 0.17335473515248795,
      "grad_norm": 2720.0,
      "learning_rate": 2.6e-07,
      "loss": 67.5297,
      "step": 27
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 2688.0,
      "learning_rate": 2.7e-07,
      "loss": 64.6912,
      "step": 28
    },
    {
      "epoch": 0.18619582664526485,
      "grad_norm": 2736.0,
      "learning_rate": 2.8e-07,
      "loss": 68.0737,
      "step": 29
    },
    {
      "epoch": 0.1926163723916533,
      "grad_norm": 2560.0,
      "learning_rate": 2.9e-07,
      "loss": 68.6904,
      "step": 30
    },
    {
      "epoch": 0.19903691813804172,
      "grad_norm": 2688.0,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 68.1428,
      "step": 31
    },
    {
      "epoch": 0.20545746388443017,
      "grad_norm": 2704.0,
      "learning_rate": 3.1e-07,
      "loss": 66.533,
      "step": 32
    },
    {
      "epoch": 0.21187800963081863,
      "grad_norm": 2560.0,
      "learning_rate": 3.2e-07,
      "loss": 64.648,
      "step": 33
    },
    {
      "epoch": 0.21829855537720708,
      "grad_norm": 2576.0,
      "learning_rate": 3.3e-07,
      "loss": 64.7796,
      "step": 34
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 2496.0,
      "learning_rate": 3.4e-07,
      "loss": 64.8224,
      "step": 35
    },
    {
      "epoch": 0.23113964686998395,
      "grad_norm": 2704.0,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 66.5291,
      "step": 36
    },
    {
      "epoch": 0.2375601926163724,
      "grad_norm": 3136.0,
      "learning_rate": 3.6e-07,
      "loss": 63.2267,
      "step": 37
    },
    {
      "epoch": 0.24398073836276082,
      "grad_norm": 2608.0,
      "learning_rate": 3.7000000000000006e-07,
      "loss": 67.5298,
      "step": 38
    },
    {
      "epoch": 0.2504012841091493,
      "grad_norm": 3936.0,
      "learning_rate": 3.8e-07,
      "loss": 66.8556,
      "step": 39
    },
    {
      "epoch": 0.2568218298555377,
      "grad_norm": 2640.0,
      "learning_rate": 3.9e-07,
      "loss": 65.8693,
      "step": 40
    },
    {
      "epoch": 0.26324237560192615,
      "grad_norm": 2752.0,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 65.4138,
      "step": 41
    },
    {
      "epoch": 0.2696629213483146,
      "grad_norm": 2544.0,
      "learning_rate": 4.1000000000000004e-07,
      "loss": 66.6703,
      "step": 42
    },
    {
      "epoch": 0.27608346709470305,
      "grad_norm": 2624.0,
      "learning_rate": 4.2e-07,
      "loss": 63.8722,
      "step": 43
    },
    {
      "epoch": 0.2825040128410915,
      "grad_norm": 2544.0,
      "learning_rate": 4.3e-07,
      "loss": 68.2989,
      "step": 44
    },
    {
      "epoch": 0.28892455858747995,
      "grad_norm": 2624.0,
      "learning_rate": 4.4e-07,
      "loss": 65.0818,
      "step": 45
    },
    {
      "epoch": 0.2953451043338684,
      "grad_norm": 2576.0,
      "learning_rate": 4.5e-07,
      "loss": 64.2289,
      "step": 46
    },
    {
      "epoch": 0.3017656500802568,
      "grad_norm": 2704.0,
      "learning_rate": 4.6e-07,
      "loss": 66.8615,
      "step": 47
    },
    {
      "epoch": 0.30818619582664525,
      "grad_norm": 2560.0,
      "learning_rate": 4.7000000000000005e-07,
      "loss": 66.4824,
      "step": 48
    },
    {
      "epoch": 0.3146067415730337,
      "grad_norm": 2656.0,
      "learning_rate": 4.8e-07,
      "loss": 68.432,
      "step": 49
    },
    {
      "epoch": 0.32102728731942215,
      "grad_norm": 2512.0,
      "learning_rate": 4.9e-07,
      "loss": 63.132,
      "step": 50
    },
    {
      "epoch": 0.3274478330658106,
      "grad_norm": 2592.0,
      "learning_rate": 5.000000000000001e-07,
      "loss": 66.8897,
      "step": 51
    },
    {
      "epoch": 0.33386837881219905,
      "grad_norm": 2592.0,
      "learning_rate": 5.100000000000001e-07,
      "loss": 65.4496,
      "step": 52
    },
    {
      "epoch": 0.3402889245585875,
      "grad_norm": 2448.0,
      "learning_rate": 5.2e-07,
      "loss": 63.2593,
      "step": 53
    },
    {
      "epoch": 0.3467094703049759,
      "grad_norm": 2480.0,
      "learning_rate": 5.3e-07,
      "loss": 64.0896,
      "step": 54
    },
    {
      "epoch": 0.35313001605136435,
      "grad_norm": 2432.0,
      "learning_rate": 5.4e-07,
      "loss": 62.3197,
      "step": 55
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 2624.0,
      "learning_rate": 5.5e-07,
      "loss": 63.0738,
      "step": 56
    },
    {
      "epoch": 0.36597110754414125,
      "grad_norm": 2528.0,
      "learning_rate": 5.6e-07,
      "loss": 59.7323,
      "step": 57
    },
    {
      "epoch": 0.3723916532905297,
      "grad_norm": 2528.0,
      "learning_rate": 5.7e-07,
      "loss": 65.4428,
      "step": 58
    },
    {
      "epoch": 0.37881219903691815,
      "grad_norm": 2560.0,
      "learning_rate": 5.8e-07,
      "loss": 59.6848,
      "step": 59
    },
    {
      "epoch": 0.3852327447833066,
      "grad_norm": 2432.0,
      "learning_rate": 5.9e-07,
      "loss": 62.4228,
      "step": 60
    },
    {
      "epoch": 0.391653290529695,
      "grad_norm": 2512.0,
      "learning_rate": 6.000000000000001e-07,
      "loss": 61.2024,
      "step": 61
    },
    {
      "epoch": 0.39807383627608345,
      "grad_norm": 2384.0,
      "learning_rate": 6.100000000000001e-07,
      "loss": 59.5451,
      "step": 62
    },
    {
      "epoch": 0.4044943820224719,
      "grad_norm": 2480.0,
      "learning_rate": 6.2e-07,
      "loss": 61.2986,
      "step": 63
    },
    {
      "epoch": 0.41091492776886035,
      "grad_norm": 2784.0,
      "learning_rate": 6.3e-07,
      "loss": 60.3593,
      "step": 64
    },
    {
      "epoch": 0.4173354735152488,
      "grad_norm": 2384.0,
      "learning_rate": 6.4e-07,
      "loss": 60.3284,
      "step": 65
    },
    {
      "epoch": 0.42375601926163725,
      "grad_norm": 2384.0,
      "learning_rate": 6.5e-07,
      "loss": 60.9094,
      "step": 66
    },
    {
      "epoch": 0.4301765650080257,
      "grad_norm": 2400.0,
      "learning_rate": 6.6e-07,
      "loss": 59.482,
      "step": 67
    },
    {
      "epoch": 0.43659711075441415,
      "grad_norm": 2384.0,
      "learning_rate": 6.7e-07,
      "loss": 58.1856,
      "step": 68
    },
    {
      "epoch": 0.44301765650080255,
      "grad_norm": 2432.0,
      "learning_rate": 6.8e-07,
      "loss": 60.6586,
      "step": 69
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 2400.0,
      "learning_rate": 6.900000000000001e-07,
      "loss": 61.602,
      "step": 70
    },
    {
      "epoch": 0.45585874799357945,
      "grad_norm": 2416.0,
      "learning_rate": 7.000000000000001e-07,
      "loss": 57.5153,
      "step": 71
    },
    {
      "epoch": 0.4622792937399679,
      "grad_norm": 2368.0,
      "learning_rate": 7.100000000000001e-07,
      "loss": 59.7101,
      "step": 72
    },
    {
      "epoch": 0.46869983948635635,
      "grad_norm": 2400.0,
      "learning_rate": 7.2e-07,
      "loss": 58.1877,
      "step": 73
    },
    {
      "epoch": 0.4751203852327448,
      "grad_norm": 2352.0,
      "learning_rate": 7.3e-07,
      "loss": 57.8943,
      "step": 74
    },
    {
      "epoch": 0.48154093097913325,
      "grad_norm": 2176.0,
      "learning_rate": 7.400000000000001e-07,
      "loss": 56.3821,
      "step": 75
    },
    {
      "epoch": 0.48796147672552165,
      "grad_norm": 2336.0,
      "learning_rate": 7.5e-07,
      "loss": 56.0092,
      "step": 76
    },
    {
      "epoch": 0.4943820224719101,
      "grad_norm": 2432.0,
      "learning_rate": 7.6e-07,
      "loss": 55.8024,
      "step": 77
    },
    {
      "epoch": 0.5008025682182986,
      "grad_norm": 2288.0,
      "learning_rate": 7.7e-07,
      "loss": 57.4868,
      "step": 78
    },
    {
      "epoch": 0.507223113964687,
      "grad_norm": 2480.0,
      "learning_rate": 7.8e-07,
      "loss": 57.2675,
      "step": 79
    },
    {
      "epoch": 0.5136436597110754,
      "grad_norm": 2160.0,
      "learning_rate": 7.900000000000002e-07,
      "loss": 55.0669,
      "step": 80
    },
    {
      "epoch": 0.5200642054574639,
      "grad_norm": 2368.0,
      "learning_rate": 8.000000000000001e-07,
      "loss": 58.2571,
      "step": 81
    },
    {
      "epoch": 0.5264847512038523,
      "grad_norm": 2176.0,
      "learning_rate": 8.1e-07,
      "loss": 55.8729,
      "step": 82
    },
    {
      "epoch": 0.5329052969502408,
      "grad_norm": 2352.0,
      "learning_rate": 8.200000000000001e-07,
      "loss": 58.6975,
      "step": 83
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 2320.0,
      "learning_rate": 8.300000000000001e-07,
      "loss": 58.1886,
      "step": 84
    },
    {
      "epoch": 0.5457463884430177,
      "grad_norm": 2192.0,
      "learning_rate": 8.4e-07,
      "loss": 55.7109,
      "step": 85
    },
    {
      "epoch": 0.5521669341894061,
      "grad_norm": 2272.0,
      "learning_rate": 8.500000000000001e-07,
      "loss": 56.3032,
      "step": 86
    },
    {
      "epoch": 0.5585874799357945,
      "grad_norm": 2224.0,
      "learning_rate": 8.6e-07,
      "loss": 51.9766,
      "step": 87
    },
    {
      "epoch": 0.565008025682183,
      "grad_norm": 2112.0,
      "learning_rate": 8.699999999999999e-07,
      "loss": 53.2478,
      "step": 88
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2352.0,
      "learning_rate": 8.8e-07,
      "loss": 58.7411,
      "step": 89
    },
    {
      "epoch": 0.5778491171749599,
      "grad_norm": 2240.0,
      "learning_rate": 8.900000000000001e-07,
      "loss": 53.9676,
      "step": 90
    },
    {
      "epoch": 0.5842696629213483,
      "grad_norm": 2112.0,
      "learning_rate": 9e-07,
      "loss": 51.1115,
      "step": 91
    },
    {
      "epoch": 0.5906902086677368,
      "grad_norm": 2080.0,
      "learning_rate": 9.100000000000001e-07,
      "loss": 53.7419,
      "step": 92
    },
    {
      "epoch": 0.5971107544141252,
      "grad_norm": 2176.0,
      "learning_rate": 9.2e-07,
      "loss": 52.9354,
      "step": 93
    },
    {
      "epoch": 0.6035313001605136,
      "grad_norm": 2128.0,
      "learning_rate": 9.3e-07,
      "loss": 51.0074,
      "step": 94
    },
    {
      "epoch": 0.6099518459069021,
      "grad_norm": 2112.0,
      "learning_rate": 9.400000000000001e-07,
      "loss": 50.9071,
      "step": 95
    },
    {
      "epoch": 0.6163723916532905,
      "grad_norm": 1984.0,
      "learning_rate": 9.5e-07,
      "loss": 50.7536,
      "step": 96
    },
    {
      "epoch": 0.622792937399679,
      "grad_norm": 1944.0,
      "learning_rate": 9.6e-07,
      "loss": 48.3991,
      "step": 97
    },
    {
      "epoch": 0.6292134831460674,
      "grad_norm": 2128.0,
      "learning_rate": 9.7e-07,
      "loss": 51.7168,
      "step": 98
    },
    {
      "epoch": 0.6356340288924559,
      "grad_norm": 2080.0,
      "learning_rate": 9.8e-07,
      "loss": 48.4291,
      "step": 99
    },
    {
      "epoch": 0.6420545746388443,
      "grad_norm": 2000.0,
      "learning_rate": 9.9e-07,
      "loss": 51.0727,
      "step": 100
    },
    {
      "epoch": 0.6484751203852327,
      "grad_norm": 2048.0,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 47.2227,
      "step": 101
    },
    {
      "epoch": 0.6548956661316212,
      "grad_norm": 1696.0,
      "learning_rate": 1.01e-06,
      "loss": 43.2169,
      "step": 102
    },
    {
      "epoch": 0.6613162118780096,
      "grad_norm": 1920.0,
      "learning_rate": 1.0200000000000002e-06,
      "loss": 48.9871,
      "step": 103
    },
    {
      "epoch": 0.6677367576243981,
      "grad_norm": 1920.0,
      "learning_rate": 1.03e-06,
      "loss": 47.0749,
      "step": 104
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 1768.0,
      "learning_rate": 1.04e-06,
      "loss": 42.987,
      "step": 105
    },
    {
      "epoch": 0.680577849117175,
      "grad_norm": 1816.0,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 44.7319,
      "step": 106
    },
    {
      "epoch": 0.6869983948635634,
      "grad_norm": 1944.0,
      "learning_rate": 1.06e-06,
      "loss": 46.5783,
      "step": 107
    },
    {
      "epoch": 0.6934189406099518,
      "grad_norm": 1672.0,
      "learning_rate": 1.07e-06,
      "loss": 43.2166,
      "step": 108
    },
    {
      "epoch": 0.6998394863563403,
      "grad_norm": 1632.0,
      "learning_rate": 1.08e-06,
      "loss": 42.2282,
      "step": 109
    },
    {
      "epoch": 0.7062600321027287,
      "grad_norm": 1624.0,
      "learning_rate": 1.0900000000000002e-06,
      "loss": 42.7272,
      "step": 110
    },
    {
      "epoch": 0.7126805778491172,
      "grad_norm": 1752.0,
      "learning_rate": 1.1e-06,
      "loss": 42.4449,
      "step": 111
    },
    {
      "epoch": 0.7191011235955056,
      "grad_norm": 1808.0,
      "learning_rate": 1.1100000000000002e-06,
      "loss": 40.9031,
      "step": 112
    },
    {
      "epoch": 0.7255216693418941,
      "grad_norm": 1784.0,
      "learning_rate": 1.12e-06,
      "loss": 40.4737,
      "step": 113
    },
    {
      "epoch": 0.7319422150882825,
      "grad_norm": 1536.0,
      "learning_rate": 1.13e-06,
      "loss": 40.0722,
      "step": 114
    },
    {
      "epoch": 0.7383627608346709,
      "grad_norm": 1592.0,
      "learning_rate": 1.14e-06,
      "loss": 39.246,
      "step": 115
    },
    {
      "epoch": 0.7447833065810594,
      "grad_norm": 1624.0,
      "learning_rate": 1.15e-06,
      "loss": 40.4427,
      "step": 116
    },
    {
      "epoch": 0.7512038523274478,
      "grad_norm": 1424.0,
      "learning_rate": 1.16e-06,
      "loss": 37.9958,
      "step": 117
    },
    {
      "epoch": 0.7576243980738363,
      "grad_norm": 1328.0,
      "learning_rate": 1.17e-06,
      "loss": 37.7608,
      "step": 118
    },
    {
      "epoch": 0.7640449438202247,
      "grad_norm": 1392.0,
      "learning_rate": 1.18e-06,
      "loss": 38.096,
      "step": 119
    },
    {
      "epoch": 0.7704654895666132,
      "grad_norm": 1400.0,
      "learning_rate": 1.19e-06,
      "loss": 37.8043,
      "step": 120
    },
    {
      "epoch": 0.7768860353130016,
      "grad_norm": 1424.0,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 36.2898,
      "step": 121
    },
    {
      "epoch": 0.78330658105939,
      "grad_norm": 1400.0,
      "learning_rate": 1.21e-06,
      "loss": 37.9876,
      "step": 122
    },
    {
      "epoch": 0.7897271268057785,
      "grad_norm": 1296.0,
      "learning_rate": 1.2200000000000002e-06,
      "loss": 35.8186,
      "step": 123
    },
    {
      "epoch": 0.7961476725521669,
      "grad_norm": 1336.0,
      "learning_rate": 1.23e-06,
      "loss": 34.5259,
      "step": 124
    },
    {
      "epoch": 0.8025682182985554,
      "grad_norm": 1256.0,
      "learning_rate": 1.24e-06,
      "loss": 35.6749,
      "step": 125
    },
    {
      "epoch": 0.8089887640449438,
      "grad_norm": 1336.0,
      "learning_rate": 1.25e-06,
      "loss": 34.393,
      "step": 126
    },
    {
      "epoch": 0.8154093097913323,
      "grad_norm": 1112.0,
      "learning_rate": 1.26e-06,
      "loss": 33.0221,
      "step": 127
    },
    {
      "epoch": 0.8218298555377207,
      "grad_norm": 1400.0,
      "learning_rate": 1.27e-06,
      "loss": 34.6581,
      "step": 128
    },
    {
      "epoch": 0.8282504012841091,
      "grad_norm": 1048.0,
      "learning_rate": 1.28e-06,
      "loss": 31.1044,
      "step": 129
    },
    {
      "epoch": 0.8346709470304976,
      "grad_norm": 1128.0,
      "learning_rate": 1.2900000000000001e-06,
      "loss": 31.8478,
      "step": 130
    },
    {
      "epoch": 0.841091492776886,
      "grad_norm": 1168.0,
      "learning_rate": 1.3e-06,
      "loss": 31.3836,
      "step": 131
    },
    {
      "epoch": 0.8475120385232745,
      "grad_norm": 1160.0,
      "learning_rate": 1.3100000000000002e-06,
      "loss": 31.4803,
      "step": 132
    },
    {
      "epoch": 0.8539325842696629,
      "grad_norm": 1040.0,
      "learning_rate": 1.32e-06,
      "loss": 30.0389,
      "step": 133
    },
    {
      "epoch": 0.8603531300160514,
      "grad_norm": 1088.0,
      "learning_rate": 1.33e-06,
      "loss": 31.3586,
      "step": 134
    },
    {
      "epoch": 0.8667736757624398,
      "grad_norm": 1120.0,
      "learning_rate": 1.34e-06,
      "loss": 31.1174,
      "step": 135
    },
    {
      "epoch": 0.8731942215088283,
      "grad_norm": 948.0,
      "learning_rate": 1.35e-06,
      "loss": 29.0728,
      "step": 136
    },
    {
      "epoch": 0.8796147672552167,
      "grad_norm": 900.0,
      "learning_rate": 1.36e-06,
      "loss": 28.5329,
      "step": 137
    },
    {
      "epoch": 0.8860353130016051,
      "grad_norm": 936.0,
      "learning_rate": 1.37e-06,
      "loss": 29.1405,
      "step": 138
    },
    {
      "epoch": 0.8924558587479936,
      "grad_norm": 952.0,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 28.8426,
      "step": 139
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 880.0,
      "learning_rate": 1.39e-06,
      "loss": 28.2564,
      "step": 140
    },
    {
      "epoch": 0.9052969502407705,
      "grad_norm": 860.0,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 27.616,
      "step": 141
    },
    {
      "epoch": 0.9117174959871589,
      "grad_norm": 828.0,
      "learning_rate": 1.41e-06,
      "loss": 28.4123,
      "step": 142
    },
    {
      "epoch": 0.9181380417335474,
      "grad_norm": 776.0,
      "learning_rate": 1.4200000000000002e-06,
      "loss": 26.376,
      "step": 143
    },
    {
      "epoch": 0.9245585874799358,
      "grad_norm": 828.0,
      "learning_rate": 1.43e-06,
      "loss": 27.0611,
      "step": 144
    },
    {
      "epoch": 0.9309791332263242,
      "grad_norm": 656.0,
      "learning_rate": 1.44e-06,
      "loss": 24.8307,
      "step": 145
    },
    {
      "epoch": 0.9373996789727127,
      "grad_norm": 692.0,
      "learning_rate": 1.45e-06,
      "loss": 25.3216,
      "step": 146
    },
    {
      "epoch": 0.9438202247191011,
      "grad_norm": 692.0,
      "learning_rate": 1.46e-06,
      "loss": 23.6355,
      "step": 147
    },
    {
      "epoch": 0.9502407704654896,
      "grad_norm": 776.0,
      "learning_rate": 1.4700000000000001e-06,
      "loss": 23.4762,
      "step": 148
    },
    {
      "epoch": 0.956661316211878,
      "grad_norm": 708.0,
      "learning_rate": 1.4800000000000002e-06,
      "loss": 24.8771,
      "step": 149
    },
    {
      "epoch": 0.9630818619582665,
      "grad_norm": 684.0,
      "learning_rate": 1.4900000000000001e-06,
      "loss": 25.1315,
      "step": 150
    },
    {
      "epoch": 0.9695024077046549,
      "grad_norm": 636.0,
      "learning_rate": 1.5e-06,
      "loss": 23.2373,
      "step": 151
    },
    {
      "epoch": 0.9759229534510433,
      "grad_norm": 500.0,
      "learning_rate": 1.5100000000000002e-06,
      "loss": 23.485,
      "step": 152
    },
    {
      "epoch": 0.9823434991974318,
      "grad_norm": 548.0,
      "learning_rate": 1.52e-06,
      "loss": 23.3626,
      "step": 153
    },
    {
      "epoch": 0.9887640449438202,
      "grad_norm": 700.0,
      "learning_rate": 1.53e-06,
      "loss": 23.1542,
      "step": 154
    },
    {
      "epoch": 0.9951845906902087,
      "grad_norm": 584.0,
      "learning_rate": 1.54e-06,
      "loss": 22.8096,
      "step": 155
    },
    {
      "epoch": 1.0,
      "grad_norm": 318.0,
      "learning_rate": 1.55e-06,
      "loss": 14.6985,
      "step": 156
    },
    {
      "epoch": 1.0064205457463884,
      "grad_norm": 552.0,
      "learning_rate": 1.56e-06,
      "loss": 21.8381,
      "step": 157
    },
    {
      "epoch": 1.0128410914927768,
      "grad_norm": 462.0,
      "learning_rate": 1.57e-06,
      "loss": 20.6383,
      "step": 158
    },
    {
      "epoch": 1.0192616372391654,
      "grad_norm": 512.0,
      "learning_rate": 1.5800000000000003e-06,
      "loss": 20.5827,
      "step": 159
    },
    {
      "epoch": 1.0256821829855538,
      "grad_norm": 510.0,
      "learning_rate": 1.5900000000000002e-06,
      "loss": 19.6922,
      "step": 160
    },
    {
      "epoch": 1.0321027287319422,
      "grad_norm": 446.0,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 19.9688,
      "step": 161
    },
    {
      "epoch": 1.0385232744783306,
      "grad_norm": 468.0,
      "learning_rate": 1.61e-06,
      "loss": 20.4126,
      "step": 162
    },
    {
      "epoch": 1.0449438202247192,
      "grad_norm": 402.0,
      "learning_rate": 1.62e-06,
      "loss": 18.945,
      "step": 163
    },
    {
      "epoch": 1.0513643659711076,
      "grad_norm": 358.0,
      "learning_rate": 1.6299999999999999e-06,
      "loss": 18.9673,
      "step": 164
    },
    {
      "epoch": 1.057784911717496,
      "grad_norm": 350.0,
      "learning_rate": 1.6400000000000002e-06,
      "loss": 19.0214,
      "step": 165
    },
    {
      "epoch": 1.0642054574638844,
      "grad_norm": 372.0,
      "learning_rate": 1.65e-06,
      "loss": 18.854,
      "step": 166
    },
    {
      "epoch": 1.0706260032102728,
      "grad_norm": 350.0,
      "learning_rate": 1.6600000000000002e-06,
      "loss": 18.4359,
      "step": 167
    },
    {
      "epoch": 1.0770465489566614,
      "grad_norm": 372.0,
      "learning_rate": 1.67e-06,
      "loss": 18.9277,
      "step": 168
    },
    {
      "epoch": 1.0834670947030498,
      "grad_norm": 310.0,
      "learning_rate": 1.68e-06,
      "loss": 17.8542,
      "step": 169
    },
    {
      "epoch": 1.0898876404494382,
      "grad_norm": 298.0,
      "learning_rate": 1.69e-06,
      "loss": 17.6669,
      "step": 170
    },
    {
      "epoch": 1.0963081861958266,
      "grad_norm": 334.0,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 18.6033,
      "step": 171
    },
    {
      "epoch": 1.102728731942215,
      "grad_norm": 306.0,
      "learning_rate": 1.7100000000000001e-06,
      "loss": 17.9333,
      "step": 172
    },
    {
      "epoch": 1.1091492776886036,
      "grad_norm": 258.0,
      "learning_rate": 1.72e-06,
      "loss": 16.9446,
      "step": 173
    },
    {
      "epoch": 1.115569823434992,
      "grad_norm": 286.0,
      "learning_rate": 1.73e-06,
      "loss": 16.2851,
      "step": 174
    },
    {
      "epoch": 1.1219903691813804,
      "grad_norm": 272.0,
      "learning_rate": 1.7399999999999999e-06,
      "loss": 16.2929,
      "step": 175
    },
    {
      "epoch": 1.1284109149277688,
      "grad_norm": 229.0,
      "learning_rate": 1.7500000000000002e-06,
      "loss": 16.4739,
      "step": 176
    },
    {
      "epoch": 1.1348314606741572,
      "grad_norm": 348.0,
      "learning_rate": 1.76e-06,
      "loss": 17.45,
      "step": 177
    },
    {
      "epoch": 1.1412520064205458,
      "grad_norm": 211.0,
      "learning_rate": 1.7700000000000002e-06,
      "loss": 15.2997,
      "step": 178
    },
    {
      "epoch": 1.1476725521669342,
      "grad_norm": 288.0,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 16.4745,
      "step": 179
    },
    {
      "epoch": 1.1540930979133226,
      "grad_norm": 211.0,
      "learning_rate": 1.79e-06,
      "loss": 15.3642,
      "step": 180
    },
    {
      "epoch": 1.160513643659711,
      "grad_norm": 204.0,
      "learning_rate": 1.8e-06,
      "loss": 15.7648,
      "step": 181
    },
    {
      "epoch": 1.1669341894060996,
      "grad_norm": 180.0,
      "learning_rate": 1.8100000000000002e-06,
      "loss": 14.727,
      "step": 182
    },
    {
      "epoch": 1.173354735152488,
      "grad_norm": 195.0,
      "learning_rate": 1.8200000000000002e-06,
      "loss": 14.532,
      "step": 183
    },
    {
      "epoch": 1.1797752808988764,
      "grad_norm": 199.0,
      "learning_rate": 1.83e-06,
      "loss": 14.1641,
      "step": 184
    },
    {
      "epoch": 1.1861958266452648,
      "grad_norm": 231.0,
      "learning_rate": 1.84e-06,
      "loss": 14.8582,
      "step": 185
    },
    {
      "epoch": 1.1926163723916532,
      "grad_norm": 212.0,
      "learning_rate": 1.85e-06,
      "loss": 14.3754,
      "step": 186
    },
    {
      "epoch": 1.1990369181380418,
      "grad_norm": 187.0,
      "learning_rate": 1.86e-06,
      "loss": 13.9474,
      "step": 187
    },
    {
      "epoch": 1.2054574638844302,
      "grad_norm": 191.0,
      "learning_rate": 1.8700000000000003e-06,
      "loss": 13.8309,
      "step": 188
    },
    {
      "epoch": 1.2118780096308186,
      "grad_norm": 183.0,
      "learning_rate": 1.8800000000000002e-06,
      "loss": 12.8467,
      "step": 189
    },
    {
      "epoch": 1.218298555377207,
      "grad_norm": 143.0,
      "learning_rate": 1.8900000000000001e-06,
      "loss": 12.9361,
      "step": 190
    },
    {
      "epoch": 1.2247191011235956,
      "grad_norm": 193.0,
      "learning_rate": 1.9e-06,
      "loss": 13.0388,
      "step": 191
    },
    {
      "epoch": 1.231139646869984,
      "grad_norm": 164.0,
      "learning_rate": 1.91e-06,
      "loss": 12.5101,
      "step": 192
    },
    {
      "epoch": 1.2375601926163724,
      "grad_norm": 156.0,
      "learning_rate": 1.92e-06,
      "loss": 12.295,
      "step": 193
    },
    {
      "epoch": 1.2439807383627608,
      "grad_norm": 155.0,
      "learning_rate": 1.93e-06,
      "loss": 11.9348,
      "step": 194
    },
    {
      "epoch": 1.2504012841091492,
      "grad_norm": 173.0,
      "learning_rate": 1.94e-06,
      "loss": 12.6231,
      "step": 195
    },
    {
      "epoch": 1.2568218298555376,
      "grad_norm": 126.5,
      "learning_rate": 1.95e-06,
      "loss": 11.3234,
      "step": 196
    },
    {
      "epoch": 1.2632423756019262,
      "grad_norm": 154.0,
      "learning_rate": 1.96e-06,
      "loss": 11.7104,
      "step": 197
    },
    {
      "epoch": 1.2696629213483146,
      "grad_norm": 156.0,
      "learning_rate": 1.9699999999999998e-06,
      "loss": 11.3229,
      "step": 198
    },
    {
      "epoch": 1.276083467094703,
      "grad_norm": 128.0,
      "learning_rate": 1.98e-06,
      "loss": 11.1193,
      "step": 199
    },
    {
      "epoch": 1.2825040128410916,
      "grad_norm": 162.0,
      "learning_rate": 1.99e-06,
      "loss": 10.9172,
      "step": 200
    },
    {
      "epoch": 1.28892455858748,
      "grad_norm": 109.5,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 11.4688,
      "step": 201
    },
    {
      "epoch": 1.2953451043338684,
      "grad_norm": 130.0,
      "learning_rate": 2.0100000000000002e-06,
      "loss": 10.6257,
      "step": 202
    },
    {
      "epoch": 1.3017656500802568,
      "grad_norm": 140.0,
      "learning_rate": 2.02e-06,
      "loss": 10.7059,
      "step": 203
    },
    {
      "epoch": 1.3081861958266452,
      "grad_norm": 123.5,
      "learning_rate": 2.03e-06,
      "loss": 10.4435,
      "step": 204
    },
    {
      "epoch": 1.3146067415730336,
      "grad_norm": 111.0,
      "learning_rate": 2.0400000000000004e-06,
      "loss": 10.3587,
      "step": 205
    },
    {
      "epoch": 1.3210272873194222,
      "grad_norm": 112.0,
      "learning_rate": 2.0500000000000003e-06,
      "loss": 10.2,
      "step": 206
    },
    {
      "epoch": 1.3274478330658106,
      "grad_norm": 115.0,
      "learning_rate": 2.06e-06,
      "loss": 9.8934,
      "step": 207
    },
    {
      "epoch": 1.333868378812199,
      "grad_norm": 107.0,
      "learning_rate": 2.07e-06,
      "loss": 9.7525,
      "step": 208
    },
    {
      "epoch": 1.3402889245585876,
      "grad_norm": 129.0,
      "learning_rate": 2.08e-06,
      "loss": 10.2076,
      "step": 209
    },
    {
      "epoch": 1.346709470304976,
      "grad_norm": 118.0,
      "learning_rate": 2.09e-06,
      "loss": 9.6471,
      "step": 210
    },
    {
      "epoch": 1.3531300160513644,
      "grad_norm": 106.5,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 9.4667,
      "step": 211
    },
    {
      "epoch": 1.3595505617977528,
      "grad_norm": 114.0,
      "learning_rate": 2.11e-06,
      "loss": 9.2211,
      "step": 212
    },
    {
      "epoch": 1.3659711075441412,
      "grad_norm": 81.0,
      "learning_rate": 2.12e-06,
      "loss": 9.0104,
      "step": 213
    },
    {
      "epoch": 1.3723916532905296,
      "grad_norm": 96.0,
      "learning_rate": 2.13e-06,
      "loss": 9.2574,
      "step": 214
    },
    {
      "epoch": 1.3788121990369182,
      "grad_norm": 78.0,
      "learning_rate": 2.14e-06,
      "loss": 9.0517,
      "step": 215
    },
    {
      "epoch": 1.3852327447833066,
      "grad_norm": 59.75,
      "learning_rate": 2.1499999999999997e-06,
      "loss": 8.3138,
      "step": 216
    },
    {
      "epoch": 1.391653290529695,
      "grad_norm": 77.0,
      "learning_rate": 2.16e-06,
      "loss": 8.5174,
      "step": 217
    },
    {
      "epoch": 1.3980738362760834,
      "grad_norm": 89.0,
      "learning_rate": 2.17e-06,
      "loss": 8.3941,
      "step": 218
    },
    {
      "epoch": 1.404494382022472,
      "grad_norm": 68.5,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 8.1793,
      "step": 219
    },
    {
      "epoch": 1.4109149277688604,
      "grad_norm": 61.75,
      "learning_rate": 2.19e-06,
      "loss": 8.0112,
      "step": 220
    },
    {
      "epoch": 1.4173354735152488,
      "grad_norm": 63.25,
      "learning_rate": 2.2e-06,
      "loss": 8.1937,
      "step": 221
    },
    {
      "epoch": 1.4237560192616372,
      "grad_norm": 68.0,
      "learning_rate": 2.2100000000000004e-06,
      "loss": 7.9726,
      "step": 222
    },
    {
      "epoch": 1.4301765650080256,
      "grad_norm": 74.5,
      "learning_rate": 2.2200000000000003e-06,
      "loss": 7.8816,
      "step": 223
    },
    {
      "epoch": 1.4365971107544142,
      "grad_norm": 65.5,
      "learning_rate": 2.2300000000000002e-06,
      "loss": 7.8442,
      "step": 224
    },
    {
      "epoch": 1.4430176565008026,
      "grad_norm": 68.5,
      "learning_rate": 2.24e-06,
      "loss": 7.6176,
      "step": 225
    },
    {
      "epoch": 1.449438202247191,
      "grad_norm": 55.25,
      "learning_rate": 2.25e-06,
      "loss": 7.3147,
      "step": 226
    },
    {
      "epoch": 1.4558587479935794,
      "grad_norm": 73.0,
      "learning_rate": 2.26e-06,
      "loss": 7.5017,
      "step": 227
    },
    {
      "epoch": 1.462279293739968,
      "grad_norm": 61.75,
      "learning_rate": 2.2700000000000003e-06,
      "loss": 7.1296,
      "step": 228
    },
    {
      "epoch": 1.4686998394863564,
      "grad_norm": 48.25,
      "learning_rate": 2.28e-06,
      "loss": 7.1482,
      "step": 229
    },
    {
      "epoch": 1.4751203852327448,
      "grad_norm": 49.75,
      "learning_rate": 2.29e-06,
      "loss": 6.9477,
      "step": 230
    },
    {
      "epoch": 1.4815409309791332,
      "grad_norm": 46.5,
      "learning_rate": 2.3e-06,
      "loss": 6.8802,
      "step": 231
    },
    {
      "epoch": 1.4879614767255216,
      "grad_norm": 57.75,
      "learning_rate": 2.31e-06,
      "loss": 6.9703,
      "step": 232
    },
    {
      "epoch": 1.49438202247191,
      "grad_norm": 80.5,
      "learning_rate": 2.32e-06,
      "loss": 6.8221,
      "step": 233
    },
    {
      "epoch": 1.5008025682182986,
      "grad_norm": 52.0,
      "learning_rate": 2.33e-06,
      "loss": 6.5523,
      "step": 234
    },
    {
      "epoch": 1.507223113964687,
      "grad_norm": 45.75,
      "learning_rate": 2.34e-06,
      "loss": 6.5678,
      "step": 235
    },
    {
      "epoch": 1.5136436597110754,
      "grad_norm": 48.25,
      "learning_rate": 2.35e-06,
      "loss": 6.5961,
      "step": 236
    },
    {
      "epoch": 1.520064205457464,
      "grad_norm": 46.75,
      "learning_rate": 2.36e-06,
      "loss": 6.3501,
      "step": 237
    },
    {
      "epoch": 1.5264847512038524,
      "grad_norm": 44.0,
      "learning_rate": 2.37e-06,
      "loss": 6.4374,
      "step": 238
    },
    {
      "epoch": 1.5329052969502408,
      "grad_norm": 55.5,
      "learning_rate": 2.38e-06,
      "loss": 6.2247,
      "step": 239
    },
    {
      "epoch": 1.5393258426966292,
      "grad_norm": 65.0,
      "learning_rate": 2.3900000000000004e-06,
      "loss": 6.3353,
      "step": 240
    },
    {
      "epoch": 1.5457463884430176,
      "grad_norm": 43.5,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 6.19,
      "step": 241
    },
    {
      "epoch": 1.552166934189406,
      "grad_norm": 64.0,
      "learning_rate": 2.4100000000000002e-06,
      "loss": 5.9859,
      "step": 242
    },
    {
      "epoch": 1.5585874799357944,
      "grad_norm": 40.0,
      "learning_rate": 2.42e-06,
      "loss": 6.1023,
      "step": 243
    },
    {
      "epoch": 1.565008025682183,
      "grad_norm": 40.75,
      "learning_rate": 2.43e-06,
      "loss": 5.9572,
      "step": 244
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 31.375,
      "learning_rate": 2.4400000000000004e-06,
      "loss": 5.825,
      "step": 245
    },
    {
      "epoch": 1.57784911717496,
      "grad_norm": 39.5,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 5.8367,
      "step": 246
    },
    {
      "epoch": 1.5842696629213484,
      "grad_norm": 25.625,
      "learning_rate": 2.46e-06,
      "loss": 5.6145,
      "step": 247
    },
    {
      "epoch": 1.5906902086677368,
      "grad_norm": 29.25,
      "learning_rate": 2.47e-06,
      "loss": 5.5492,
      "step": 248
    },
    {
      "epoch": 1.5971107544141252,
      "grad_norm": 24.75,
      "learning_rate": 2.48e-06,
      "loss": 5.5288,
      "step": 249
    },
    {
      "epoch": 1.6035313001605136,
      "grad_norm": 26.875,
      "learning_rate": 2.49e-06,
      "loss": 5.4456,
      "step": 250
    },
    {
      "epoch": 1.609951845906902,
      "grad_norm": 24.5,
      "learning_rate": 2.5e-06,
      "loss": 5.4041,
      "step": 251
    },
    {
      "epoch": 1.6163723916532904,
      "grad_norm": 48.75,
      "learning_rate": 2.51e-06,
      "loss": 5.4765,
      "step": 252
    },
    {
      "epoch": 1.622792937399679,
      "grad_norm": 35.25,
      "learning_rate": 2.52e-06,
      "loss": 5.4415,
      "step": 253
    },
    {
      "epoch": 1.6292134831460674,
      "grad_norm": 22.25,
      "learning_rate": 2.53e-06,
      "loss": 5.2895,
      "step": 254
    },
    {
      "epoch": 1.635634028892456,
      "grad_norm": 24.5,
      "learning_rate": 2.54e-06,
      "loss": 5.2233,
      "step": 255
    },
    {
      "epoch": 1.6420545746388444,
      "grad_norm": 21.625,
      "learning_rate": 2.55e-06,
      "loss": 5.1632,
      "step": 256
    },
    {
      "epoch": 1.6484751203852328,
      "grad_norm": 31.75,
      "learning_rate": 2.56e-06,
      "loss": 5.3393,
      "step": 257
    },
    {
      "epoch": 1.6548956661316212,
      "grad_norm": 22.625,
      "learning_rate": 2.5700000000000004e-06,
      "loss": 5.0161,
      "step": 258
    },
    {
      "epoch": 1.6613162118780096,
      "grad_norm": 19.375,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 5.1114,
      "step": 259
    },
    {
      "epoch": 1.667736757624398,
      "grad_norm": 18.125,
      "learning_rate": 2.59e-06,
      "loss": 5.1154,
      "step": 260
    },
    {
      "epoch": 1.6741573033707864,
      "grad_norm": 21.625,
      "learning_rate": 2.6e-06,
      "loss": 5.0329,
      "step": 261
    },
    {
      "epoch": 1.680577849117175,
      "grad_norm": 31.5,
      "learning_rate": 2.6100000000000004e-06,
      "loss": 4.9979,
      "step": 262
    },
    {
      "epoch": 1.6869983948635634,
      "grad_norm": 25.25,
      "learning_rate": 2.6200000000000003e-06,
      "loss": 5.0385,
      "step": 263
    },
    {
      "epoch": 1.6934189406099518,
      "grad_norm": 25.875,
      "learning_rate": 2.6300000000000002e-06,
      "loss": 5.0432,
      "step": 264
    },
    {
      "epoch": 1.6998394863563404,
      "grad_norm": 15.375,
      "learning_rate": 2.64e-06,
      "loss": 4.8379,
      "step": 265
    },
    {
      "epoch": 1.7062600321027288,
      "grad_norm": 21.125,
      "learning_rate": 2.65e-06,
      "loss": 4.8686,
      "step": 266
    },
    {
      "epoch": 1.7126805778491172,
      "grad_norm": 19.125,
      "learning_rate": 2.66e-06,
      "loss": 4.9208,
      "step": 267
    },
    {
      "epoch": 1.7191011235955056,
      "grad_norm": 13.375,
      "learning_rate": 2.6700000000000003e-06,
      "loss": 4.7493,
      "step": 268
    },
    {
      "epoch": 1.725521669341894,
      "grad_norm": 20.375,
      "learning_rate": 2.68e-06,
      "loss": 4.8399,
      "step": 269
    },
    {
      "epoch": 1.7319422150882824,
      "grad_norm": 12.9375,
      "learning_rate": 2.69e-06,
      "loss": 4.7345,
      "step": 270
    },
    {
      "epoch": 1.7383627608346708,
      "grad_norm": 20.75,
      "learning_rate": 2.7e-06,
      "loss": 4.7447,
      "step": 271
    },
    {
      "epoch": 1.7447833065810594,
      "grad_norm": 24.0,
      "learning_rate": 2.71e-06,
      "loss": 4.7403,
      "step": 272
    },
    {
      "epoch": 1.7512038523274478,
      "grad_norm": 16.25,
      "learning_rate": 2.72e-06,
      "loss": 4.7301,
      "step": 273
    },
    {
      "epoch": 1.7576243980738364,
      "grad_norm": 10.5,
      "learning_rate": 2.73e-06,
      "loss": 4.6193,
      "step": 274
    },
    {
      "epoch": 1.7640449438202248,
      "grad_norm": 8.875,
      "learning_rate": 2.74e-06,
      "loss": 4.5934,
      "step": 275
    },
    {
      "epoch": 1.7704654895666132,
      "grad_norm": 11.6875,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 4.6304,
      "step": 276
    },
    {
      "epoch": 1.7768860353130016,
      "grad_norm": 9.375,
      "learning_rate": 2.7600000000000003e-06,
      "loss": 4.5637,
      "step": 277
    },
    {
      "epoch": 1.78330658105939,
      "grad_norm": 11.0,
      "learning_rate": 2.77e-06,
      "loss": 4.5497,
      "step": 278
    },
    {
      "epoch": 1.7897271268057784,
      "grad_norm": 22.375,
      "learning_rate": 2.78e-06,
      "loss": 4.6591,
      "step": 279
    },
    {
      "epoch": 1.7961476725521668,
      "grad_norm": 8.5,
      "learning_rate": 2.7900000000000004e-06,
      "loss": 4.4922,
      "step": 280
    },
    {
      "epoch": 1.8025682182985554,
      "grad_norm": 21.625,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 4.5627,
      "step": 281
    },
    {
      "epoch": 1.8089887640449438,
      "grad_norm": 7.0625,
      "learning_rate": 2.81e-06,
      "loss": 4.4748,
      "step": 282
    },
    {
      "epoch": 1.8154093097913324,
      "grad_norm": 19.5,
      "learning_rate": 2.82e-06,
      "loss": 4.4487,
      "step": 283
    },
    {
      "epoch": 1.8218298555377208,
      "grad_norm": 12.8125,
      "learning_rate": 2.83e-06,
      "loss": 4.4716,
      "step": 284
    },
    {
      "epoch": 1.8282504012841092,
      "grad_norm": 14.875,
      "learning_rate": 2.8400000000000003e-06,
      "loss": 4.5438,
      "step": 285
    },
    {
      "epoch": 1.8346709470304976,
      "grad_norm": 13.375,
      "learning_rate": 2.8500000000000002e-06,
      "loss": 4.4262,
      "step": 286
    },
    {
      "epoch": 1.841091492776886,
      "grad_norm": 17.875,
      "learning_rate": 2.86e-06,
      "loss": 4.4127,
      "step": 287
    },
    {
      "epoch": 1.8475120385232744,
      "grad_norm": 10.0625,
      "learning_rate": 2.87e-06,
      "loss": 4.4463,
      "step": 288
    },
    {
      "epoch": 1.8539325842696628,
      "grad_norm": 10.875,
      "learning_rate": 2.88e-06,
      "loss": 4.4437,
      "step": 289
    },
    {
      "epoch": 1.8603531300160514,
      "grad_norm": 16.125,
      "learning_rate": 2.89e-06,
      "loss": 4.3908,
      "step": 290
    },
    {
      "epoch": 1.8667736757624398,
      "grad_norm": 9.9375,
      "learning_rate": 2.9e-06,
      "loss": 4.4305,
      "step": 291
    },
    {
      "epoch": 1.8731942215088284,
      "grad_norm": 5.21875,
      "learning_rate": 2.91e-06,
      "loss": 4.3518,
      "step": 292
    },
    {
      "epoch": 1.8796147672552168,
      "grad_norm": 8.4375,
      "learning_rate": 2.92e-06,
      "loss": 4.4349,
      "step": 293
    },
    {
      "epoch": 1.8860353130016052,
      "grad_norm": 9.75,
      "learning_rate": 2.93e-06,
      "loss": 4.3817,
      "step": 294
    },
    {
      "epoch": 1.8924558587479936,
      "grad_norm": 7.40625,
      "learning_rate": 2.9400000000000002e-06,
      "loss": 4.3206,
      "step": 295
    },
    {
      "epoch": 1.898876404494382,
      "grad_norm": 9.75,
      "learning_rate": 2.95e-06,
      "loss": 4.4141,
      "step": 296
    },
    {
      "epoch": 1.9052969502407704,
      "grad_norm": 12.4375,
      "learning_rate": 2.9600000000000005e-06,
      "loss": 4.3508,
      "step": 297
    },
    {
      "epoch": 1.9117174959871588,
      "grad_norm": 7.78125,
      "learning_rate": 2.9700000000000004e-06,
      "loss": 4.3224,
      "step": 298
    },
    {
      "epoch": 1.9181380417335474,
      "grad_norm": 24.875,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 4.3644,
      "step": 299
    },
    {
      "epoch": 1.9245585874799358,
      "grad_norm": 5.375,
      "learning_rate": 2.99e-06,
      "loss": 4.3205,
      "step": 300
    },
    {
      "epoch": 1.9309791332263242,
      "grad_norm": 4.3125,
      "learning_rate": 3e-06,
      "loss": 4.2883,
      "step": 301
    },
    {
      "epoch": 1.9373996789727128,
      "grad_norm": 8.125,
      "learning_rate": 3.01e-06,
      "loss": 4.2833,
      "step": 302
    },
    {
      "epoch": 1.9438202247191012,
      "grad_norm": 7.875,
      "learning_rate": 3.0200000000000003e-06,
      "loss": 4.3265,
      "step": 303
    },
    {
      "epoch": 1.9502407704654896,
      "grad_norm": 9.4375,
      "learning_rate": 3.0300000000000002e-06,
      "loss": 4.3152,
      "step": 304
    },
    {
      "epoch": 1.956661316211878,
      "grad_norm": 6.0625,
      "learning_rate": 3.04e-06,
      "loss": 4.2631,
      "step": 305
    },
    {
      "epoch": 1.9630818619582664,
      "grad_norm": 4.375,
      "learning_rate": 3.05e-06,
      "loss": 4.2309,
      "step": 306
    },
    {
      "epoch": 1.9695024077046548,
      "grad_norm": 6.5625,
      "learning_rate": 3.06e-06,
      "loss": 4.2553,
      "step": 307
    },
    {
      "epoch": 1.9759229534510432,
      "grad_norm": 5.8125,
      "learning_rate": 3.0700000000000003e-06,
      "loss": 4.2342,
      "step": 308
    },
    {
      "epoch": 1.9823434991974318,
      "grad_norm": 7.46875,
      "learning_rate": 3.08e-06,
      "loss": 4.26,
      "step": 309
    },
    {
      "epoch": 1.9887640449438202,
      "grad_norm": 12.1875,
      "learning_rate": 3.09e-06,
      "loss": 4.2549,
      "step": 310
    },
    {
      "epoch": 1.9951845906902088,
      "grad_norm": 4.5,
      "learning_rate": 3.1e-06,
      "loss": 4.2134,
      "step": 311
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.3125,
      "learning_rate": 3.11e-06,
      "loss": 3.1615,
      "step": 312
    },
    {
      "epoch": 2.0064205457463884,
      "grad_norm": 7.90625,
      "learning_rate": 3.12e-06,
      "loss": 4.2399,
      "step": 313
    },
    {
      "epoch": 2.012841091492777,
      "grad_norm": 6.9375,
      "learning_rate": 3.13e-06,
      "loss": 4.1947,
      "step": 314
    },
    {
      "epoch": 2.019261637239165,
      "grad_norm": 3.5625,
      "learning_rate": 3.14e-06,
      "loss": 4.1862,
      "step": 315
    },
    {
      "epoch": 2.0256821829855536,
      "grad_norm": 4.3125,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 4.2002,
      "step": 316
    },
    {
      "epoch": 2.0321027287319424,
      "grad_norm": 3.75,
      "learning_rate": 3.1600000000000007e-06,
      "loss": 4.1832,
      "step": 317
    },
    {
      "epoch": 2.038523274478331,
      "grad_norm": 3.5,
      "learning_rate": 3.17e-06,
      "loss": 4.1961,
      "step": 318
    },
    {
      "epoch": 2.044943820224719,
      "grad_norm": 5.21875,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 4.1732,
      "step": 319
    },
    {
      "epoch": 2.0513643659711076,
      "grad_norm": 3.375,
      "learning_rate": 3.19e-06,
      "loss": 4.1945,
      "step": 320
    },
    {
      "epoch": 2.057784911717496,
      "grad_norm": 2.6875,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 4.18,
      "step": 321
    },
    {
      "epoch": 2.0642054574638844,
      "grad_norm": 5.0,
      "learning_rate": 3.2099999999999998e-06,
      "loss": 4.2207,
      "step": 322
    },
    {
      "epoch": 2.070626003210273,
      "grad_norm": 4.1875,
      "learning_rate": 3.22e-06,
      "loss": 4.1885,
      "step": 323
    },
    {
      "epoch": 2.077046548956661,
      "grad_norm": 3.34375,
      "learning_rate": 3.2300000000000004e-06,
      "loss": 4.1569,
      "step": 324
    },
    {
      "epoch": 2.0834670947030496,
      "grad_norm": 4.09375,
      "learning_rate": 3.24e-06,
      "loss": 4.1438,
      "step": 325
    },
    {
      "epoch": 2.0898876404494384,
      "grad_norm": 2.046875,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 4.1505,
      "step": 326
    },
    {
      "epoch": 2.096308186195827,
      "grad_norm": 5.1875,
      "learning_rate": 3.2599999999999997e-06,
      "loss": 4.1533,
      "step": 327
    },
    {
      "epoch": 2.102728731942215,
      "grad_norm": 3.375,
      "learning_rate": 3.27e-06,
      "loss": 4.1264,
      "step": 328
    },
    {
      "epoch": 2.1091492776886036,
      "grad_norm": 8.6875,
      "learning_rate": 3.2800000000000004e-06,
      "loss": 4.1112,
      "step": 329
    },
    {
      "epoch": 2.115569823434992,
      "grad_norm": 4.53125,
      "learning_rate": 3.29e-06,
      "loss": 4.149,
      "step": 330
    },
    {
      "epoch": 2.1219903691813804,
      "grad_norm": 2.96875,
      "learning_rate": 3.3e-06,
      "loss": 4.1277,
      "step": 331
    },
    {
      "epoch": 2.128410914927769,
      "grad_norm": 3.109375,
      "learning_rate": 3.31e-06,
      "loss": 4.1461,
      "step": 332
    },
    {
      "epoch": 2.134831460674157,
      "grad_norm": 2.40625,
      "learning_rate": 3.3200000000000004e-06,
      "loss": 4.1228,
      "step": 333
    },
    {
      "epoch": 2.1412520064205456,
      "grad_norm": 20.25,
      "learning_rate": 3.3300000000000003e-06,
      "loss": 4.2054,
      "step": 334
    },
    {
      "epoch": 2.1476725521669344,
      "grad_norm": 5.34375,
      "learning_rate": 3.34e-06,
      "loss": 4.1536,
      "step": 335
    },
    {
      "epoch": 2.154093097913323,
      "grad_norm": 10.25,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 4.1162,
      "step": 336
    },
    {
      "epoch": 2.160513643659711,
      "grad_norm": 6.5,
      "learning_rate": 3.36e-06,
      "loss": 4.1375,
      "step": 337
    },
    {
      "epoch": 2.1669341894060996,
      "grad_norm": 11.4375,
      "learning_rate": 3.3700000000000003e-06,
      "loss": 4.1674,
      "step": 338
    },
    {
      "epoch": 2.173354735152488,
      "grad_norm": 3.5625,
      "learning_rate": 3.38e-06,
      "loss": 4.1092,
      "step": 339
    },
    {
      "epoch": 2.1797752808988764,
      "grad_norm": 28.0,
      "learning_rate": 3.39e-06,
      "loss": 4.1105,
      "step": 340
    },
    {
      "epoch": 2.186195826645265,
      "grad_norm": 2.0625,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 4.0914,
      "step": 341
    },
    {
      "epoch": 2.192616372391653,
      "grad_norm": 10.625,
      "learning_rate": 3.41e-06,
      "loss": 4.1199,
      "step": 342
    },
    {
      "epoch": 2.1990369181380416,
      "grad_norm": 2.078125,
      "learning_rate": 3.4200000000000003e-06,
      "loss": 4.1136,
      "step": 343
    },
    {
      "epoch": 2.20545746388443,
      "grad_norm": 2.078125,
      "learning_rate": 3.4299999999999998e-06,
      "loss": 4.1005,
      "step": 344
    },
    {
      "epoch": 2.211878009630819,
      "grad_norm": 1.2890625,
      "learning_rate": 3.44e-06,
      "loss": 4.082,
      "step": 345
    },
    {
      "epoch": 2.218298555377207,
      "grad_norm": 6.34375,
      "learning_rate": 3.4500000000000004e-06,
      "loss": 4.1012,
      "step": 346
    },
    {
      "epoch": 2.2247191011235956,
      "grad_norm": 2.875,
      "learning_rate": 3.46e-06,
      "loss": 4.0978,
      "step": 347
    },
    {
      "epoch": 2.231139646869984,
      "grad_norm": 40.75,
      "learning_rate": 3.4700000000000002e-06,
      "loss": 4.1382,
      "step": 348
    },
    {
      "epoch": 2.2375601926163724,
      "grad_norm": 6.46875,
      "learning_rate": 3.4799999999999997e-06,
      "loss": 4.1005,
      "step": 349
    },
    {
      "epoch": 2.243980738362761,
      "grad_norm": 5.84375,
      "learning_rate": 3.49e-06,
      "loss": 4.0952,
      "step": 350
    },
    {
      "epoch": 2.250401284109149,
      "grad_norm": 4.3125,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 4.0907,
      "step": 351
    },
    {
      "epoch": 2.2568218298555376,
      "grad_norm": 6.28125,
      "learning_rate": 3.5100000000000003e-06,
      "loss": 4.1052,
      "step": 352
    },
    {
      "epoch": 2.263242375601926,
      "grad_norm": 3.65625,
      "learning_rate": 3.52e-06,
      "loss": 4.1076,
      "step": 353
    },
    {
      "epoch": 2.2696629213483144,
      "grad_norm": 2.03125,
      "learning_rate": 3.53e-06,
      "loss": 4.1015,
      "step": 354
    },
    {
      "epoch": 2.276083467094703,
      "grad_norm": 4.78125,
      "learning_rate": 3.5400000000000004e-06,
      "loss": 4.09,
      "step": 355
    },
    {
      "epoch": 2.2825040128410916,
      "grad_norm": 11.625,
      "learning_rate": 3.55e-06,
      "loss": 4.0933,
      "step": 356
    },
    {
      "epoch": 2.28892455858748,
      "grad_norm": 1.203125,
      "learning_rate": 3.5600000000000002e-06,
      "loss": 4.0857,
      "step": 357
    },
    {
      "epoch": 2.2953451043338684,
      "grad_norm": 2.59375,
      "learning_rate": 3.5700000000000005e-06,
      "loss": 4.0942,
      "step": 358
    },
    {
      "epoch": 2.301765650080257,
      "grad_norm": 1.6796875,
      "learning_rate": 3.58e-06,
      "loss": 4.0925,
      "step": 359
    },
    {
      "epoch": 2.308186195826645,
      "grad_norm": 1.515625,
      "learning_rate": 3.5900000000000004e-06,
      "loss": 4.0665,
      "step": 360
    },
    {
      "epoch": 2.3146067415730336,
      "grad_norm": 1.1796875,
      "learning_rate": 3.6e-06,
      "loss": 4.0716,
      "step": 361
    },
    {
      "epoch": 2.321027287319422,
      "grad_norm": 6.53125,
      "learning_rate": 3.61e-06,
      "loss": 4.107,
      "step": 362
    },
    {
      "epoch": 2.3274478330658104,
      "grad_norm": 4.4375,
      "learning_rate": 3.6200000000000005e-06,
      "loss": 4.0748,
      "step": 363
    },
    {
      "epoch": 2.333868378812199,
      "grad_norm": 4.71875,
      "learning_rate": 3.63e-06,
      "loss": 4.0687,
      "step": 364
    },
    {
      "epoch": 2.3402889245585876,
      "grad_norm": 1.1171875,
      "learning_rate": 3.6400000000000003e-06,
      "loss": 4.0712,
      "step": 365
    },
    {
      "epoch": 2.346709470304976,
      "grad_norm": 1.1953125,
      "learning_rate": 3.6499999999999998e-06,
      "loss": 4.1146,
      "step": 366
    },
    {
      "epoch": 2.3531300160513644,
      "grad_norm": 1.2578125,
      "learning_rate": 3.66e-06,
      "loss": 4.0619,
      "step": 367
    },
    {
      "epoch": 2.359550561797753,
      "grad_norm": 2.125,
      "learning_rate": 3.6700000000000004e-06,
      "loss": 4.0759,
      "step": 368
    },
    {
      "epoch": 2.365971107544141,
      "grad_norm": 3.703125,
      "learning_rate": 3.68e-06,
      "loss": 4.0774,
      "step": 369
    },
    {
      "epoch": 2.3723916532905296,
      "grad_norm": 2.84375,
      "learning_rate": 3.6900000000000002e-06,
      "loss": 4.0755,
      "step": 370
    },
    {
      "epoch": 2.378812199036918,
      "grad_norm": 1.4609375,
      "learning_rate": 3.7e-06,
      "loss": 4.0727,
      "step": 371
    },
    {
      "epoch": 2.3852327447833064,
      "grad_norm": 8.25,
      "learning_rate": 3.7100000000000005e-06,
      "loss": 4.1228,
      "step": 372
    },
    {
      "epoch": 2.391653290529695,
      "grad_norm": 4.34375,
      "learning_rate": 3.72e-06,
      "loss": 4.0669,
      "step": 373
    },
    {
      "epoch": 2.3980738362760836,
      "grad_norm": 8.5625,
      "learning_rate": 3.7300000000000003e-06,
      "loss": 4.0941,
      "step": 374
    },
    {
      "epoch": 2.404494382022472,
      "grad_norm": 2.09375,
      "learning_rate": 3.7400000000000006e-06,
      "loss": 4.0692,
      "step": 375
    },
    {
      "epoch": 2.4109149277688604,
      "grad_norm": 3.65625,
      "learning_rate": 3.75e-06,
      "loss": 4.0725,
      "step": 376
    },
    {
      "epoch": 2.417335473515249,
      "grad_norm": 8.1875,
      "learning_rate": 3.7600000000000004e-06,
      "loss": 4.0785,
      "step": 377
    },
    {
      "epoch": 2.423756019261637,
      "grad_norm": 4.875,
      "learning_rate": 3.77e-06,
      "loss": 4.0699,
      "step": 378
    },
    {
      "epoch": 2.4301765650080256,
      "grad_norm": 4.875,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 4.0564,
      "step": 379
    },
    {
      "epoch": 2.436597110754414,
      "grad_norm": 3.078125,
      "learning_rate": 3.7900000000000006e-06,
      "loss": 4.0627,
      "step": 380
    },
    {
      "epoch": 2.4430176565008024,
      "grad_norm": 1.3828125,
      "learning_rate": 3.8e-06,
      "loss": 4.0638,
      "step": 381
    },
    {
      "epoch": 2.449438202247191,
      "grad_norm": 2.390625,
      "learning_rate": 3.8100000000000004e-06,
      "loss": 4.0799,
      "step": 382
    },
    {
      "epoch": 2.4558587479935796,
      "grad_norm": 5.84375,
      "learning_rate": 3.82e-06,
      "loss": 4.0693,
      "step": 383
    },
    {
      "epoch": 2.462279293739968,
      "grad_norm": 3.40625,
      "learning_rate": 3.830000000000001e-06,
      "loss": 4.0594,
      "step": 384
    },
    {
      "epoch": 2.4686998394863564,
      "grad_norm": 1.5,
      "learning_rate": 3.84e-06,
      "loss": 4.0492,
      "step": 385
    },
    {
      "epoch": 2.475120385232745,
      "grad_norm": 2.34375,
      "learning_rate": 3.85e-06,
      "loss": 4.059,
      "step": 386
    },
    {
      "epoch": 2.481540930979133,
      "grad_norm": 3.609375,
      "learning_rate": 3.86e-06,
      "loss": 4.0871,
      "step": 387
    },
    {
      "epoch": 2.4879614767255216,
      "grad_norm": 8.3125,
      "learning_rate": 3.87e-06,
      "loss": 4.0489,
      "step": 388
    },
    {
      "epoch": 2.49438202247191,
      "grad_norm": 5.6875,
      "learning_rate": 3.88e-06,
      "loss": 4.0586,
      "step": 389
    },
    {
      "epoch": 2.5008025682182984,
      "grad_norm": 4.53125,
      "learning_rate": 3.89e-06,
      "loss": 4.051,
      "step": 390
    },
    {
      "epoch": 2.5072231139646872,
      "grad_norm": 1.21875,
      "learning_rate": 3.9e-06,
      "loss": 4.0547,
      "step": 391
    },
    {
      "epoch": 2.513643659711075,
      "grad_norm": 4.3125,
      "learning_rate": 3.910000000000001e-06,
      "loss": 4.0584,
      "step": 392
    },
    {
      "epoch": 2.520064205457464,
      "grad_norm": 1.5234375,
      "learning_rate": 3.92e-06,
      "loss": 4.0787,
      "step": 393
    },
    {
      "epoch": 2.5264847512038524,
      "grad_norm": 3.859375,
      "learning_rate": 3.9300000000000005e-06,
      "loss": 4.0694,
      "step": 394
    },
    {
      "epoch": 2.532905296950241,
      "grad_norm": 1.515625,
      "learning_rate": 3.9399999999999995e-06,
      "loss": 4.0557,
      "step": 395
    },
    {
      "epoch": 2.539325842696629,
      "grad_norm": 1.140625,
      "learning_rate": 3.95e-06,
      "loss": 4.0455,
      "step": 396
    },
    {
      "epoch": 2.5457463884430176,
      "grad_norm": 2.34375,
      "learning_rate": 3.96e-06,
      "loss": 4.0568,
      "step": 397
    },
    {
      "epoch": 2.552166934189406,
      "grad_norm": 40.5,
      "learning_rate": 3.97e-06,
      "loss": 4.1181,
      "step": 398
    },
    {
      "epoch": 2.5585874799357944,
      "grad_norm": 5.78125,
      "learning_rate": 3.98e-06,
      "loss": 4.0634,
      "step": 399
    },
    {
      "epoch": 2.5650080256821832,
      "grad_norm": 2.484375,
      "learning_rate": 3.99e-06,
      "loss": 4.0493,
      "step": 400
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 1.65625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 4.0593,
      "step": 401
    },
    {
      "epoch": 2.57784911717496,
      "grad_norm": 4.40625,
      "learning_rate": 4.01e-06,
      "loss": 4.0425,
      "step": 402
    },
    {
      "epoch": 2.5842696629213484,
      "grad_norm": 3.9375,
      "learning_rate": 4.0200000000000005e-06,
      "loss": 4.0429,
      "step": 403
    },
    {
      "epoch": 2.590690208667737,
      "grad_norm": 3.296875,
      "learning_rate": 4.03e-06,
      "loss": 4.0476,
      "step": 404
    },
    {
      "epoch": 2.597110754414125,
      "grad_norm": 2.265625,
      "learning_rate": 4.04e-06,
      "loss": 4.06,
      "step": 405
    },
    {
      "epoch": 2.6035313001605136,
      "grad_norm": 1.8203125,
      "learning_rate": 4.05e-06,
      "loss": 4.0492,
      "step": 406
    },
    {
      "epoch": 2.609951845906902,
      "grad_norm": 3.53125,
      "learning_rate": 4.06e-06,
      "loss": 4.06,
      "step": 407
    },
    {
      "epoch": 2.6163723916532904,
      "grad_norm": 3.328125,
      "learning_rate": 4.07e-06,
      "loss": 4.0693,
      "step": 408
    },
    {
      "epoch": 2.6227929373996792,
      "grad_norm": 4.46875,
      "learning_rate": 4.080000000000001e-06,
      "loss": 4.0563,
      "step": 409
    },
    {
      "epoch": 2.629213483146067,
      "grad_norm": 2.5,
      "learning_rate": 4.09e-06,
      "loss": 4.0413,
      "step": 410
    },
    {
      "epoch": 2.635634028892456,
      "grad_norm": 2.1875,
      "learning_rate": 4.1000000000000006e-06,
      "loss": 4.0359,
      "step": 411
    },
    {
      "epoch": 2.6420545746388444,
      "grad_norm": 1.296875,
      "learning_rate": 4.11e-06,
      "loss": 4.0433,
      "step": 412
    },
    {
      "epoch": 2.648475120385233,
      "grad_norm": 3.21875,
      "learning_rate": 4.12e-06,
      "loss": 4.0376,
      "step": 413
    },
    {
      "epoch": 2.654895666131621,
      "grad_norm": 5.0625,
      "learning_rate": 4.13e-06,
      "loss": 4.0386,
      "step": 414
    },
    {
      "epoch": 2.6613162118780096,
      "grad_norm": 6.96875,
      "learning_rate": 4.14e-06,
      "loss": 4.0753,
      "step": 415
    },
    {
      "epoch": 2.667736757624398,
      "grad_norm": 1.421875,
      "learning_rate": 4.15e-06,
      "loss": 4.0429,
      "step": 416
    },
    {
      "epoch": 2.6741573033707864,
      "grad_norm": 2.0625,
      "learning_rate": 4.16e-06,
      "loss": 4.0534,
      "step": 417
    },
    {
      "epoch": 2.6805778491171752,
      "grad_norm": 1.859375,
      "learning_rate": 4.17e-06,
      "loss": 4.0364,
      "step": 418
    },
    {
      "epoch": 2.686998394863563,
      "grad_norm": 3.9375,
      "learning_rate": 4.18e-06,
      "loss": 4.047,
      "step": 419
    },
    {
      "epoch": 2.693418940609952,
      "grad_norm": 1.578125,
      "learning_rate": 4.1900000000000005e-06,
      "loss": 4.0549,
      "step": 420
    },
    {
      "epoch": 2.6998394863563404,
      "grad_norm": 7.125,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 4.0576,
      "step": 421
    },
    {
      "epoch": 2.706260032102729,
      "grad_norm": 1.0546875,
      "learning_rate": 4.21e-06,
      "loss": 4.0314,
      "step": 422
    },
    {
      "epoch": 2.712680577849117,
      "grad_norm": 0.82421875,
      "learning_rate": 4.22e-06,
      "loss": 4.0282,
      "step": 423
    },
    {
      "epoch": 2.7191011235955056,
      "grad_norm": 5.3125,
      "learning_rate": 4.23e-06,
      "loss": 4.0467,
      "step": 424
    },
    {
      "epoch": 2.725521669341894,
      "grad_norm": 1.8046875,
      "learning_rate": 4.24e-06,
      "loss": 4.0302,
      "step": 425
    },
    {
      "epoch": 2.7319422150882824,
      "grad_norm": 3.171875,
      "learning_rate": 4.250000000000001e-06,
      "loss": 4.0441,
      "step": 426
    },
    {
      "epoch": 2.738362760834671,
      "grad_norm": 0.6875,
      "learning_rate": 4.26e-06,
      "loss": 4.0405,
      "step": 427
    },
    {
      "epoch": 2.744783306581059,
      "grad_norm": 2.78125,
      "learning_rate": 4.270000000000001e-06,
      "loss": 4.0466,
      "step": 428
    },
    {
      "epoch": 2.751203852327448,
      "grad_norm": 2.90625,
      "learning_rate": 4.28e-06,
      "loss": 4.0432,
      "step": 429
    },
    {
      "epoch": 2.7576243980738364,
      "grad_norm": 1.2734375,
      "learning_rate": 4.2900000000000004e-06,
      "loss": 4.0422,
      "step": 430
    },
    {
      "epoch": 2.764044943820225,
      "grad_norm": 0.60546875,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 4.0364,
      "step": 431
    },
    {
      "epoch": 2.770465489566613,
      "grad_norm": 1.0078125,
      "learning_rate": 4.31e-06,
      "loss": 4.034,
      "step": 432
    },
    {
      "epoch": 2.7768860353130016,
      "grad_norm": 2.28125,
      "learning_rate": 4.32e-06,
      "loss": 4.0355,
      "step": 433
    },
    {
      "epoch": 2.78330658105939,
      "grad_norm": 1.03125,
      "learning_rate": 4.33e-06,
      "loss": 4.0456,
      "step": 434
    },
    {
      "epoch": 2.7897271268057784,
      "grad_norm": 0.94140625,
      "learning_rate": 4.34e-06,
      "loss": 4.0502,
      "step": 435
    },
    {
      "epoch": 2.796147672552167,
      "grad_norm": 0.703125,
      "learning_rate": 4.35e-06,
      "loss": 4.0428,
      "step": 436
    },
    {
      "epoch": 2.802568218298555,
      "grad_norm": 1.2578125,
      "learning_rate": 4.360000000000001e-06,
      "loss": 4.0361,
      "step": 437
    },
    {
      "epoch": 2.808988764044944,
      "grad_norm": 0.625,
      "learning_rate": 4.3700000000000005e-06,
      "loss": 4.0362,
      "step": 438
    },
    {
      "epoch": 2.8154093097913324,
      "grad_norm": 1.0625,
      "learning_rate": 4.38e-06,
      "loss": 4.0375,
      "step": 439
    },
    {
      "epoch": 2.821829855537721,
      "grad_norm": 0.51953125,
      "learning_rate": 4.39e-06,
      "loss": 4.0272,
      "step": 440
    },
    {
      "epoch": 2.828250401284109,
      "grad_norm": 1.125,
      "learning_rate": 4.4e-06,
      "loss": 4.0266,
      "step": 441
    },
    {
      "epoch": 2.8346709470304976,
      "grad_norm": 13.4375,
      "learning_rate": 4.41e-06,
      "loss": 4.0546,
      "step": 442
    },
    {
      "epoch": 2.841091492776886,
      "grad_norm": 1.046875,
      "learning_rate": 4.420000000000001e-06,
      "loss": 4.0325,
      "step": 443
    },
    {
      "epoch": 2.8475120385232744,
      "grad_norm": 5.71875,
      "learning_rate": 4.43e-06,
      "loss": 4.0256,
      "step": 444
    },
    {
      "epoch": 2.853932584269663,
      "grad_norm": 1.09375,
      "learning_rate": 4.440000000000001e-06,
      "loss": 4.0234,
      "step": 445
    },
    {
      "epoch": 2.860353130016051,
      "grad_norm": 1.0859375,
      "learning_rate": 4.45e-06,
      "loss": 4.0318,
      "step": 446
    },
    {
      "epoch": 2.86677367576244,
      "grad_norm": 1.8828125,
      "learning_rate": 4.4600000000000005e-06,
      "loss": 4.0427,
      "step": 447
    },
    {
      "epoch": 2.8731942215088284,
      "grad_norm": 1.2578125,
      "learning_rate": 4.4699999999999996e-06,
      "loss": 4.033,
      "step": 448
    },
    {
      "epoch": 2.879614767255217,
      "grad_norm": 1.1484375,
      "learning_rate": 4.48e-06,
      "loss": 4.0295,
      "step": 449
    },
    {
      "epoch": 2.886035313001605,
      "grad_norm": 0.7421875,
      "learning_rate": 4.49e-06,
      "loss": 4.0233,
      "step": 450
    },
    {
      "epoch": 2.8924558587479936,
      "grad_norm": 1.15625,
      "learning_rate": 4.5e-06,
      "loss": 4.0212,
      "step": 451
    },
    {
      "epoch": 2.898876404494382,
      "grad_norm": 0.79296875,
      "learning_rate": 4.51e-06,
      "loss": 4.0297,
      "step": 452
    },
    {
      "epoch": 2.9052969502407704,
      "grad_norm": 0.92578125,
      "learning_rate": 4.52e-06,
      "loss": 4.0354,
      "step": 453
    },
    {
      "epoch": 2.911717495987159,
      "grad_norm": 1.296875,
      "learning_rate": 4.53e-06,
      "loss": 4.0328,
      "step": 454
    },
    {
      "epoch": 2.918138041733547,
      "grad_norm": 1.34375,
      "learning_rate": 4.540000000000001e-06,
      "loss": 4.0276,
      "step": 455
    },
    {
      "epoch": 2.924558587479936,
      "grad_norm": 2.3125,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 4.0442,
      "step": 456
    },
    {
      "epoch": 2.930979133226324,
      "grad_norm": 8.625,
      "learning_rate": 4.56e-06,
      "loss": 4.0343,
      "step": 457
    },
    {
      "epoch": 2.937399678972713,
      "grad_norm": 3.75,
      "learning_rate": 4.57e-06,
      "loss": 4.0224,
      "step": 458
    },
    {
      "epoch": 2.943820224719101,
      "grad_norm": 1.515625,
      "learning_rate": 4.58e-06,
      "loss": 4.0251,
      "step": 459
    },
    {
      "epoch": 2.9502407704654896,
      "grad_norm": 1.3203125,
      "learning_rate": 4.590000000000001e-06,
      "loss": 4.0187,
      "step": 460
    },
    {
      "epoch": 2.956661316211878,
      "grad_norm": 1.46875,
      "learning_rate": 4.6e-06,
      "loss": 4.0195,
      "step": 461
    },
    {
      "epoch": 2.9630818619582664,
      "grad_norm": 1.4765625,
      "learning_rate": 4.610000000000001e-06,
      "loss": 4.0311,
      "step": 462
    },
    {
      "epoch": 2.969502407704655,
      "grad_norm": 0.8671875,
      "learning_rate": 4.62e-06,
      "loss": 4.0342,
      "step": 463
    },
    {
      "epoch": 2.975922953451043,
      "grad_norm": 2.953125,
      "learning_rate": 4.6300000000000006e-06,
      "loss": 4.0225,
      "step": 464
    },
    {
      "epoch": 2.982343499197432,
      "grad_norm": 2.34375,
      "learning_rate": 4.64e-06,
      "loss": 4.0415,
      "step": 465
    },
    {
      "epoch": 2.98876404494382,
      "grad_norm": 2.453125,
      "learning_rate": 4.65e-06,
      "loss": 4.0355,
      "step": 466
    },
    {
      "epoch": 2.995184590690209,
      "grad_norm": 1.265625,
      "learning_rate": 4.66e-06,
      "loss": 4.0247,
      "step": 467
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9375,
      "learning_rate": 4.67e-06,
      "loss": 3.0141,
      "step": 468
    },
    {
      "epoch": 3.0064205457463884,
      "grad_norm": 1.046875,
      "learning_rate": 4.68e-06,
      "loss": 4.0249,
      "step": 469
    },
    {
      "epoch": 3.012841091492777,
      "grad_norm": 1.328125,
      "learning_rate": 4.69e-06,
      "loss": 4.0274,
      "step": 470
    },
    {
      "epoch": 3.019261637239165,
      "grad_norm": 0.9375,
      "learning_rate": 4.7e-06,
      "loss": 4.0271,
      "step": 471
    },
    {
      "epoch": 3.0256821829855536,
      "grad_norm": 3.765625,
      "learning_rate": 4.710000000000001e-06,
      "loss": 4.038,
      "step": 472
    },
    {
      "epoch": 3.0321027287319424,
      "grad_norm": 3.78125,
      "learning_rate": 4.72e-06,
      "loss": 4.0421,
      "step": 473
    },
    {
      "epoch": 3.038523274478331,
      "grad_norm": 0.51171875,
      "learning_rate": 4.7300000000000005e-06,
      "loss": 4.0129,
      "step": 474
    },
    {
      "epoch": 3.044943820224719,
      "grad_norm": 0.52734375,
      "learning_rate": 4.74e-06,
      "loss": 4.0258,
      "step": 475
    },
    {
      "epoch": 3.0513643659711076,
      "grad_norm": 0.390625,
      "learning_rate": 4.75e-06,
      "loss": 4.0211,
      "step": 476
    },
    {
      "epoch": 3.057784911717496,
      "grad_norm": 0.81640625,
      "learning_rate": 4.76e-06,
      "loss": 4.0198,
      "step": 477
    },
    {
      "epoch": 3.0642054574638844,
      "grad_norm": 0.55078125,
      "learning_rate": 4.77e-06,
      "loss": 4.0285,
      "step": 478
    },
    {
      "epoch": 3.070626003210273,
      "grad_norm": 0.48828125,
      "learning_rate": 4.780000000000001e-06,
      "loss": 4.0266,
      "step": 479
    },
    {
      "epoch": 3.077046548956661,
      "grad_norm": 1.3671875,
      "learning_rate": 4.79e-06,
      "loss": 4.027,
      "step": 480
    },
    {
      "epoch": 3.0834670947030496,
      "grad_norm": 1.8203125,
      "learning_rate": 4.800000000000001e-06,
      "loss": 4.0227,
      "step": 481
    },
    {
      "epoch": 3.0898876404494384,
      "grad_norm": 4.3125,
      "learning_rate": 4.81e-06,
      "loss": 4.0377,
      "step": 482
    },
    {
      "epoch": 3.096308186195827,
      "grad_norm": 1.078125,
      "learning_rate": 4.8200000000000004e-06,
      "loss": 4.0194,
      "step": 483
    },
    {
      "epoch": 3.102728731942215,
      "grad_norm": 0.68359375,
      "learning_rate": 4.83e-06,
      "loss": 4.0224,
      "step": 484
    },
    {
      "epoch": 3.1091492776886036,
      "grad_norm": 0.67578125,
      "learning_rate": 4.84e-06,
      "loss": 4.0229,
      "step": 485
    },
    {
      "epoch": 3.115569823434992,
      "grad_norm": 0.5234375,
      "learning_rate": 4.85e-06,
      "loss": 4.0158,
      "step": 486
    },
    {
      "epoch": 3.1219903691813804,
      "grad_norm": 2.0625,
      "learning_rate": 4.86e-06,
      "loss": 4.0292,
      "step": 487
    },
    {
      "epoch": 3.128410914927769,
      "grad_norm": 1.8125,
      "learning_rate": 4.87e-06,
      "loss": 4.0381,
      "step": 488
    },
    {
      "epoch": 3.134831460674157,
      "grad_norm": 0.494140625,
      "learning_rate": 4.880000000000001e-06,
      "loss": 4.0215,
      "step": 489
    },
    {
      "epoch": 3.1412520064205456,
      "grad_norm": 0.408203125,
      "learning_rate": 4.89e-06,
      "loss": 4.0253,
      "step": 490
    },
    {
      "epoch": 3.1476725521669344,
      "grad_norm": 0.451171875,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 4.0134,
      "step": 491
    },
    {
      "epoch": 3.154093097913323,
      "grad_norm": 0.8515625,
      "learning_rate": 4.9100000000000004e-06,
      "loss": 4.0187,
      "step": 492
    },
    {
      "epoch": 3.160513643659711,
      "grad_norm": 0.470703125,
      "learning_rate": 4.92e-06,
      "loss": 4.0337,
      "step": 493
    },
    {
      "epoch": 3.1669341894060996,
      "grad_norm": 1.0703125,
      "learning_rate": 4.93e-06,
      "loss": 4.0126,
      "step": 494
    },
    {
      "epoch": 3.173354735152488,
      "grad_norm": 0.5859375,
      "learning_rate": 4.94e-06,
      "loss": 4.0192,
      "step": 495
    },
    {
      "epoch": 3.1797752808988764,
      "grad_norm": 0.412109375,
      "learning_rate": 4.950000000000001e-06,
      "loss": 4.0217,
      "step": 496
    },
    {
      "epoch": 3.186195826645265,
      "grad_norm": 0.546875,
      "learning_rate": 4.96e-06,
      "loss": 4.0153,
      "step": 497
    },
    {
      "epoch": 3.192616372391653,
      "grad_norm": 0.7578125,
      "learning_rate": 4.970000000000001e-06,
      "loss": 4.0224,
      "step": 498
    },
    {
      "epoch": 3.1990369181380416,
      "grad_norm": 0.4375,
      "learning_rate": 4.98e-06,
      "loss": 4.02,
      "step": 499
    },
    {
      "epoch": 3.20545746388443,
      "grad_norm": 1.359375,
      "learning_rate": 4.9900000000000005e-06,
      "loss": 4.0135,
      "step": 500
    },
    {
      "epoch": 3.211878009630819,
      "grad_norm": 0.439453125,
      "learning_rate": 5e-06,
      "loss": 4.0134,
      "step": 501
    },
    {
      "epoch": 3.218298555377207,
      "grad_norm": 0.40625,
      "learning_rate": 5.01e-06,
      "loss": 4.0141,
      "step": 502
    },
    {
      "epoch": 3.2247191011235956,
      "grad_norm": 9.9375,
      "learning_rate": 5.02e-06,
      "loss": 4.0259,
      "step": 503
    },
    {
      "epoch": 3.231139646869984,
      "grad_norm": 2.609375,
      "learning_rate": 5.03e-06,
      "loss": 4.0355,
      "step": 504
    },
    {
      "epoch": 3.2375601926163724,
      "grad_norm": 1.40625,
      "learning_rate": 5.04e-06,
      "loss": 4.0243,
      "step": 505
    },
    {
      "epoch": 3.243980738362761,
      "grad_norm": 1.0546875,
      "learning_rate": 5.050000000000001e-06,
      "loss": 4.0169,
      "step": 506
    },
    {
      "epoch": 3.250401284109149,
      "grad_norm": 2.703125,
      "learning_rate": 5.06e-06,
      "loss": 4.0265,
      "step": 507
    },
    {
      "epoch": 3.2568218298555376,
      "grad_norm": 2.0625,
      "learning_rate": 5.070000000000001e-06,
      "loss": 4.0142,
      "step": 508
    },
    {
      "epoch": 3.263242375601926,
      "grad_norm": 1.3203125,
      "learning_rate": 5.08e-06,
      "loss": 4.0258,
      "step": 509
    },
    {
      "epoch": 3.2696629213483144,
      "grad_norm": 1.265625,
      "learning_rate": 5.09e-06,
      "loss": 4.0256,
      "step": 510
    },
    {
      "epoch": 3.276083467094703,
      "grad_norm": 0.3125,
      "learning_rate": 5.1e-06,
      "loss": 4.0122,
      "step": 511
    },
    {
      "epoch": 3.2825040128410916,
      "grad_norm": 0.9921875,
      "learning_rate": 5.11e-06,
      "loss": 4.0187,
      "step": 512
    },
    {
      "epoch": 3.28892455858748,
      "grad_norm": 2.96875,
      "learning_rate": 5.12e-06,
      "loss": 4.0138,
      "step": 513
    },
    {
      "epoch": 3.2953451043338684,
      "grad_norm": 0.3359375,
      "learning_rate": 5.13e-06,
      "loss": 4.0197,
      "step": 514
    },
    {
      "epoch": 3.301765650080257,
      "grad_norm": 1.7421875,
      "learning_rate": 5.140000000000001e-06,
      "loss": 4.0196,
      "step": 515
    },
    {
      "epoch": 3.308186195826645,
      "grad_norm": 1.2265625,
      "learning_rate": 5.15e-06,
      "loss": 4.0113,
      "step": 516
    },
    {
      "epoch": 3.3146067415730336,
      "grad_norm": 1.109375,
      "learning_rate": 5.1600000000000006e-06,
      "loss": 4.0234,
      "step": 517
    },
    {
      "epoch": 3.321027287319422,
      "grad_norm": 0.271484375,
      "learning_rate": 5.1700000000000005e-06,
      "loss": 4.019,
      "step": 518
    },
    {
      "epoch": 3.3274478330658104,
      "grad_norm": 0.859375,
      "learning_rate": 5.18e-06,
      "loss": 4.0169,
      "step": 519
    },
    {
      "epoch": 3.333868378812199,
      "grad_norm": 1.2421875,
      "learning_rate": 5.19e-06,
      "loss": 4.0231,
      "step": 520
    },
    {
      "epoch": 3.3402889245585876,
      "grad_norm": 0.8515625,
      "learning_rate": 5.2e-06,
      "loss": 4.0191,
      "step": 521
    },
    {
      "epoch": 3.346709470304976,
      "grad_norm": 1.109375,
      "learning_rate": 5.21e-06,
      "loss": 4.0214,
      "step": 522
    },
    {
      "epoch": 3.3531300160513644,
      "grad_norm": 1.625,
      "learning_rate": 5.220000000000001e-06,
      "loss": 4.0198,
      "step": 523
    },
    {
      "epoch": 3.359550561797753,
      "grad_norm": 1.140625,
      "learning_rate": 5.23e-06,
      "loss": 4.0097,
      "step": 524
    },
    {
      "epoch": 3.365971107544141,
      "grad_norm": 0.58203125,
      "learning_rate": 5.240000000000001e-06,
      "loss": 4.0167,
      "step": 525
    },
    {
      "epoch": 3.3723916532905296,
      "grad_norm": 0.73828125,
      "learning_rate": 5.25e-06,
      "loss": 4.0154,
      "step": 526
    },
    {
      "epoch": 3.378812199036918,
      "grad_norm": 1.140625,
      "learning_rate": 5.2600000000000005e-06,
      "loss": 4.0199,
      "step": 527
    },
    {
      "epoch": 3.3852327447833064,
      "grad_norm": 0.6171875,
      "learning_rate": 5.2699999999999995e-06,
      "loss": 4.0207,
      "step": 528
    },
    {
      "epoch": 3.391653290529695,
      "grad_norm": 0.6328125,
      "learning_rate": 5.28e-06,
      "loss": 4.009,
      "step": 529
    },
    {
      "epoch": 3.3980738362760836,
      "grad_norm": 1.8359375,
      "learning_rate": 5.29e-06,
      "loss": 4.0319,
      "step": 530
    },
    {
      "epoch": 3.404494382022472,
      "grad_norm": 3.78125,
      "learning_rate": 5.3e-06,
      "loss": 4.0295,
      "step": 531
    },
    {
      "epoch": 3.4109149277688604,
      "grad_norm": 2.390625,
      "learning_rate": 5.31e-06,
      "loss": 4.019,
      "step": 532
    },
    {
      "epoch": 3.417335473515249,
      "grad_norm": 1.625,
      "learning_rate": 5.32e-06,
      "loss": 4.0179,
      "step": 533
    },
    {
      "epoch": 3.423756019261637,
      "grad_norm": 0.859375,
      "learning_rate": 5.330000000000001e-06,
      "loss": 4.0175,
      "step": 534
    },
    {
      "epoch": 3.4301765650080256,
      "grad_norm": 1.1953125,
      "learning_rate": 5.3400000000000005e-06,
      "loss": 4.0141,
      "step": 535
    },
    {
      "epoch": 3.436597110754414,
      "grad_norm": 1.75,
      "learning_rate": 5.3500000000000004e-06,
      "loss": 4.0211,
      "step": 536
    },
    {
      "epoch": 3.4430176565008024,
      "grad_norm": 9.625,
      "learning_rate": 5.36e-06,
      "loss": 4.0307,
      "step": 537
    },
    {
      "epoch": 3.449438202247191,
      "grad_norm": 0.2890625,
      "learning_rate": 5.37e-06,
      "loss": 4.014,
      "step": 538
    },
    {
      "epoch": 3.4558587479935796,
      "grad_norm": 2.078125,
      "learning_rate": 5.38e-06,
      "loss": 4.0265,
      "step": 539
    },
    {
      "epoch": 3.462279293739968,
      "grad_norm": 0.8515625,
      "learning_rate": 5.390000000000001e-06,
      "loss": 4.0136,
      "step": 540
    },
    {
      "epoch": 3.4686998394863564,
      "grad_norm": 0.68359375,
      "learning_rate": 5.4e-06,
      "loss": 4.0147,
      "step": 541
    },
    {
      "epoch": 3.475120385232745,
      "grad_norm": 0.3515625,
      "learning_rate": 5.410000000000001e-06,
      "loss": 4.008,
      "step": 542
    },
    {
      "epoch": 3.481540930979133,
      "grad_norm": 3.09375,
      "learning_rate": 5.42e-06,
      "loss": 4.0181,
      "step": 543
    },
    {
      "epoch": 3.4879614767255216,
      "grad_norm": 0.90234375,
      "learning_rate": 5.4300000000000005e-06,
      "loss": 4.0129,
      "step": 544
    },
    {
      "epoch": 3.49438202247191,
      "grad_norm": 0.431640625,
      "learning_rate": 5.44e-06,
      "loss": 4.0201,
      "step": 545
    },
    {
      "epoch": 3.5008025682182984,
      "grad_norm": 0.306640625,
      "learning_rate": 5.45e-06,
      "loss": 4.015,
      "step": 546
    },
    {
      "epoch": 3.5072231139646872,
      "grad_norm": 1.796875,
      "learning_rate": 5.46e-06,
      "loss": 4.0205,
      "step": 547
    },
    {
      "epoch": 3.513643659711075,
      "grad_norm": 0.63671875,
      "learning_rate": 5.47e-06,
      "loss": 4.0128,
      "step": 548
    },
    {
      "epoch": 3.520064205457464,
      "grad_norm": 1.5390625,
      "learning_rate": 5.48e-06,
      "loss": 4.0154,
      "step": 549
    },
    {
      "epoch": 3.5264847512038524,
      "grad_norm": 0.52734375,
      "learning_rate": 5.49e-06,
      "loss": 4.0104,
      "step": 550
    },
    {
      "epoch": 3.532905296950241,
      "grad_norm": 0.93359375,
      "learning_rate": 5.500000000000001e-06,
      "loss": 4.0234,
      "step": 551
    },
    {
      "epoch": 3.539325842696629,
      "grad_norm": 0.63671875,
      "learning_rate": 5.510000000000001e-06,
      "loss": 4.0222,
      "step": 552
    },
    {
      "epoch": 3.5457463884430176,
      "grad_norm": 0.396484375,
      "learning_rate": 5.5200000000000005e-06,
      "loss": 4.0119,
      "step": 553
    },
    {
      "epoch": 3.552166934189406,
      "grad_norm": 0.29296875,
      "learning_rate": 5.53e-06,
      "loss": 4.0068,
      "step": 554
    },
    {
      "epoch": 3.5585874799357944,
      "grad_norm": 0.6796875,
      "learning_rate": 5.54e-06,
      "loss": 4.0321,
      "step": 555
    },
    {
      "epoch": 3.5650080256821832,
      "grad_norm": 3.390625,
      "learning_rate": 5.55e-06,
      "loss": 4.0225,
      "step": 556
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.47265625,
      "learning_rate": 5.56e-06,
      "loss": 4.0098,
      "step": 557
    },
    {
      "epoch": 3.57784911717496,
      "grad_norm": 0.8125,
      "learning_rate": 5.57e-06,
      "loss": 4.0161,
      "step": 558
    },
    {
      "epoch": 3.5842696629213484,
      "grad_norm": 0.89453125,
      "learning_rate": 5.580000000000001e-06,
      "loss": 4.0264,
      "step": 559
    },
    {
      "epoch": 3.590690208667737,
      "grad_norm": 0.59765625,
      "learning_rate": 5.59e-06,
      "loss": 4.0161,
      "step": 560
    },
    {
      "epoch": 3.597110754414125,
      "grad_norm": 0.435546875,
      "learning_rate": 5.600000000000001e-06,
      "loss": 4.0194,
      "step": 561
    },
    {
      "epoch": 3.6035313001605136,
      "grad_norm": 0.4609375,
      "learning_rate": 5.61e-06,
      "loss": 4.0109,
      "step": 562
    },
    {
      "epoch": 3.609951845906902,
      "grad_norm": 0.62109375,
      "learning_rate": 5.62e-06,
      "loss": 4.0066,
      "step": 563
    },
    {
      "epoch": 3.6163723916532904,
      "grad_norm": 0.71484375,
      "learning_rate": 5.63e-06,
      "loss": 4.0181,
      "step": 564
    },
    {
      "epoch": 3.6227929373996792,
      "grad_norm": 0.294921875,
      "learning_rate": 5.64e-06,
      "loss": 4.0082,
      "step": 565
    },
    {
      "epoch": 3.629213483146067,
      "grad_norm": 0.2333984375,
      "learning_rate": 5.65e-06,
      "loss": 4.0158,
      "step": 566
    },
    {
      "epoch": 3.635634028892456,
      "grad_norm": 0.4296875,
      "learning_rate": 5.66e-06,
      "loss": 4.0125,
      "step": 567
    },
    {
      "epoch": 3.6420545746388444,
      "grad_norm": 0.609375,
      "learning_rate": 5.67e-06,
      "loss": 4.0087,
      "step": 568
    },
    {
      "epoch": 3.648475120385233,
      "grad_norm": 0.58984375,
      "learning_rate": 5.680000000000001e-06,
      "loss": 4.017,
      "step": 569
    },
    {
      "epoch": 3.654895666131621,
      "grad_norm": 0.20703125,
      "learning_rate": 5.690000000000001e-06,
      "loss": 4.012,
      "step": 570
    },
    {
      "epoch": 3.6613162118780096,
      "grad_norm": 0.474609375,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 4.0107,
      "step": 571
    },
    {
      "epoch": 3.667736757624398,
      "grad_norm": 0.2275390625,
      "learning_rate": 5.71e-06,
      "loss": 4.0076,
      "step": 572
    },
    {
      "epoch": 3.6741573033707864,
      "grad_norm": 0.30078125,
      "learning_rate": 5.72e-06,
      "loss": 4.0077,
      "step": 573
    },
    {
      "epoch": 3.6805778491171752,
      "grad_norm": 2.546875,
      "learning_rate": 5.73e-06,
      "loss": 4.0175,
      "step": 574
    },
    {
      "epoch": 3.686998394863563,
      "grad_norm": 0.84765625,
      "learning_rate": 5.74e-06,
      "loss": 4.0116,
      "step": 575
    },
    {
      "epoch": 3.693418940609952,
      "grad_norm": 2.265625,
      "learning_rate": 5.750000000000001e-06,
      "loss": 4.0128,
      "step": 576
    },
    {
      "epoch": 3.6998394863563404,
      "grad_norm": 0.44140625,
      "learning_rate": 5.76e-06,
      "loss": 4.0142,
      "step": 577
    },
    {
      "epoch": 3.706260032102729,
      "grad_norm": 0.6640625,
      "learning_rate": 5.770000000000001e-06,
      "loss": 4.0121,
      "step": 578
    },
    {
      "epoch": 3.712680577849117,
      "grad_norm": 0.2734375,
      "learning_rate": 5.78e-06,
      "loss": 4.0078,
      "step": 579
    },
    {
      "epoch": 3.7191011235955056,
      "grad_norm": 0.51953125,
      "learning_rate": 5.7900000000000005e-06,
      "loss": 4.0158,
      "step": 580
    },
    {
      "epoch": 3.725521669341894,
      "grad_norm": 0.54296875,
      "learning_rate": 5.8e-06,
      "loss": 4.0111,
      "step": 581
    },
    {
      "epoch": 3.7319422150882824,
      "grad_norm": 0.296875,
      "learning_rate": 5.81e-06,
      "loss": 4.0154,
      "step": 582
    },
    {
      "epoch": 3.738362760834671,
      "grad_norm": 0.33984375,
      "learning_rate": 5.82e-06,
      "loss": 4.011,
      "step": 583
    },
    {
      "epoch": 3.744783306581059,
      "grad_norm": 0.384765625,
      "learning_rate": 5.83e-06,
      "loss": 4.0218,
      "step": 584
    },
    {
      "epoch": 3.751203852327448,
      "grad_norm": 0.23046875,
      "learning_rate": 5.84e-06,
      "loss": 4.0145,
      "step": 585
    },
    {
      "epoch": 3.7576243980738364,
      "grad_norm": 0.671875,
      "learning_rate": 5.850000000000001e-06,
      "loss": 4.0126,
      "step": 586
    },
    {
      "epoch": 3.764044943820225,
      "grad_norm": 0.55078125,
      "learning_rate": 5.86e-06,
      "loss": 4.0095,
      "step": 587
    },
    {
      "epoch": 3.770465489566613,
      "grad_norm": 0.287109375,
      "learning_rate": 5.8700000000000005e-06,
      "loss": 4.0191,
      "step": 588
    },
    {
      "epoch": 3.7768860353130016,
      "grad_norm": 0.50390625,
      "learning_rate": 5.8800000000000005e-06,
      "loss": 4.0248,
      "step": 589
    },
    {
      "epoch": 3.78330658105939,
      "grad_norm": 0.78515625,
      "learning_rate": 5.89e-06,
      "loss": 4.0088,
      "step": 590
    },
    {
      "epoch": 3.7897271268057784,
      "grad_norm": 1.421875,
      "learning_rate": 5.9e-06,
      "loss": 4.011,
      "step": 591
    },
    {
      "epoch": 3.796147672552167,
      "grad_norm": 0.45703125,
      "learning_rate": 5.91e-06,
      "loss": 4.0124,
      "step": 592
    },
    {
      "epoch": 3.802568218298555,
      "grad_norm": 0.5234375,
      "learning_rate": 5.920000000000001e-06,
      "loss": 4.0069,
      "step": 593
    },
    {
      "epoch": 3.808988764044944,
      "grad_norm": 0.890625,
      "learning_rate": 5.93e-06,
      "loss": 4.0158,
      "step": 594
    },
    {
      "epoch": 3.8154093097913324,
      "grad_norm": 1.4609375,
      "learning_rate": 5.940000000000001e-06,
      "loss": 4.0121,
      "step": 595
    },
    {
      "epoch": 3.821829855537721,
      "grad_norm": 0.73046875,
      "learning_rate": 5.95e-06,
      "loss": 4.0066,
      "step": 596
    },
    {
      "epoch": 3.828250401284109,
      "grad_norm": 0.73046875,
      "learning_rate": 5.9600000000000005e-06,
      "loss": 4.0177,
      "step": 597
    },
    {
      "epoch": 3.8346709470304976,
      "grad_norm": 0.40625,
      "learning_rate": 5.9700000000000004e-06,
      "loss": 4.0061,
      "step": 598
    },
    {
      "epoch": 3.841091492776886,
      "grad_norm": 1.7109375,
      "learning_rate": 5.98e-06,
      "loss": 4.0124,
      "step": 599
    },
    {
      "epoch": 3.8475120385232744,
      "grad_norm": 0.47265625,
      "learning_rate": 5.99e-06,
      "loss": 4.0217,
      "step": 600
    },
    {
      "epoch": 3.853932584269663,
      "grad_norm": 2.921875,
      "learning_rate": 6e-06,
      "loss": 4.0103,
      "step": 601
    },
    {
      "epoch": 3.860353130016051,
      "grad_norm": 0.359375,
      "learning_rate": 6.01e-06,
      "loss": 4.0133,
      "step": 602
    },
    {
      "epoch": 3.86677367576244,
      "grad_norm": 0.294921875,
      "learning_rate": 6.02e-06,
      "loss": 4.0106,
      "step": 603
    },
    {
      "epoch": 3.8731942215088284,
      "grad_norm": 0.240234375,
      "learning_rate": 6.03e-06,
      "loss": 4.0058,
      "step": 604
    },
    {
      "epoch": 3.879614767255217,
      "grad_norm": 0.58203125,
      "learning_rate": 6.040000000000001e-06,
      "loss": 4.0097,
      "step": 605
    },
    {
      "epoch": 3.886035313001605,
      "grad_norm": 0.5,
      "learning_rate": 6.0500000000000005e-06,
      "loss": 4.0071,
      "step": 606
    },
    {
      "epoch": 3.8924558587479936,
      "grad_norm": 0.59375,
      "learning_rate": 6.0600000000000004e-06,
      "loss": 4.0286,
      "step": 607
    },
    {
      "epoch": 3.898876404494382,
      "grad_norm": 0.37109375,
      "learning_rate": 6.07e-06,
      "loss": 4.012,
      "step": 608
    },
    {
      "epoch": 3.9052969502407704,
      "grad_norm": 0.5546875,
      "learning_rate": 6.08e-06,
      "loss": 4.0098,
      "step": 609
    },
    {
      "epoch": 3.911717495987159,
      "grad_norm": 0.30078125,
      "learning_rate": 6.090000000000001e-06,
      "loss": 4.0113,
      "step": 610
    },
    {
      "epoch": 3.918138041733547,
      "grad_norm": 0.474609375,
      "learning_rate": 6.1e-06,
      "loss": 4.0097,
      "step": 611
    },
    {
      "epoch": 3.924558587479936,
      "grad_norm": 0.57421875,
      "learning_rate": 6.110000000000001e-06,
      "loss": 4.0074,
      "step": 612
    },
    {
      "epoch": 3.930979133226324,
      "grad_norm": 0.9453125,
      "learning_rate": 6.12e-06,
      "loss": 4.0156,
      "step": 613
    },
    {
      "epoch": 3.937399678972713,
      "grad_norm": 7.46875,
      "learning_rate": 6.130000000000001e-06,
      "loss": 4.0204,
      "step": 614
    },
    {
      "epoch": 3.943820224719101,
      "grad_norm": 1.8671875,
      "learning_rate": 6.1400000000000005e-06,
      "loss": 4.0124,
      "step": 615
    },
    {
      "epoch": 3.9502407704654896,
      "grad_norm": 1.8203125,
      "learning_rate": 6.15e-06,
      "loss": 4.0145,
      "step": 616
    },
    {
      "epoch": 3.956661316211878,
      "grad_norm": 0.71484375,
      "learning_rate": 6.16e-06,
      "loss": 4.0121,
      "step": 617
    },
    {
      "epoch": 3.9630818619582664,
      "grad_norm": 2.328125,
      "learning_rate": 6.17e-06,
      "loss": 4.0109,
      "step": 618
    },
    {
      "epoch": 3.969502407704655,
      "grad_norm": 2.625,
      "learning_rate": 6.18e-06,
      "loss": 4.0157,
      "step": 619
    },
    {
      "epoch": 3.975922953451043,
      "grad_norm": 2.3125,
      "learning_rate": 6.19e-06,
      "loss": 4.0093,
      "step": 620
    },
    {
      "epoch": 3.982343499197432,
      "grad_norm": 0.30078125,
      "learning_rate": 6.2e-06,
      "loss": 4.0112,
      "step": 621
    },
    {
      "epoch": 3.98876404494382,
      "grad_norm": 1.6484375,
      "learning_rate": 6.210000000000001e-06,
      "loss": 4.0053,
      "step": 622
    },
    {
      "epoch": 3.995184590690209,
      "grad_norm": 1.828125,
      "learning_rate": 6.22e-06,
      "loss": 4.0146,
      "step": 623
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0078125,
      "learning_rate": 6.2300000000000005e-06,
      "loss": 3.0102,
      "step": 624
    },
    {
      "epoch": 4.006420545746389,
      "grad_norm": 0.78515625,
      "learning_rate": 6.24e-06,
      "loss": 4.0109,
      "step": 625
    },
    {
      "epoch": 4.012841091492777,
      "grad_norm": 0.96875,
      "learning_rate": 6.25e-06,
      "loss": 4.0131,
      "step": 626
    },
    {
      "epoch": 4.019261637239166,
      "grad_norm": 0.6796875,
      "learning_rate": 6.26e-06,
      "loss": 4.0107,
      "step": 627
    },
    {
      "epoch": 4.025682182985554,
      "grad_norm": 0.291015625,
      "learning_rate": 6.270000000000001e-06,
      "loss": 4.0101,
      "step": 628
    },
    {
      "epoch": 4.032102728731942,
      "grad_norm": 1.484375,
      "learning_rate": 6.28e-06,
      "loss": 4.01,
      "step": 629
    },
    {
      "epoch": 4.03852327447833,
      "grad_norm": 0.96484375,
      "learning_rate": 6.29e-06,
      "loss": 4.0079,
      "step": 630
    },
    {
      "epoch": 4.044943820224719,
      "grad_norm": 0.486328125,
      "learning_rate": 6.300000000000001e-06,
      "loss": 4.0078,
      "step": 631
    },
    {
      "epoch": 4.051364365971107,
      "grad_norm": 1.359375,
      "learning_rate": 6.3100000000000006e-06,
      "loss": 4.0074,
      "step": 632
    },
    {
      "epoch": 4.057784911717496,
      "grad_norm": 1.3828125,
      "learning_rate": 6.320000000000001e-06,
      "loss": 4.0134,
      "step": 633
    },
    {
      "epoch": 4.064205457463885,
      "grad_norm": 1.1640625,
      "learning_rate": 6.3299999999999995e-06,
      "loss": 4.011,
      "step": 634
    },
    {
      "epoch": 4.070626003210273,
      "grad_norm": 1.328125,
      "learning_rate": 6.34e-06,
      "loss": 4.0063,
      "step": 635
    },
    {
      "epoch": 4.077046548956662,
      "grad_norm": 1.09375,
      "learning_rate": 6.35e-06,
      "loss": 4.0073,
      "step": 636
    },
    {
      "epoch": 4.08346709470305,
      "grad_norm": 0.7265625,
      "learning_rate": 6.360000000000001e-06,
      "loss": 4.0052,
      "step": 637
    },
    {
      "epoch": 4.089887640449438,
      "grad_norm": 1.1484375,
      "learning_rate": 6.370000000000001e-06,
      "loss": 4.0217,
      "step": 638
    },
    {
      "epoch": 4.096308186195826,
      "grad_norm": 1.0859375,
      "learning_rate": 6.38e-06,
      "loss": 4.0093,
      "step": 639
    },
    {
      "epoch": 4.102728731942215,
      "grad_norm": 0.65625,
      "learning_rate": 6.39e-06,
      "loss": 4.012,
      "step": 640
    },
    {
      "epoch": 4.109149277688603,
      "grad_norm": 0.73046875,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 4.0064,
      "step": 641
    },
    {
      "epoch": 4.115569823434992,
      "grad_norm": 1.0546875,
      "learning_rate": 6.4100000000000005e-06,
      "loss": 4.0069,
      "step": 642
    },
    {
      "epoch": 4.121990369181381,
      "grad_norm": 0.93359375,
      "learning_rate": 6.4199999999999995e-06,
      "loss": 4.0122,
      "step": 643
    },
    {
      "epoch": 4.128410914927769,
      "grad_norm": 0.9453125,
      "learning_rate": 6.43e-06,
      "loss": 4.0122,
      "step": 644
    },
    {
      "epoch": 4.134831460674158,
      "grad_norm": 1.078125,
      "learning_rate": 6.44e-06,
      "loss": 4.0102,
      "step": 645
    },
    {
      "epoch": 4.141252006420546,
      "grad_norm": 1.1875,
      "learning_rate": 6.45e-06,
      "loss": 4.0141,
      "step": 646
    },
    {
      "epoch": 4.147672552166934,
      "grad_norm": 0.25390625,
      "learning_rate": 6.460000000000001e-06,
      "loss": 4.0103,
      "step": 647
    },
    {
      "epoch": 4.154093097913322,
      "grad_norm": 0.76953125,
      "learning_rate": 6.47e-06,
      "loss": 4.0106,
      "step": 648
    },
    {
      "epoch": 4.160513643659711,
      "grad_norm": 0.72265625,
      "learning_rate": 6.48e-06,
      "loss": 4.0116,
      "step": 649
    },
    {
      "epoch": 4.166934189406099,
      "grad_norm": 0.2197265625,
      "learning_rate": 6.4900000000000005e-06,
      "loss": 4.0083,
      "step": 650
    },
    {
      "epoch": 4.173354735152488,
      "grad_norm": 0.376953125,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 4.0085,
      "step": 651
    },
    {
      "epoch": 4.179775280898877,
      "grad_norm": 0.80859375,
      "learning_rate": 6.510000000000001e-06,
      "loss": 4.0089,
      "step": 652
    },
    {
      "epoch": 4.186195826645265,
      "grad_norm": 0.62890625,
      "learning_rate": 6.519999999999999e-06,
      "loss": 4.0127,
      "step": 653
    },
    {
      "epoch": 4.192616372391654,
      "grad_norm": 0.2490234375,
      "learning_rate": 6.53e-06,
      "loss": 4.0119,
      "step": 654
    },
    {
      "epoch": 4.199036918138042,
      "grad_norm": 0.69921875,
      "learning_rate": 6.54e-06,
      "loss": 4.0028,
      "step": 655
    },
    {
      "epoch": 4.20545746388443,
      "grad_norm": 1.7578125,
      "learning_rate": 6.550000000000001e-06,
      "loss": 4.0063,
      "step": 656
    },
    {
      "epoch": 4.211878009630818,
      "grad_norm": 3.109375,
      "learning_rate": 6.560000000000001e-06,
      "loss": 4.0168,
      "step": 657
    },
    {
      "epoch": 4.218298555377207,
      "grad_norm": 0.330078125,
      "learning_rate": 6.57e-06,
      "loss": 4.0103,
      "step": 658
    },
    {
      "epoch": 4.224719101123595,
      "grad_norm": 4.53125,
      "learning_rate": 6.58e-06,
      "loss": 4.0091,
      "step": 659
    },
    {
      "epoch": 4.231139646869984,
      "grad_norm": 0.7265625,
      "learning_rate": 6.5900000000000004e-06,
      "loss": 4.0129,
      "step": 660
    },
    {
      "epoch": 4.237560192616373,
      "grad_norm": 0.291015625,
      "learning_rate": 6.6e-06,
      "loss": 4.0081,
      "step": 661
    },
    {
      "epoch": 4.243980738362761,
      "grad_norm": 0.318359375,
      "learning_rate": 6.610000000000001e-06,
      "loss": 4.0057,
      "step": 662
    },
    {
      "epoch": 4.25040128410915,
      "grad_norm": 0.26171875,
      "learning_rate": 6.62e-06,
      "loss": 4.0117,
      "step": 663
    },
    {
      "epoch": 4.256821829855538,
      "grad_norm": 0.494140625,
      "learning_rate": 6.63e-06,
      "loss": 4.005,
      "step": 664
    },
    {
      "epoch": 4.263242375601926,
      "grad_norm": 8.4375,
      "learning_rate": 6.640000000000001e-06,
      "loss": 4.0234,
      "step": 665
    },
    {
      "epoch": 4.269662921348314,
      "grad_norm": 0.484375,
      "learning_rate": 6.650000000000001e-06,
      "loss": 4.0096,
      "step": 666
    },
    {
      "epoch": 4.276083467094703,
      "grad_norm": 0.232421875,
      "learning_rate": 6.660000000000001e-06,
      "loss": 4.0053,
      "step": 667
    },
    {
      "epoch": 4.282504012841091,
      "grad_norm": 0.65625,
      "learning_rate": 6.67e-06,
      "loss": 4.0104,
      "step": 668
    },
    {
      "epoch": 4.28892455858748,
      "grad_norm": 0.48046875,
      "learning_rate": 6.68e-06,
      "loss": 4.0147,
      "step": 669
    },
    {
      "epoch": 4.295345104333869,
      "grad_norm": 0.1962890625,
      "learning_rate": 6.69e-06,
      "loss": 4.0083,
      "step": 670
    },
    {
      "epoch": 4.301765650080257,
      "grad_norm": 0.330078125,
      "learning_rate": 6.700000000000001e-06,
      "loss": 4.0051,
      "step": 671
    },
    {
      "epoch": 4.308186195826646,
      "grad_norm": 0.63671875,
      "learning_rate": 6.710000000000001e-06,
      "loss": 4.0153,
      "step": 672
    },
    {
      "epoch": 4.314606741573034,
      "grad_norm": 0.423828125,
      "learning_rate": 6.72e-06,
      "loss": 4.0093,
      "step": 673
    },
    {
      "epoch": 4.321027287319422,
      "grad_norm": 0.4296875,
      "learning_rate": 6.73e-06,
      "loss": 4.0085,
      "step": 674
    },
    {
      "epoch": 4.32744783306581,
      "grad_norm": 0.33984375,
      "learning_rate": 6.740000000000001e-06,
      "loss": 4.0037,
      "step": 675
    },
    {
      "epoch": 4.333868378812199,
      "grad_norm": 1.3359375,
      "learning_rate": 6.750000000000001e-06,
      "loss": 4.0102,
      "step": 676
    },
    {
      "epoch": 4.340288924558587,
      "grad_norm": 0.5390625,
      "learning_rate": 6.76e-06,
      "loss": 4.0097,
      "step": 677
    },
    {
      "epoch": 4.346709470304976,
      "grad_norm": 0.2353515625,
      "learning_rate": 6.7699999999999996e-06,
      "loss": 4.014,
      "step": 678
    },
    {
      "epoch": 4.353130016051364,
      "grad_norm": 0.2353515625,
      "learning_rate": 6.78e-06,
      "loss": 4.0085,
      "step": 679
    },
    {
      "epoch": 4.359550561797753,
      "grad_norm": 0.259765625,
      "learning_rate": 6.79e-06,
      "loss": 4.0087,
      "step": 680
    },
    {
      "epoch": 4.365971107544142,
      "grad_norm": 0.5234375,
      "learning_rate": 6.800000000000001e-06,
      "loss": 4.0197,
      "step": 681
    },
    {
      "epoch": 4.37239165329053,
      "grad_norm": 0.5625,
      "learning_rate": 6.81e-06,
      "loss": 4.0048,
      "step": 682
    },
    {
      "epoch": 4.378812199036918,
      "grad_norm": 0.373046875,
      "learning_rate": 6.82e-06,
      "loss": 4.0038,
      "step": 683
    },
    {
      "epoch": 4.385232744783306,
      "grad_norm": 0.2255859375,
      "learning_rate": 6.830000000000001e-06,
      "loss": 4.0082,
      "step": 684
    },
    {
      "epoch": 4.391653290529695,
      "grad_norm": 2.21875,
      "learning_rate": 6.840000000000001e-06,
      "loss": 4.0132,
      "step": 685
    },
    {
      "epoch": 4.398073836276083,
      "grad_norm": 0.71484375,
      "learning_rate": 6.8500000000000005e-06,
      "loss": 4.0077,
      "step": 686
    },
    {
      "epoch": 4.404494382022472,
      "grad_norm": 1.6484375,
      "learning_rate": 6.8599999999999995e-06,
      "loss": 4.0057,
      "step": 687
    },
    {
      "epoch": 4.41091492776886,
      "grad_norm": 0.484375,
      "learning_rate": 6.87e-06,
      "loss": 4.0099,
      "step": 688
    },
    {
      "epoch": 4.417335473515249,
      "grad_norm": 0.4296875,
      "learning_rate": 6.88e-06,
      "loss": 4.0104,
      "step": 689
    },
    {
      "epoch": 4.423756019261638,
      "grad_norm": 0.6015625,
      "learning_rate": 6.890000000000001e-06,
      "loss": 4.0078,
      "step": 690
    },
    {
      "epoch": 4.430176565008026,
      "grad_norm": 1.515625,
      "learning_rate": 6.900000000000001e-06,
      "loss": 4.0193,
      "step": 691
    },
    {
      "epoch": 4.436597110754414,
      "grad_norm": 0.7578125,
      "learning_rate": 6.91e-06,
      "loss": 4.0097,
      "step": 692
    },
    {
      "epoch": 4.443017656500802,
      "grad_norm": 1.421875,
      "learning_rate": 6.92e-06,
      "loss": 4.0032,
      "step": 693
    },
    {
      "epoch": 4.449438202247191,
      "grad_norm": 3.03125,
      "learning_rate": 6.9300000000000006e-06,
      "loss": 4.0085,
      "step": 694
    },
    {
      "epoch": 4.455858747993579,
      "grad_norm": 0.1767578125,
      "learning_rate": 6.9400000000000005e-06,
      "loss": 4.003,
      "step": 695
    },
    {
      "epoch": 4.462279293739968,
      "grad_norm": 1.234375,
      "learning_rate": 6.950000000000001e-06,
      "loss": 4.0136,
      "step": 696
    },
    {
      "epoch": 4.468699839486356,
      "grad_norm": 0.52734375,
      "learning_rate": 6.9599999999999994e-06,
      "loss": 4.0083,
      "step": 697
    },
    {
      "epoch": 4.475120385232745,
      "grad_norm": 0.2041015625,
      "learning_rate": 6.97e-06,
      "loss": 4.0069,
      "step": 698
    },
    {
      "epoch": 4.481540930979134,
      "grad_norm": 1.0625,
      "learning_rate": 6.98e-06,
      "loss": 4.0071,
      "step": 699
    },
    {
      "epoch": 4.487961476725522,
      "grad_norm": 2.953125,
      "learning_rate": 6.990000000000001e-06,
      "loss": 4.0221,
      "step": 700
    },
    {
      "epoch": 4.49438202247191,
      "grad_norm": 0.369140625,
      "learning_rate": 7.000000000000001e-06,
      "loss": 4.0132,
      "step": 701
    },
    {
      "epoch": 4.500802568218298,
      "grad_norm": 0.61328125,
      "learning_rate": 7.01e-06,
      "loss": 4.0132,
      "step": 702
    },
    {
      "epoch": 4.507223113964687,
      "grad_norm": 0.4140625,
      "learning_rate": 7.0200000000000006e-06,
      "loss": 4.0165,
      "step": 703
    },
    {
      "epoch": 4.513643659711075,
      "grad_norm": 0.1923828125,
      "learning_rate": 7.0300000000000005e-06,
      "loss": 4.004,
      "step": 704
    },
    {
      "epoch": 4.520064205457464,
      "grad_norm": 0.77734375,
      "learning_rate": 7.04e-06,
      "loss": 4.0098,
      "step": 705
    },
    {
      "epoch": 4.526484751203852,
      "grad_norm": 0.337890625,
      "learning_rate": 7.049999999999999e-06,
      "loss": 4.0146,
      "step": 706
    },
    {
      "epoch": 4.532905296950241,
      "grad_norm": 0.24609375,
      "learning_rate": 7.06e-06,
      "loss": 4.0041,
      "step": 707
    },
    {
      "epoch": 4.539325842696629,
      "grad_norm": 0.78125,
      "learning_rate": 7.07e-06,
      "loss": 4.0072,
      "step": 708
    },
    {
      "epoch": 4.545746388443018,
      "grad_norm": 0.50390625,
      "learning_rate": 7.080000000000001e-06,
      "loss": 4.0135,
      "step": 709
    },
    {
      "epoch": 4.552166934189406,
      "grad_norm": 0.2470703125,
      "learning_rate": 7.090000000000001e-06,
      "loss": 4.0083,
      "step": 710
    },
    {
      "epoch": 4.558587479935794,
      "grad_norm": 0.59765625,
      "learning_rate": 7.1e-06,
      "loss": 4.0044,
      "step": 711
    },
    {
      "epoch": 4.565008025682183,
      "grad_norm": 0.91796875,
      "learning_rate": 7.11e-06,
      "loss": 4.0107,
      "step": 712
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.19921875,
      "learning_rate": 7.1200000000000004e-06,
      "loss": 4.0126,
      "step": 713
    },
    {
      "epoch": 4.57784911717496,
      "grad_norm": 0.279296875,
      "learning_rate": 7.13e-06,
      "loss": 4.0035,
      "step": 714
    },
    {
      "epoch": 4.584269662921348,
      "grad_norm": 5.59375,
      "learning_rate": 7.140000000000001e-06,
      "loss": 4.0094,
      "step": 715
    },
    {
      "epoch": 4.590690208667737,
      "grad_norm": 0.9609375,
      "learning_rate": 7.15e-06,
      "loss": 4.0077,
      "step": 716
    },
    {
      "epoch": 4.597110754414125,
      "grad_norm": 0.4921875,
      "learning_rate": 7.16e-06,
      "loss": 4.0057,
      "step": 717
    },
    {
      "epoch": 4.603531300160514,
      "grad_norm": 3.0625,
      "learning_rate": 7.17e-06,
      "loss": 4.0071,
      "step": 718
    },
    {
      "epoch": 4.609951845906902,
      "grad_norm": 1.15625,
      "learning_rate": 7.180000000000001e-06,
      "loss": 4.0041,
      "step": 719
    },
    {
      "epoch": 4.61637239165329,
      "grad_norm": 6.0,
      "learning_rate": 7.190000000000001e-06,
      "loss": 4.0139,
      "step": 720
    },
    {
      "epoch": 4.622792937399679,
      "grad_norm": 0.7734375,
      "learning_rate": 7.2e-06,
      "loss": 4.007,
      "step": 721
    },
    {
      "epoch": 4.629213483146067,
      "grad_norm": 0.76953125,
      "learning_rate": 7.2100000000000004e-06,
      "loss": 4.0058,
      "step": 722
    },
    {
      "epoch": 4.635634028892456,
      "grad_norm": 6.125,
      "learning_rate": 7.22e-06,
      "loss": 4.0126,
      "step": 723
    },
    {
      "epoch": 4.642054574638844,
      "grad_norm": 0.94921875,
      "learning_rate": 7.230000000000001e-06,
      "loss": 4.009,
      "step": 724
    },
    {
      "epoch": 4.648475120385233,
      "grad_norm": 0.4609375,
      "learning_rate": 7.240000000000001e-06,
      "loss": 4.0148,
      "step": 725
    },
    {
      "epoch": 4.654895666131621,
      "grad_norm": 1.015625,
      "learning_rate": 7.25e-06,
      "loss": 4.0179,
      "step": 726
    },
    {
      "epoch": 4.66131621187801,
      "grad_norm": 0.89453125,
      "learning_rate": 7.26e-06,
      "loss": 4.0081,
      "step": 727
    },
    {
      "epoch": 4.667736757624398,
      "grad_norm": 0.2197265625,
      "learning_rate": 7.270000000000001e-06,
      "loss": 4.01,
      "step": 728
    },
    {
      "epoch": 4.674157303370786,
      "grad_norm": 1.0859375,
      "learning_rate": 7.280000000000001e-06,
      "loss": 4.0105,
      "step": 729
    },
    {
      "epoch": 4.680577849117175,
      "grad_norm": 0.46484375,
      "learning_rate": 7.290000000000001e-06,
      "loss": 4.0086,
      "step": 730
    },
    {
      "epoch": 4.686998394863563,
      "grad_norm": 1.2890625,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 4.0082,
      "step": 731
    },
    {
      "epoch": 4.693418940609952,
      "grad_norm": 0.72265625,
      "learning_rate": 7.31e-06,
      "loss": 4.0032,
      "step": 732
    },
    {
      "epoch": 4.69983948635634,
      "grad_norm": 0.953125,
      "learning_rate": 7.32e-06,
      "loss": 4.0189,
      "step": 733
    },
    {
      "epoch": 4.706260032102729,
      "grad_norm": 0.259765625,
      "learning_rate": 7.330000000000001e-06,
      "loss": 4.0119,
      "step": 734
    },
    {
      "epoch": 4.712680577849117,
      "grad_norm": 0.765625,
      "learning_rate": 7.340000000000001e-06,
      "loss": 4.0055,
      "step": 735
    },
    {
      "epoch": 4.719101123595506,
      "grad_norm": 0.271484375,
      "learning_rate": 7.35e-06,
      "loss": 4.0092,
      "step": 736
    },
    {
      "epoch": 4.725521669341894,
      "grad_norm": 0.279296875,
      "learning_rate": 7.36e-06,
      "loss": 4.0069,
      "step": 737
    },
    {
      "epoch": 4.731942215088282,
      "grad_norm": 0.1640625,
      "learning_rate": 7.370000000000001e-06,
      "loss": 4.0057,
      "step": 738
    },
    {
      "epoch": 4.738362760834671,
      "grad_norm": 0.17578125,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 4.0052,
      "step": 739
    },
    {
      "epoch": 4.744783306581059,
      "grad_norm": 0.1796875,
      "learning_rate": 7.3899999999999995e-06,
      "loss": 4.0091,
      "step": 740
    },
    {
      "epoch": 4.751203852327448,
      "grad_norm": 1.3515625,
      "learning_rate": 7.4e-06,
      "loss": 4.0123,
      "step": 741
    },
    {
      "epoch": 4.757624398073836,
      "grad_norm": 0.2099609375,
      "learning_rate": 7.41e-06,
      "loss": 4.0096,
      "step": 742
    },
    {
      "epoch": 4.764044943820225,
      "grad_norm": 0.703125,
      "learning_rate": 7.420000000000001e-06,
      "loss": 4.0089,
      "step": 743
    },
    {
      "epoch": 4.770465489566613,
      "grad_norm": 0.341796875,
      "learning_rate": 7.430000000000001e-06,
      "loss": 4.0182,
      "step": 744
    },
    {
      "epoch": 4.776886035313002,
      "grad_norm": 0.28515625,
      "learning_rate": 7.44e-06,
      "loss": 4.0182,
      "step": 745
    },
    {
      "epoch": 4.78330658105939,
      "grad_norm": 0.271484375,
      "learning_rate": 7.45e-06,
      "loss": 4.0077,
      "step": 746
    },
    {
      "epoch": 4.789727126805778,
      "grad_norm": 0.369140625,
      "learning_rate": 7.4600000000000006e-06,
      "loss": 4.0033,
      "step": 747
    },
    {
      "epoch": 4.796147672552167,
      "grad_norm": 0.1494140625,
      "learning_rate": 7.4700000000000005e-06,
      "loss": 4.0067,
      "step": 748
    },
    {
      "epoch": 4.802568218298555,
      "grad_norm": 0.3125,
      "learning_rate": 7.480000000000001e-06,
      "loss": 4.004,
      "step": 749
    },
    {
      "epoch": 4.808988764044944,
      "grad_norm": 0.1494140625,
      "learning_rate": 7.4899999999999994e-06,
      "loss": 4.0042,
      "step": 750
    },
    {
      "epoch": 4.815409309791332,
      "grad_norm": 0.39453125,
      "learning_rate": 7.5e-06,
      "loss": 4.0029,
      "step": 751
    },
    {
      "epoch": 4.821829855537721,
      "grad_norm": 0.7421875,
      "learning_rate": 7.51e-06,
      "loss": 4.0064,
      "step": 752
    },
    {
      "epoch": 4.828250401284109,
      "grad_norm": 0.36328125,
      "learning_rate": 7.520000000000001e-06,
      "loss": 4.0064,
      "step": 753
    },
    {
      "epoch": 4.834670947030498,
      "grad_norm": 0.2197265625,
      "learning_rate": 7.530000000000001e-06,
      "loss": 4.0119,
      "step": 754
    },
    {
      "epoch": 4.841091492776886,
      "grad_norm": 1.2421875,
      "learning_rate": 7.54e-06,
      "loss": 4.0038,
      "step": 755
    },
    {
      "epoch": 4.847512038523274,
      "grad_norm": 0.2451171875,
      "learning_rate": 7.55e-06,
      "loss": 4.0073,
      "step": 756
    },
    {
      "epoch": 4.853932584269663,
      "grad_norm": 0.1845703125,
      "learning_rate": 7.5600000000000005e-06,
      "loss": 4.012,
      "step": 757
    },
    {
      "epoch": 4.860353130016051,
      "grad_norm": 0.14453125,
      "learning_rate": 7.57e-06,
      "loss": 4.0064,
      "step": 758
    },
    {
      "epoch": 4.86677367576244,
      "grad_norm": 2.765625,
      "learning_rate": 7.580000000000001e-06,
      "loss": 4.0075,
      "step": 759
    },
    {
      "epoch": 4.873194221508828,
      "grad_norm": 2.078125,
      "learning_rate": 7.59e-06,
      "loss": 4.0123,
      "step": 760
    },
    {
      "epoch": 4.879614767255217,
      "grad_norm": 1.6953125,
      "learning_rate": 7.6e-06,
      "loss": 4.0072,
      "step": 761
    },
    {
      "epoch": 4.886035313001605,
      "grad_norm": 0.29296875,
      "learning_rate": 7.610000000000001e-06,
      "loss": 4.0041,
      "step": 762
    },
    {
      "epoch": 4.892455858747994,
      "grad_norm": 0.26953125,
      "learning_rate": 7.620000000000001e-06,
      "loss": 4.0043,
      "step": 763
    },
    {
      "epoch": 4.898876404494382,
      "grad_norm": 0.150390625,
      "learning_rate": 7.630000000000001e-06,
      "loss": 4.0047,
      "step": 764
    },
    {
      "epoch": 4.90529695024077,
      "grad_norm": 0.484375,
      "learning_rate": 7.64e-06,
      "loss": 4.0202,
      "step": 765
    },
    {
      "epoch": 4.911717495987159,
      "grad_norm": 0.34375,
      "learning_rate": 7.65e-06,
      "loss": 4.004,
      "step": 766
    },
    {
      "epoch": 4.918138041733547,
      "grad_norm": 0.361328125,
      "learning_rate": 7.660000000000001e-06,
      "loss": 4.0061,
      "step": 767
    },
    {
      "epoch": 4.924558587479936,
      "grad_norm": 0.2099609375,
      "learning_rate": 7.670000000000001e-06,
      "loss": 4.0045,
      "step": 768
    },
    {
      "epoch": 4.930979133226324,
      "grad_norm": 0.2060546875,
      "learning_rate": 7.68e-06,
      "loss": 4.0053,
      "step": 769
    },
    {
      "epoch": 4.937399678972713,
      "grad_norm": 0.2138671875,
      "learning_rate": 7.69e-06,
      "loss": 4.0056,
      "step": 770
    },
    {
      "epoch": 4.943820224719101,
      "grad_norm": 0.19140625,
      "learning_rate": 7.7e-06,
      "loss": 4.0067,
      "step": 771
    },
    {
      "epoch": 4.95024077046549,
      "grad_norm": 0.58203125,
      "learning_rate": 7.71e-06,
      "loss": 4.0077,
      "step": 772
    },
    {
      "epoch": 4.956661316211878,
      "grad_norm": 0.373046875,
      "learning_rate": 7.72e-06,
      "loss": 4.0092,
      "step": 773
    },
    {
      "epoch": 4.963081861958266,
      "grad_norm": 2.25,
      "learning_rate": 7.73e-06,
      "loss": 4.0052,
      "step": 774
    },
    {
      "epoch": 4.969502407704655,
      "grad_norm": 1.7890625,
      "learning_rate": 7.74e-06,
      "loss": 4.0093,
      "step": 775
    },
    {
      "epoch": 4.975922953451043,
      "grad_norm": 1.3671875,
      "learning_rate": 7.75e-06,
      "loss": 4.0098,
      "step": 776
    },
    {
      "epoch": 4.982343499197432,
      "grad_norm": 0.51171875,
      "learning_rate": 7.76e-06,
      "loss": 4.0048,
      "step": 777
    },
    {
      "epoch": 4.98876404494382,
      "grad_norm": 0.4609375,
      "learning_rate": 7.77e-06,
      "loss": 4.0042,
      "step": 778
    },
    {
      "epoch": 4.995184590690209,
      "grad_norm": 0.64453125,
      "learning_rate": 7.78e-06,
      "loss": 4.0044,
      "step": 779
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.26953125,
      "learning_rate": 7.79e-06,
      "loss": 3.0075,
      "step": 780
    },
    {
      "epoch": 5.006420545746389,
      "grad_norm": 0.373046875,
      "learning_rate": 7.8e-06,
      "loss": 4.0144,
      "step": 781
    },
    {
      "epoch": 5.012841091492777,
      "grad_norm": 0.6796875,
      "learning_rate": 7.810000000000001e-06,
      "loss": 4.0145,
      "step": 782
    },
    {
      "epoch": 5.019261637239166,
      "grad_norm": 0.251953125,
      "learning_rate": 7.820000000000001e-06,
      "loss": 4.0097,
      "step": 783
    },
    {
      "epoch": 5.025682182985554,
      "grad_norm": 0.36328125,
      "learning_rate": 7.83e-06,
      "loss": 4.01,
      "step": 784
    },
    {
      "epoch": 5.032102728731942,
      "grad_norm": 0.240234375,
      "learning_rate": 7.84e-06,
      "loss": 4.0117,
      "step": 785
    },
    {
      "epoch": 5.03852327447833,
      "grad_norm": 0.158203125,
      "learning_rate": 7.850000000000001e-06,
      "loss": 4.0056,
      "step": 786
    },
    {
      "epoch": 5.044943820224719,
      "grad_norm": 0.427734375,
      "learning_rate": 7.860000000000001e-06,
      "loss": 4.011,
      "step": 787
    },
    {
      "epoch": 5.051364365971107,
      "grad_norm": 0.8203125,
      "learning_rate": 7.870000000000001e-06,
      "loss": 4.0185,
      "step": 788
    },
    {
      "epoch": 5.057784911717496,
      "grad_norm": 0.328125,
      "learning_rate": 7.879999999999999e-06,
      "loss": 4.0083,
      "step": 789
    },
    {
      "epoch": 5.064205457463885,
      "grad_norm": 0.2265625,
      "learning_rate": 7.89e-06,
      "loss": 4.0031,
      "step": 790
    },
    {
      "epoch": 5.070626003210273,
      "grad_norm": 0.30859375,
      "learning_rate": 7.9e-06,
      "loss": 4.0068,
      "step": 791
    },
    {
      "epoch": 5.077046548956662,
      "grad_norm": 0.259765625,
      "learning_rate": 7.91e-06,
      "loss": 4.0018,
      "step": 792
    },
    {
      "epoch": 5.08346709470305,
      "grad_norm": 0.265625,
      "learning_rate": 7.92e-06,
      "loss": 4.0259,
      "step": 793
    },
    {
      "epoch": 5.089887640449438,
      "grad_norm": 3.125,
      "learning_rate": 7.93e-06,
      "loss": 4.0058,
      "step": 794
    },
    {
      "epoch": 5.096308186195826,
      "grad_norm": 0.1640625,
      "learning_rate": 7.94e-06,
      "loss": 4.0124,
      "step": 795
    },
    {
      "epoch": 5.102728731942215,
      "grad_norm": 0.404296875,
      "learning_rate": 7.95e-06,
      "loss": 4.0116,
      "step": 796
    },
    {
      "epoch": 5.109149277688603,
      "grad_norm": 0.158203125,
      "learning_rate": 7.96e-06,
      "loss": 4.0047,
      "step": 797
    },
    {
      "epoch": 5.115569823434992,
      "grad_norm": 0.1484375,
      "learning_rate": 7.97e-06,
      "loss": 4.0022,
      "step": 798
    },
    {
      "epoch": 5.121990369181381,
      "grad_norm": 0.439453125,
      "learning_rate": 7.98e-06,
      "loss": 4.0047,
      "step": 799
    },
    {
      "epoch": 5.128410914927769,
      "grad_norm": 0.283203125,
      "learning_rate": 7.99e-06,
      "loss": 4.0057,
      "step": 800
    },
    {
      "epoch": 5.134831460674158,
      "grad_norm": 0.302734375,
      "learning_rate": 8.000000000000001e-06,
      "loss": 4.0078,
      "step": 801
    },
    {
      "epoch": 5.141252006420546,
      "grad_norm": 1.359375,
      "learning_rate": 8.010000000000001e-06,
      "loss": 4.0117,
      "step": 802
    },
    {
      "epoch": 5.147672552166934,
      "grad_norm": 0.1416015625,
      "learning_rate": 8.02e-06,
      "loss": 4.0039,
      "step": 803
    },
    {
      "epoch": 5.154093097913322,
      "grad_norm": 0.2421875,
      "learning_rate": 8.03e-06,
      "loss": 4.0073,
      "step": 804
    },
    {
      "epoch": 5.160513643659711,
      "grad_norm": 1.484375,
      "learning_rate": 8.040000000000001e-06,
      "loss": 4.0061,
      "step": 805
    },
    {
      "epoch": 5.166934189406099,
      "grad_norm": 0.302734375,
      "learning_rate": 8.050000000000001e-06,
      "loss": 4.0075,
      "step": 806
    },
    {
      "epoch": 5.173354735152488,
      "grad_norm": 0.412109375,
      "learning_rate": 8.06e-06,
      "loss": 4.0063,
      "step": 807
    },
    {
      "epoch": 5.179775280898877,
      "grad_norm": 0.322265625,
      "learning_rate": 8.069999999999999e-06,
      "loss": 4.0055,
      "step": 808
    },
    {
      "epoch": 5.186195826645265,
      "grad_norm": 0.259765625,
      "learning_rate": 8.08e-06,
      "loss": 4.0084,
      "step": 809
    },
    {
      "epoch": 5.192616372391654,
      "grad_norm": 0.80078125,
      "learning_rate": 8.09e-06,
      "loss": 4.0047,
      "step": 810
    },
    {
      "epoch": 5.199036918138042,
      "grad_norm": 0.1865234375,
      "learning_rate": 8.1e-06,
      "loss": 4.011,
      "step": 811
    },
    {
      "epoch": 5.20545746388443,
      "grad_norm": 0.462890625,
      "learning_rate": 8.11e-06,
      "loss": 4.0062,
      "step": 812
    },
    {
      "epoch": 5.211878009630818,
      "grad_norm": 0.322265625,
      "learning_rate": 8.12e-06,
      "loss": 4.0068,
      "step": 813
    },
    {
      "epoch": 5.218298555377207,
      "grad_norm": 0.2734375,
      "learning_rate": 8.13e-06,
      "loss": 4.0046,
      "step": 814
    },
    {
      "epoch": 5.224719101123595,
      "grad_norm": 0.3046875,
      "learning_rate": 8.14e-06,
      "loss": 4.0069,
      "step": 815
    },
    {
      "epoch": 5.231139646869984,
      "grad_norm": 0.498046875,
      "learning_rate": 8.15e-06,
      "loss": 4.0053,
      "step": 816
    },
    {
      "epoch": 5.237560192616373,
      "grad_norm": 0.2275390625,
      "learning_rate": 8.160000000000001e-06,
      "loss": 4.0106,
      "step": 817
    },
    {
      "epoch": 5.243980738362761,
      "grad_norm": 0.1875,
      "learning_rate": 8.17e-06,
      "loss": 4.0037,
      "step": 818
    },
    {
      "epoch": 5.25040128410915,
      "grad_norm": 0.1708984375,
      "learning_rate": 8.18e-06,
      "loss": 4.0034,
      "step": 819
    },
    {
      "epoch": 5.256821829855538,
      "grad_norm": 0.419921875,
      "learning_rate": 8.190000000000001e-06,
      "loss": 4.0071,
      "step": 820
    },
    {
      "epoch": 5.263242375601926,
      "grad_norm": 0.26171875,
      "learning_rate": 8.200000000000001e-06,
      "loss": 4.0089,
      "step": 821
    },
    {
      "epoch": 5.269662921348314,
      "grad_norm": 0.3125,
      "learning_rate": 8.210000000000001e-06,
      "loss": 4.0047,
      "step": 822
    },
    {
      "epoch": 5.276083467094703,
      "grad_norm": 0.294921875,
      "learning_rate": 8.22e-06,
      "loss": 4.0064,
      "step": 823
    },
    {
      "epoch": 5.282504012841091,
      "grad_norm": 0.28515625,
      "learning_rate": 8.23e-06,
      "loss": 4.0017,
      "step": 824
    },
    {
      "epoch": 5.28892455858748,
      "grad_norm": 0.1552734375,
      "learning_rate": 8.24e-06,
      "loss": 4.008,
      "step": 825
    },
    {
      "epoch": 5.295345104333869,
      "grad_norm": 0.57421875,
      "learning_rate": 8.25e-06,
      "loss": 4.0095,
      "step": 826
    },
    {
      "epoch": 5.301765650080257,
      "grad_norm": 0.578125,
      "learning_rate": 8.26e-06,
      "loss": 4.008,
      "step": 827
    },
    {
      "epoch": 5.308186195826646,
      "grad_norm": 2.375,
      "learning_rate": 8.27e-06,
      "loss": 4.0046,
      "step": 828
    },
    {
      "epoch": 5.314606741573034,
      "grad_norm": 0.30859375,
      "learning_rate": 8.28e-06,
      "loss": 4.0053,
      "step": 829
    },
    {
      "epoch": 5.321027287319422,
      "grad_norm": 0.322265625,
      "learning_rate": 8.29e-06,
      "loss": 4.007,
      "step": 830
    },
    {
      "epoch": 5.32744783306581,
      "grad_norm": 1.390625,
      "learning_rate": 8.3e-06,
      "loss": 4.0123,
      "step": 831
    },
    {
      "epoch": 5.333868378812199,
      "grad_norm": 0.279296875,
      "learning_rate": 8.31e-06,
      "loss": 4.0058,
      "step": 832
    },
    {
      "epoch": 5.340288924558587,
      "grad_norm": 0.21484375,
      "learning_rate": 8.32e-06,
      "loss": 4.0052,
      "step": 833
    },
    {
      "epoch": 5.346709470304976,
      "grad_norm": 0.201171875,
      "learning_rate": 8.33e-06,
      "loss": 4.0065,
      "step": 834
    },
    {
      "epoch": 5.353130016051364,
      "grad_norm": 0.19140625,
      "learning_rate": 8.34e-06,
      "loss": 4.0048,
      "step": 835
    },
    {
      "epoch": 5.359550561797753,
      "grad_norm": 0.396484375,
      "learning_rate": 8.350000000000001e-06,
      "loss": 4.0064,
      "step": 836
    },
    {
      "epoch": 5.365971107544142,
      "grad_norm": 0.255859375,
      "learning_rate": 8.36e-06,
      "loss": 4.0076,
      "step": 837
    },
    {
      "epoch": 5.37239165329053,
      "grad_norm": 0.59375,
      "learning_rate": 8.37e-06,
      "loss": 4.006,
      "step": 838
    },
    {
      "epoch": 5.378812199036918,
      "grad_norm": 0.1796875,
      "learning_rate": 8.380000000000001e-06,
      "loss": 4.0061,
      "step": 839
    },
    {
      "epoch": 5.385232744783306,
      "grad_norm": 0.1591796875,
      "learning_rate": 8.390000000000001e-06,
      "loss": 4.0074,
      "step": 840
    },
    {
      "epoch": 5.391653290529695,
      "grad_norm": 0.2275390625,
      "learning_rate": 8.400000000000001e-06,
      "loss": 4.0035,
      "step": 841
    },
    {
      "epoch": 5.398073836276083,
      "grad_norm": 0.337890625,
      "learning_rate": 8.409999999999999e-06,
      "loss": 4.0068,
      "step": 842
    },
    {
      "epoch": 5.404494382022472,
      "grad_norm": 0.1650390625,
      "learning_rate": 8.42e-06,
      "loss": 4.0119,
      "step": 843
    },
    {
      "epoch": 5.41091492776886,
      "grad_norm": 0.453125,
      "learning_rate": 8.43e-06,
      "loss": 4.0034,
      "step": 844
    },
    {
      "epoch": 5.417335473515249,
      "grad_norm": 0.205078125,
      "learning_rate": 8.44e-06,
      "loss": 4.0094,
      "step": 845
    },
    {
      "epoch": 5.423756019261638,
      "grad_norm": 0.2734375,
      "learning_rate": 8.45e-06,
      "loss": 4.0081,
      "step": 846
    },
    {
      "epoch": 5.430176565008026,
      "grad_norm": 0.197265625,
      "learning_rate": 8.46e-06,
      "loss": 4.0119,
      "step": 847
    },
    {
      "epoch": 5.436597110754414,
      "grad_norm": 0.412109375,
      "learning_rate": 8.47e-06,
      "loss": 4.0103,
      "step": 848
    },
    {
      "epoch": 5.443017656500802,
      "grad_norm": 0.408203125,
      "learning_rate": 8.48e-06,
      "loss": 4.0059,
      "step": 849
    },
    {
      "epoch": 5.449438202247191,
      "grad_norm": 0.326171875,
      "learning_rate": 8.49e-06,
      "loss": 4.0037,
      "step": 850
    },
    {
      "epoch": 5.455858747993579,
      "grad_norm": 0.228515625,
      "learning_rate": 8.500000000000002e-06,
      "loss": 4.0048,
      "step": 851
    },
    {
      "epoch": 5.462279293739968,
      "grad_norm": 0.4609375,
      "learning_rate": 8.51e-06,
      "loss": 4.0117,
      "step": 852
    },
    {
      "epoch": 5.468699839486356,
      "grad_norm": 3.109375,
      "learning_rate": 8.52e-06,
      "loss": 4.0572,
      "step": 853
    },
    {
      "epoch": 5.475120385232745,
      "grad_norm": 0.419921875,
      "learning_rate": 8.53e-06,
      "loss": 4.0025,
      "step": 854
    },
    {
      "epoch": 5.481540930979134,
      "grad_norm": 0.494140625,
      "learning_rate": 8.540000000000001e-06,
      "loss": 4.0053,
      "step": 855
    },
    {
      "epoch": 5.487961476725522,
      "grad_norm": 0.31640625,
      "learning_rate": 8.550000000000001e-06,
      "loss": 4.0052,
      "step": 856
    },
    {
      "epoch": 5.49438202247191,
      "grad_norm": 0.287109375,
      "learning_rate": 8.56e-06,
      "loss": 4.0038,
      "step": 857
    },
    {
      "epoch": 5.500802568218298,
      "grad_norm": 0.7421875,
      "learning_rate": 8.570000000000001e-06,
      "loss": 4.0041,
      "step": 858
    },
    {
      "epoch": 5.507223113964687,
      "grad_norm": 0.2578125,
      "learning_rate": 8.580000000000001e-06,
      "loss": 4.0148,
      "step": 859
    },
    {
      "epoch": 5.513643659711075,
      "grad_norm": 0.4765625,
      "learning_rate": 8.59e-06,
      "loss": 4.0053,
      "step": 860
    },
    {
      "epoch": 5.520064205457464,
      "grad_norm": 0.27734375,
      "learning_rate": 8.599999999999999e-06,
      "loss": 4.0067,
      "step": 861
    },
    {
      "epoch": 5.526484751203852,
      "grad_norm": 0.30859375,
      "learning_rate": 8.61e-06,
      "loss": 4.0118,
      "step": 862
    },
    {
      "epoch": 5.532905296950241,
      "grad_norm": 0.353515625,
      "learning_rate": 8.62e-06,
      "loss": 4.0005,
      "step": 863
    },
    {
      "epoch": 5.539325842696629,
      "grad_norm": 0.765625,
      "learning_rate": 8.63e-06,
      "loss": 4.0073,
      "step": 864
    },
    {
      "epoch": 5.545746388443018,
      "grad_norm": 0.39453125,
      "learning_rate": 8.64e-06,
      "loss": 4.0087,
      "step": 865
    },
    {
      "epoch": 5.552166934189406,
      "grad_norm": 0.59765625,
      "learning_rate": 8.65e-06,
      "loss": 4.012,
      "step": 866
    },
    {
      "epoch": 5.558587479935794,
      "grad_norm": 0.83203125,
      "learning_rate": 8.66e-06,
      "loss": 4.0079,
      "step": 867
    },
    {
      "epoch": 5.565008025682183,
      "grad_norm": 0.30859375,
      "learning_rate": 8.67e-06,
      "loss": 4.0042,
      "step": 868
    },
    {
      "epoch": 5.571428571428571,
      "grad_norm": 0.56640625,
      "learning_rate": 8.68e-06,
      "loss": 4.0066,
      "step": 869
    },
    {
      "epoch": 5.57784911717496,
      "grad_norm": 0.412109375,
      "learning_rate": 8.690000000000002e-06,
      "loss": 4.0066,
      "step": 870
    },
    {
      "epoch": 5.584269662921348,
      "grad_norm": 0.1826171875,
      "learning_rate": 8.7e-06,
      "loss": 4.0006,
      "step": 871
    },
    {
      "epoch": 5.590690208667737,
      "grad_norm": 0.1494140625,
      "learning_rate": 8.71e-06,
      "loss": 4.0043,
      "step": 872
    },
    {
      "epoch": 5.597110754414125,
      "grad_norm": 0.4375,
      "learning_rate": 8.720000000000001e-06,
      "loss": 4.0078,
      "step": 873
    },
    {
      "epoch": 5.603531300160514,
      "grad_norm": 0.28515625,
      "learning_rate": 8.730000000000001e-06,
      "loss": 4.0011,
      "step": 874
    },
    {
      "epoch": 5.609951845906902,
      "grad_norm": 0.34375,
      "learning_rate": 8.740000000000001e-06,
      "loss": 4.0096,
      "step": 875
    },
    {
      "epoch": 5.61637239165329,
      "grad_norm": 0.1669921875,
      "learning_rate": 8.75e-06,
      "loss": 4.0047,
      "step": 876
    },
    {
      "epoch": 5.622792937399679,
      "grad_norm": 0.177734375,
      "learning_rate": 8.76e-06,
      "loss": 4.0043,
      "step": 877
    },
    {
      "epoch": 5.629213483146067,
      "grad_norm": 0.82421875,
      "learning_rate": 8.77e-06,
      "loss": 4.001,
      "step": 878
    },
    {
      "epoch": 5.635634028892456,
      "grad_norm": 0.26171875,
      "learning_rate": 8.78e-06,
      "loss": 4.0095,
      "step": 879
    },
    {
      "epoch": 5.642054574638844,
      "grad_norm": 0.216796875,
      "learning_rate": 8.79e-06,
      "loss": 4.0066,
      "step": 880
    },
    {
      "epoch": 5.648475120385233,
      "grad_norm": 0.1796875,
      "learning_rate": 8.8e-06,
      "loss": 4.0074,
      "step": 881
    },
    {
      "epoch": 5.654895666131621,
      "grad_norm": 2.578125,
      "learning_rate": 8.81e-06,
      "loss": 4.0087,
      "step": 882
    },
    {
      "epoch": 5.66131621187801,
      "grad_norm": 0.52734375,
      "learning_rate": 8.82e-06,
      "loss": 4.0082,
      "step": 883
    },
    {
      "epoch": 5.667736757624398,
      "grad_norm": 0.361328125,
      "learning_rate": 8.83e-06,
      "loss": 4.0022,
      "step": 884
    },
    {
      "epoch": 5.674157303370786,
      "grad_norm": 0.216796875,
      "learning_rate": 8.840000000000002e-06,
      "loss": 4.0067,
      "step": 885
    },
    {
      "epoch": 5.680577849117175,
      "grad_norm": 0.51171875,
      "learning_rate": 8.85e-06,
      "loss": 4.005,
      "step": 886
    },
    {
      "epoch": 5.686998394863563,
      "grad_norm": 0.29296875,
      "learning_rate": 8.86e-06,
      "loss": 4.0079,
      "step": 887
    },
    {
      "epoch": 5.693418940609952,
      "grad_norm": 0.134765625,
      "learning_rate": 8.87e-06,
      "loss": 3.9997,
      "step": 888
    },
    {
      "epoch": 5.69983948635634,
      "grad_norm": 0.2119140625,
      "learning_rate": 8.880000000000001e-06,
      "loss": 4.0038,
      "step": 889
    },
    {
      "epoch": 5.706260032102729,
      "grad_norm": 0.2021484375,
      "learning_rate": 8.890000000000001e-06,
      "loss": 4.0031,
      "step": 890
    },
    {
      "epoch": 5.712680577849117,
      "grad_norm": 11.25,
      "learning_rate": 8.9e-06,
      "loss": 4.0239,
      "step": 891
    },
    {
      "epoch": 5.719101123595506,
      "grad_norm": 0.294921875,
      "learning_rate": 8.910000000000001e-06,
      "loss": 4.0052,
      "step": 892
    },
    {
      "epoch": 5.725521669341894,
      "grad_norm": 0.158203125,
      "learning_rate": 8.920000000000001e-06,
      "loss": 4.005,
      "step": 893
    },
    {
      "epoch": 5.731942215088282,
      "grad_norm": 0.404296875,
      "learning_rate": 8.930000000000001e-06,
      "loss": 4.0152,
      "step": 894
    },
    {
      "epoch": 5.738362760834671,
      "grad_norm": 0.2236328125,
      "learning_rate": 8.939999999999999e-06,
      "loss": 4.0048,
      "step": 895
    },
    {
      "epoch": 5.744783306581059,
      "grad_norm": 0.7109375,
      "learning_rate": 8.95e-06,
      "loss": 4.0134,
      "step": 896
    },
    {
      "epoch": 5.751203852327448,
      "grad_norm": 0.275390625,
      "learning_rate": 8.96e-06,
      "loss": 4.0027,
      "step": 897
    },
    {
      "epoch": 5.757624398073836,
      "grad_norm": 0.478515625,
      "learning_rate": 8.97e-06,
      "loss": 4.0088,
      "step": 898
    },
    {
      "epoch": 5.764044943820225,
      "grad_norm": 0.1591796875,
      "learning_rate": 8.98e-06,
      "loss": 4.0033,
      "step": 899
    },
    {
      "epoch": 5.770465489566613,
      "grad_norm": 0.5546875,
      "learning_rate": 8.99e-06,
      "loss": 4.0049,
      "step": 900
    },
    {
      "epoch": 5.776886035313002,
      "grad_norm": 1.9296875,
      "learning_rate": 9e-06,
      "loss": 4.0148,
      "step": 901
    },
    {
      "epoch": 5.78330658105939,
      "grad_norm": 0.294921875,
      "learning_rate": 9.01e-06,
      "loss": 4.0042,
      "step": 902
    },
    {
      "epoch": 5.789727126805778,
      "grad_norm": 0.53125,
      "learning_rate": 9.02e-06,
      "loss": 4.0049,
      "step": 903
    },
    {
      "epoch": 5.796147672552167,
      "grad_norm": 0.5546875,
      "learning_rate": 9.030000000000002e-06,
      "loss": 4.0022,
      "step": 904
    },
    {
      "epoch": 5.802568218298555,
      "grad_norm": 0.439453125,
      "learning_rate": 9.04e-06,
      "loss": 4.0056,
      "step": 905
    },
    {
      "epoch": 5.808988764044944,
      "grad_norm": 0.470703125,
      "learning_rate": 9.05e-06,
      "loss": 4.0101,
      "step": 906
    },
    {
      "epoch": 5.815409309791332,
      "grad_norm": 0.47265625,
      "learning_rate": 9.06e-06,
      "loss": 4.0048,
      "step": 907
    },
    {
      "epoch": 5.821829855537721,
      "grad_norm": 0.150390625,
      "learning_rate": 9.070000000000001e-06,
      "loss": 4.0035,
      "step": 908
    },
    {
      "epoch": 5.828250401284109,
      "grad_norm": 0.197265625,
      "learning_rate": 9.080000000000001e-06,
      "loss": 4.0088,
      "step": 909
    },
    {
      "epoch": 5.834670947030498,
      "grad_norm": 0.6796875,
      "learning_rate": 9.09e-06,
      "loss": 4.0042,
      "step": 910
    },
    {
      "epoch": 5.841091492776886,
      "grad_norm": 0.72265625,
      "learning_rate": 9.100000000000001e-06,
      "loss": 4.0052,
      "step": 911
    },
    {
      "epoch": 5.847512038523274,
      "grad_norm": 0.69921875,
      "learning_rate": 9.110000000000001e-06,
      "loss": 4.0083,
      "step": 912
    },
    {
      "epoch": 5.853932584269663,
      "grad_norm": 0.76953125,
      "learning_rate": 9.12e-06,
      "loss": 4.0127,
      "step": 913
    },
    {
      "epoch": 5.860353130016051,
      "grad_norm": 0.2421875,
      "learning_rate": 9.13e-06,
      "loss": 4.0072,
      "step": 914
    },
    {
      "epoch": 5.86677367576244,
      "grad_norm": 0.50390625,
      "learning_rate": 9.14e-06,
      "loss": 4.0071,
      "step": 915
    },
    {
      "epoch": 5.873194221508828,
      "grad_norm": 0.4140625,
      "learning_rate": 9.15e-06,
      "loss": 4.0038,
      "step": 916
    },
    {
      "epoch": 5.879614767255217,
      "grad_norm": 0.322265625,
      "learning_rate": 9.16e-06,
      "loss": 4.0065,
      "step": 917
    },
    {
      "epoch": 5.886035313001605,
      "grad_norm": 0.25390625,
      "learning_rate": 9.17e-06,
      "loss": 4.0058,
      "step": 918
    },
    {
      "epoch": 5.892455858747994,
      "grad_norm": 0.208984375,
      "learning_rate": 9.180000000000002e-06,
      "loss": 4.0046,
      "step": 919
    },
    {
      "epoch": 5.898876404494382,
      "grad_norm": 0.5859375,
      "learning_rate": 9.19e-06,
      "loss": 4.0076,
      "step": 920
    },
    {
      "epoch": 5.90529695024077,
      "grad_norm": 0.451171875,
      "learning_rate": 9.2e-06,
      "loss": 4.0033,
      "step": 921
    },
    {
      "epoch": 5.911717495987159,
      "grad_norm": 0.79296875,
      "learning_rate": 9.21e-06,
      "loss": 4.0051,
      "step": 922
    },
    {
      "epoch": 5.918138041733547,
      "grad_norm": 0.328125,
      "learning_rate": 9.220000000000002e-06,
      "loss": 4.0085,
      "step": 923
    },
    {
      "epoch": 5.924558587479936,
      "grad_norm": 0.2314453125,
      "learning_rate": 9.23e-06,
      "loss": 4.0071,
      "step": 924
    },
    {
      "epoch": 5.930979133226324,
      "grad_norm": 0.86328125,
      "learning_rate": 9.24e-06,
      "loss": 4.0064,
      "step": 925
    },
    {
      "epoch": 5.937399678972713,
      "grad_norm": 0.50390625,
      "learning_rate": 9.25e-06,
      "loss": 4.0035,
      "step": 926
    },
    {
      "epoch": 5.943820224719101,
      "grad_norm": 0.40234375,
      "learning_rate": 9.260000000000001e-06,
      "loss": 4.003,
      "step": 927
    },
    {
      "epoch": 5.95024077046549,
      "grad_norm": 0.2236328125,
      "learning_rate": 9.270000000000001e-06,
      "loss": 4.0084,
      "step": 928
    },
    {
      "epoch": 5.956661316211878,
      "grad_norm": 0.32421875,
      "learning_rate": 9.28e-06,
      "loss": 4.0015,
      "step": 929
    },
    {
      "epoch": 5.963081861958266,
      "grad_norm": 0.1748046875,
      "learning_rate": 9.29e-06,
      "loss": 4.0018,
      "step": 930
    },
    {
      "epoch": 5.969502407704655,
      "grad_norm": 0.296875,
      "learning_rate": 9.3e-06,
      "loss": 4.0087,
      "step": 931
    },
    {
      "epoch": 5.975922953451043,
      "grad_norm": 0.25390625,
      "learning_rate": 9.31e-06,
      "loss": 4.0063,
      "step": 932
    },
    {
      "epoch": 5.982343499197432,
      "grad_norm": 0.443359375,
      "learning_rate": 9.32e-06,
      "loss": 4.0037,
      "step": 933
    },
    {
      "epoch": 5.98876404494382,
      "grad_norm": 0.2119140625,
      "learning_rate": 9.33e-06,
      "loss": 4.0041,
      "step": 934
    },
    {
      "epoch": 5.995184590690209,
      "grad_norm": 0.359375,
      "learning_rate": 9.34e-06,
      "loss": 4.0038,
      "step": 935
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.53515625,
      "learning_rate": 9.35e-06,
      "loss": 3.0034,
      "step": 936
    },
    {
      "epoch": 6.006420545746389,
      "grad_norm": 0.279296875,
      "learning_rate": 9.36e-06,
      "loss": 4.0075,
      "step": 937
    },
    {
      "epoch": 6.012841091492777,
      "grad_norm": 0.1767578125,
      "learning_rate": 9.370000000000002e-06,
      "loss": 4.0082,
      "step": 938
    },
    {
      "epoch": 6.019261637239166,
      "grad_norm": 0.25390625,
      "learning_rate": 9.38e-06,
      "loss": 4.0046,
      "step": 939
    },
    {
      "epoch": 6.025682182985554,
      "grad_norm": 0.21875,
      "learning_rate": 9.39e-06,
      "loss": 4.0018,
      "step": 940
    },
    {
      "epoch": 6.032102728731942,
      "grad_norm": 0.6640625,
      "learning_rate": 9.4e-06,
      "loss": 4.0071,
      "step": 941
    },
    {
      "epoch": 6.03852327447833,
      "grad_norm": 0.1875,
      "learning_rate": 9.410000000000001e-06,
      "loss": 4.0029,
      "step": 942
    },
    {
      "epoch": 6.044943820224719,
      "grad_norm": 0.384765625,
      "learning_rate": 9.420000000000001e-06,
      "loss": 4.0057,
      "step": 943
    },
    {
      "epoch": 6.051364365971107,
      "grad_norm": 0.3984375,
      "learning_rate": 9.43e-06,
      "loss": 4.0049,
      "step": 944
    },
    {
      "epoch": 6.057784911717496,
      "grad_norm": 0.1376953125,
      "learning_rate": 9.44e-06,
      "loss": 4.0019,
      "step": 945
    },
    {
      "epoch": 6.064205457463885,
      "grad_norm": 0.173828125,
      "learning_rate": 9.450000000000001e-06,
      "loss": 4.0051,
      "step": 946
    },
    {
      "epoch": 6.070626003210273,
      "grad_norm": 0.41015625,
      "learning_rate": 9.460000000000001e-06,
      "loss": 4.0007,
      "step": 947
    },
    {
      "epoch": 6.077046548956662,
      "grad_norm": 0.228515625,
      "learning_rate": 9.47e-06,
      "loss": 4.0081,
      "step": 948
    },
    {
      "epoch": 6.08346709470305,
      "grad_norm": 0.1611328125,
      "learning_rate": 9.48e-06,
      "loss": 4.0011,
      "step": 949
    },
    {
      "epoch": 6.089887640449438,
      "grad_norm": 0.330078125,
      "learning_rate": 9.49e-06,
      "loss": 4.0047,
      "step": 950
    },
    {
      "epoch": 6.096308186195826,
      "grad_norm": 0.17578125,
      "learning_rate": 9.5e-06,
      "loss": 4.0064,
      "step": 951
    },
    {
      "epoch": 6.102728731942215,
      "grad_norm": 0.1357421875,
      "learning_rate": 9.51e-06,
      "loss": 4.0049,
      "step": 952
    },
    {
      "epoch": 6.109149277688603,
      "grad_norm": 0.1826171875,
      "learning_rate": 9.52e-06,
      "loss": 4.0031,
      "step": 953
    },
    {
      "epoch": 6.115569823434992,
      "grad_norm": 0.1953125,
      "learning_rate": 9.53e-06,
      "loss": 4.0108,
      "step": 954
    },
    {
      "epoch": 6.121990369181381,
      "grad_norm": 0.1982421875,
      "learning_rate": 9.54e-06,
      "loss": 4.0114,
      "step": 955
    },
    {
      "epoch": 6.128410914927769,
      "grad_norm": 0.69921875,
      "learning_rate": 9.55e-06,
      "loss": 4.0158,
      "step": 956
    },
    {
      "epoch": 6.134831460674158,
      "grad_norm": 0.2138671875,
      "learning_rate": 9.560000000000002e-06,
      "loss": 4.0071,
      "step": 957
    },
    {
      "epoch": 6.141252006420546,
      "grad_norm": 0.275390625,
      "learning_rate": 9.57e-06,
      "loss": 4.0038,
      "step": 958
    },
    {
      "epoch": 6.147672552166934,
      "grad_norm": 0.244140625,
      "learning_rate": 9.58e-06,
      "loss": 4.0032,
      "step": 959
    },
    {
      "epoch": 6.154093097913322,
      "grad_norm": 0.2275390625,
      "learning_rate": 9.59e-06,
      "loss": 4.0021,
      "step": 960
    },
    {
      "epoch": 6.160513643659711,
      "grad_norm": 0.173828125,
      "learning_rate": 9.600000000000001e-06,
      "loss": 4.0021,
      "step": 961
    },
    {
      "epoch": 6.166934189406099,
      "grad_norm": 0.162109375,
      "learning_rate": 9.610000000000001e-06,
      "loss": 4.0049,
      "step": 962
    },
    {
      "epoch": 6.173354735152488,
      "grad_norm": 0.1376953125,
      "learning_rate": 9.62e-06,
      "loss": 4.0042,
      "step": 963
    },
    {
      "epoch": 6.179775280898877,
      "grad_norm": 0.30859375,
      "learning_rate": 9.630000000000001e-06,
      "loss": 4.0041,
      "step": 964
    },
    {
      "epoch": 6.186195826645265,
      "grad_norm": 0.162109375,
      "learning_rate": 9.640000000000001e-06,
      "loss": 4.0027,
      "step": 965
    },
    {
      "epoch": 6.192616372391654,
      "grad_norm": 0.2158203125,
      "learning_rate": 9.65e-06,
      "loss": 4.0028,
      "step": 966
    },
    {
      "epoch": 6.199036918138042,
      "grad_norm": 0.1513671875,
      "learning_rate": 9.66e-06,
      "loss": 3.9995,
      "step": 967
    },
    {
      "epoch": 6.20545746388443,
      "grad_norm": 0.205078125,
      "learning_rate": 9.67e-06,
      "loss": 4.008,
      "step": 968
    },
    {
      "epoch": 6.211878009630818,
      "grad_norm": 0.404296875,
      "learning_rate": 9.68e-06,
      "loss": 4.0144,
      "step": 969
    },
    {
      "epoch": 6.218298555377207,
      "grad_norm": 0.1513671875,
      "learning_rate": 9.69e-06,
      "loss": 4.0024,
      "step": 970
    },
    {
      "epoch": 6.224719101123595,
      "grad_norm": 0.1708984375,
      "learning_rate": 9.7e-06,
      "loss": 4.001,
      "step": 971
    },
    {
      "epoch": 6.231139646869984,
      "grad_norm": 0.50390625,
      "learning_rate": 9.71e-06,
      "loss": 4.0059,
      "step": 972
    },
    {
      "epoch": 6.237560192616373,
      "grad_norm": 0.37109375,
      "learning_rate": 9.72e-06,
      "loss": 4.0023,
      "step": 973
    },
    {
      "epoch": 6.243980738362761,
      "grad_norm": 0.6328125,
      "learning_rate": 9.73e-06,
      "loss": 4.0071,
      "step": 974
    },
    {
      "epoch": 6.25040128410915,
      "grad_norm": 0.2236328125,
      "learning_rate": 9.74e-06,
      "loss": 4.0035,
      "step": 975
    },
    {
      "epoch": 6.256821829855538,
      "grad_norm": 0.177734375,
      "learning_rate": 9.750000000000002e-06,
      "loss": 4.0069,
      "step": 976
    },
    {
      "epoch": 6.263242375601926,
      "grad_norm": 0.283203125,
      "learning_rate": 9.760000000000001e-06,
      "loss": 4.0066,
      "step": 977
    },
    {
      "epoch": 6.269662921348314,
      "grad_norm": 0.3125,
      "learning_rate": 9.77e-06,
      "loss": 4.0068,
      "step": 978
    },
    {
      "epoch": 6.276083467094703,
      "grad_norm": 0.58984375,
      "learning_rate": 9.78e-06,
      "loss": 4.0049,
      "step": 979
    },
    {
      "epoch": 6.282504012841091,
      "grad_norm": 2.28125,
      "learning_rate": 9.790000000000001e-06,
      "loss": 4.009,
      "step": 980
    },
    {
      "epoch": 6.28892455858748,
      "grad_norm": 0.296875,
      "learning_rate": 9.800000000000001e-06,
      "loss": 4.0085,
      "step": 981
    },
    {
      "epoch": 6.295345104333869,
      "grad_norm": 1.4765625,
      "learning_rate": 9.810000000000001e-06,
      "loss": 4.0103,
      "step": 982
    },
    {
      "epoch": 6.301765650080257,
      "grad_norm": 0.90625,
      "learning_rate": 9.820000000000001e-06,
      "loss": 4.0114,
      "step": 983
    },
    {
      "epoch": 6.308186195826646,
      "grad_norm": 0.27734375,
      "learning_rate": 9.83e-06,
      "loss": 4.0066,
      "step": 984
    },
    {
      "epoch": 6.314606741573034,
      "grad_norm": 0.2275390625,
      "learning_rate": 9.84e-06,
      "loss": 4.0009,
      "step": 985
    },
    {
      "epoch": 6.321027287319422,
      "grad_norm": 0.302734375,
      "learning_rate": 9.85e-06,
      "loss": 4.0031,
      "step": 986
    },
    {
      "epoch": 6.32744783306581,
      "grad_norm": 0.22265625,
      "learning_rate": 9.86e-06,
      "loss": 4.0034,
      "step": 987
    },
    {
      "epoch": 6.333868378812199,
      "grad_norm": 0.3046875,
      "learning_rate": 9.87e-06,
      "loss": 4.0052,
      "step": 988
    },
    {
      "epoch": 6.340288924558587,
      "grad_norm": 0.1484375,
      "learning_rate": 9.88e-06,
      "loss": 4.0034,
      "step": 989
    },
    {
      "epoch": 6.346709470304976,
      "grad_norm": 0.275390625,
      "learning_rate": 9.89e-06,
      "loss": 4.0035,
      "step": 990
    },
    {
      "epoch": 6.353130016051364,
      "grad_norm": 2.140625,
      "learning_rate": 9.900000000000002e-06,
      "loss": 4.015,
      "step": 991
    },
    {
      "epoch": 6.359550561797753,
      "grad_norm": 0.2265625,
      "learning_rate": 9.91e-06,
      "loss": 4.0025,
      "step": 992
    },
    {
      "epoch": 6.365971107544142,
      "grad_norm": 0.197265625,
      "learning_rate": 9.92e-06,
      "loss": 4.0077,
      "step": 993
    },
    {
      "epoch": 6.37239165329053,
      "grad_norm": 1.28125,
      "learning_rate": 9.93e-06,
      "loss": 4.0028,
      "step": 994
    },
    {
      "epoch": 6.378812199036918,
      "grad_norm": 0.388671875,
      "learning_rate": 9.940000000000001e-06,
      "loss": 4.0098,
      "step": 995
    },
    {
      "epoch": 6.385232744783306,
      "grad_norm": 0.1474609375,
      "learning_rate": 9.950000000000001e-06,
      "loss": 4.0071,
      "step": 996
    },
    {
      "epoch": 6.391653290529695,
      "grad_norm": 0.1982421875,
      "learning_rate": 9.96e-06,
      "loss": 4.0036,
      "step": 997
    },
    {
      "epoch": 6.398073836276083,
      "grad_norm": 0.78515625,
      "learning_rate": 9.97e-06,
      "loss": 4.0057,
      "step": 998
    },
    {
      "epoch": 6.404494382022472,
      "grad_norm": 0.2353515625,
      "learning_rate": 9.980000000000001e-06,
      "loss": 4.0075,
      "step": 999
    },
    {
      "epoch": 6.41091492776886,
      "grad_norm": 0.13671875,
      "learning_rate": 9.990000000000001e-06,
      "loss": 4.0046,
      "step": 1000
    },
    {
      "epoch": 6.417335473515249,
      "grad_norm": 0.1953125,
      "learning_rate": 1e-05,
      "loss": 4.0031,
      "step": 1001
    },
    {
      "epoch": 6.423756019261638,
      "grad_norm": 0.294921875,
      "learning_rate": 1.001e-05,
      "loss": 4.0038,
      "step": 1002
    },
    {
      "epoch": 6.430176565008026,
      "grad_norm": 0.130859375,
      "learning_rate": 1.002e-05,
      "loss": 4.0051,
      "step": 1003
    },
    {
      "epoch": 6.436597110754414,
      "grad_norm": 0.134765625,
      "learning_rate": 1.003e-05,
      "loss": 4.0048,
      "step": 1004
    },
    {
      "epoch": 6.443017656500802,
      "grad_norm": 0.294921875,
      "learning_rate": 1.004e-05,
      "loss": 4.0018,
      "step": 1005
    },
    {
      "epoch": 6.449438202247191,
      "grad_norm": 0.1494140625,
      "learning_rate": 1.005e-05,
      "loss": 3.9995,
      "step": 1006
    },
    {
      "epoch": 6.455858747993579,
      "grad_norm": 0.216796875,
      "learning_rate": 1.006e-05,
      "loss": 4.0066,
      "step": 1007
    },
    {
      "epoch": 6.462279293739968,
      "grad_norm": 0.16015625,
      "learning_rate": 1.007e-05,
      "loss": 4.0035,
      "step": 1008
    },
    {
      "epoch": 6.468699839486356,
      "grad_norm": 0.359375,
      "learning_rate": 1.008e-05,
      "loss": 3.9993,
      "step": 1009
    },
    {
      "epoch": 6.475120385232745,
      "grad_norm": 0.140625,
      "learning_rate": 1.0090000000000002e-05,
      "loss": 4.0031,
      "step": 1010
    },
    {
      "epoch": 6.481540930979134,
      "grad_norm": 0.2001953125,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 4.0059,
      "step": 1011
    },
    {
      "epoch": 6.487961476725522,
      "grad_norm": 0.16796875,
      "learning_rate": 1.011e-05,
      "loss": 4.0033,
      "step": 1012
    },
    {
      "epoch": 6.49438202247191,
      "grad_norm": 0.2109375,
      "learning_rate": 1.012e-05,
      "loss": 4.0057,
      "step": 1013
    },
    {
      "epoch": 6.500802568218298,
      "grad_norm": 0.1708984375,
      "learning_rate": 1.0130000000000001e-05,
      "loss": 4.0026,
      "step": 1014
    },
    {
      "epoch": 6.507223113964687,
      "grad_norm": 0.140625,
      "learning_rate": 1.0140000000000001e-05,
      "loss": 4.0052,
      "step": 1015
    },
    {
      "epoch": 6.513643659711075,
      "grad_norm": 0.640625,
      "learning_rate": 1.0150000000000001e-05,
      "loss": 4.0118,
      "step": 1016
    },
    {
      "epoch": 6.520064205457464,
      "grad_norm": 0.26953125,
      "learning_rate": 1.016e-05,
      "loss": 4.0095,
      "step": 1017
    },
    {
      "epoch": 6.526484751203852,
      "grad_norm": 0.208984375,
      "learning_rate": 1.0170000000000001e-05,
      "loss": 4.0088,
      "step": 1018
    },
    {
      "epoch": 6.532905296950241,
      "grad_norm": 0.140625,
      "learning_rate": 1.018e-05,
      "loss": 4.0073,
      "step": 1019
    },
    {
      "epoch": 6.539325842696629,
      "grad_norm": 0.1640625,
      "learning_rate": 1.019e-05,
      "loss": 4.003,
      "step": 1020
    },
    {
      "epoch": 6.545746388443018,
      "grad_norm": 0.291015625,
      "learning_rate": 1.02e-05,
      "loss": 4.0084,
      "step": 1021
    },
    {
      "epoch": 6.552166934189406,
      "grad_norm": 0.2041015625,
      "learning_rate": 1.021e-05,
      "loss": 4.0024,
      "step": 1022
    },
    {
      "epoch": 6.558587479935794,
      "grad_norm": 0.263671875,
      "learning_rate": 1.022e-05,
      "loss": 4.0065,
      "step": 1023
    },
    {
      "epoch": 6.565008025682183,
      "grad_norm": 0.2578125,
      "learning_rate": 1.023e-05,
      "loss": 4.0042,
      "step": 1024
    },
    {
      "epoch": 6.571428571428571,
      "grad_norm": 0.2109375,
      "learning_rate": 1.024e-05,
      "loss": 4.0073,
      "step": 1025
    },
    {
      "epoch": 6.57784911717496,
      "grad_norm": 0.1630859375,
      "learning_rate": 1.025e-05,
      "loss": 4.0117,
      "step": 1026
    },
    {
      "epoch": 6.584269662921348,
      "grad_norm": 0.1640625,
      "learning_rate": 1.026e-05,
      "loss": 4.0121,
      "step": 1027
    },
    {
      "epoch": 6.590690208667737,
      "grad_norm": 1.21875,
      "learning_rate": 1.027e-05,
      "loss": 4.0072,
      "step": 1028
    },
    {
      "epoch": 6.597110754414125,
      "grad_norm": 0.419921875,
      "learning_rate": 1.0280000000000002e-05,
      "loss": 4.0035,
      "step": 1029
    },
    {
      "epoch": 6.603531300160514,
      "grad_norm": 0.1689453125,
      "learning_rate": 1.0290000000000001e-05,
      "loss": 4.0133,
      "step": 1030
    },
    {
      "epoch": 6.609951845906902,
      "grad_norm": 0.14453125,
      "learning_rate": 1.03e-05,
      "loss": 4.0028,
      "step": 1031
    },
    {
      "epoch": 6.61637239165329,
      "grad_norm": 0.2236328125,
      "learning_rate": 1.031e-05,
      "loss": 4.0023,
      "step": 1032
    },
    {
      "epoch": 6.622792937399679,
      "grad_norm": 0.2216796875,
      "learning_rate": 1.0320000000000001e-05,
      "loss": 4.0076,
      "step": 1033
    },
    {
      "epoch": 6.629213483146067,
      "grad_norm": 0.138671875,
      "learning_rate": 1.0330000000000001e-05,
      "loss": 4.0015,
      "step": 1034
    },
    {
      "epoch": 6.635634028892456,
      "grad_norm": 5.65625,
      "learning_rate": 1.0340000000000001e-05,
      "loss": 4.0061,
      "step": 1035
    },
    {
      "epoch": 6.642054574638844,
      "grad_norm": 0.361328125,
      "learning_rate": 1.035e-05,
      "loss": 4.0064,
      "step": 1036
    },
    {
      "epoch": 6.648475120385233,
      "grad_norm": 0.353515625,
      "learning_rate": 1.036e-05,
      "loss": 4.0038,
      "step": 1037
    },
    {
      "epoch": 6.654895666131621,
      "grad_norm": 0.376953125,
      "learning_rate": 1.037e-05,
      "loss": 4.0041,
      "step": 1038
    },
    {
      "epoch": 6.66131621187801,
      "grad_norm": 0.53515625,
      "learning_rate": 1.038e-05,
      "loss": 4.0048,
      "step": 1039
    },
    {
      "epoch": 6.667736757624398,
      "grad_norm": 0.1416015625,
      "learning_rate": 1.039e-05,
      "loss": 4.0056,
      "step": 1040
    },
    {
      "epoch": 6.674157303370786,
      "grad_norm": 0.2021484375,
      "learning_rate": 1.04e-05,
      "loss": 4.0041,
      "step": 1041
    },
    {
      "epoch": 6.680577849117175,
      "grad_norm": 0.734375,
      "learning_rate": 1.041e-05,
      "loss": 4.0012,
      "step": 1042
    },
    {
      "epoch": 6.686998394863563,
      "grad_norm": 0.3203125,
      "learning_rate": 1.042e-05,
      "loss": 4.0075,
      "step": 1043
    },
    {
      "epoch": 6.693418940609952,
      "grad_norm": 0.5078125,
      "learning_rate": 1.043e-05,
      "loss": 4.0017,
      "step": 1044
    },
    {
      "epoch": 6.69983948635634,
      "grad_norm": 0.296875,
      "learning_rate": 1.0440000000000002e-05,
      "loss": 4.0057,
      "step": 1045
    },
    {
      "epoch": 6.706260032102729,
      "grad_norm": 0.14453125,
      "learning_rate": 1.045e-05,
      "loss": 4.0033,
      "step": 1046
    },
    {
      "epoch": 6.712680577849117,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.046e-05,
      "loss": 4.0066,
      "step": 1047
    },
    {
      "epoch": 6.719101123595506,
      "grad_norm": 0.248046875,
      "learning_rate": 1.0470000000000001e-05,
      "loss": 4.0076,
      "step": 1048
    },
    {
      "epoch": 6.725521669341894,
      "grad_norm": 0.318359375,
      "learning_rate": 1.0480000000000001e-05,
      "loss": 4.0022,
      "step": 1049
    },
    {
      "epoch": 6.731942215088282,
      "grad_norm": 0.22265625,
      "learning_rate": 1.049e-05,
      "loss": 4.0016,
      "step": 1050
    },
    {
      "epoch": 6.738362760834671,
      "grad_norm": 3.21875,
      "learning_rate": 1.05e-05,
      "loss": 4.0567,
      "step": 1051
    },
    {
      "epoch": 6.744783306581059,
      "grad_norm": 0.435546875,
      "learning_rate": 1.0510000000000001e-05,
      "loss": 4.0107,
      "step": 1052
    },
    {
      "epoch": 6.751203852327448,
      "grad_norm": 0.38671875,
      "learning_rate": 1.0520000000000001e-05,
      "loss": 4.0042,
      "step": 1053
    },
    {
      "epoch": 6.757624398073836,
      "grad_norm": 0.275390625,
      "learning_rate": 1.053e-05,
      "loss": 4.001,
      "step": 1054
    },
    {
      "epoch": 6.764044943820225,
      "grad_norm": 0.150390625,
      "learning_rate": 1.0539999999999999e-05,
      "loss": 4.0028,
      "step": 1055
    },
    {
      "epoch": 6.770465489566613,
      "grad_norm": 2.09375,
      "learning_rate": 1.055e-05,
      "loss": 4.0139,
      "step": 1056
    },
    {
      "epoch": 6.776886035313002,
      "grad_norm": 0.5703125,
      "learning_rate": 1.056e-05,
      "loss": 4.0039,
      "step": 1057
    },
    {
      "epoch": 6.78330658105939,
      "grad_norm": 0.54296875,
      "learning_rate": 1.057e-05,
      "loss": 4.006,
      "step": 1058
    },
    {
      "epoch": 6.789727126805778,
      "grad_norm": 0.36328125,
      "learning_rate": 1.058e-05,
      "loss": 4.0108,
      "step": 1059
    },
    {
      "epoch": 6.796147672552167,
      "grad_norm": 0.205078125,
      "learning_rate": 1.059e-05,
      "loss": 4.0051,
      "step": 1060
    },
    {
      "epoch": 6.802568218298555,
      "grad_norm": 0.361328125,
      "learning_rate": 1.06e-05,
      "loss": 4.0094,
      "step": 1061
    },
    {
      "epoch": 6.808988764044944,
      "grad_norm": 0.3046875,
      "learning_rate": 1.061e-05,
      "loss": 4.0104,
      "step": 1062
    },
    {
      "epoch": 6.815409309791332,
      "grad_norm": 0.373046875,
      "learning_rate": 1.062e-05,
      "loss": 4.0033,
      "step": 1063
    },
    {
      "epoch": 6.821829855537721,
      "grad_norm": 0.73828125,
      "learning_rate": 1.0630000000000002e-05,
      "loss": 4.0042,
      "step": 1064
    },
    {
      "epoch": 6.828250401284109,
      "grad_norm": 0.39453125,
      "learning_rate": 1.064e-05,
      "loss": 4.003,
      "step": 1065
    },
    {
      "epoch": 6.834670947030498,
      "grad_norm": 0.75390625,
      "learning_rate": 1.065e-05,
      "loss": 4.0121,
      "step": 1066
    },
    {
      "epoch": 6.841091492776886,
      "grad_norm": 0.8984375,
      "learning_rate": 1.0660000000000001e-05,
      "loss": 4.0027,
      "step": 1067
    },
    {
      "epoch": 6.847512038523274,
      "grad_norm": 2.578125,
      "learning_rate": 1.0670000000000001e-05,
      "loss": 4.0036,
      "step": 1068
    },
    {
      "epoch": 6.853932584269663,
      "grad_norm": 2.828125,
      "learning_rate": 1.0680000000000001e-05,
      "loss": 4.0157,
      "step": 1069
    },
    {
      "epoch": 6.860353130016051,
      "grad_norm": 0.58984375,
      "learning_rate": 1.069e-05,
      "loss": 4.0045,
      "step": 1070
    },
    {
      "epoch": 6.86677367576244,
      "grad_norm": 0.275390625,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 4.0025,
      "step": 1071
    },
    {
      "epoch": 6.873194221508828,
      "grad_norm": 0.337890625,
      "learning_rate": 1.071e-05,
      "loss": 4.0013,
      "step": 1072
    },
    {
      "epoch": 6.879614767255217,
      "grad_norm": 0.419921875,
      "learning_rate": 1.072e-05,
      "loss": 4.0034,
      "step": 1073
    },
    {
      "epoch": 6.886035313001605,
      "grad_norm": 0.1494140625,
      "learning_rate": 1.073e-05,
      "loss": 4.0066,
      "step": 1074
    },
    {
      "epoch": 6.892455858747994,
      "grad_norm": 0.408203125,
      "learning_rate": 1.074e-05,
      "loss": 4.0014,
      "step": 1075
    },
    {
      "epoch": 6.898876404494382,
      "grad_norm": 0.26171875,
      "learning_rate": 1.075e-05,
      "loss": 4.0023,
      "step": 1076
    },
    {
      "epoch": 6.90529695024077,
      "grad_norm": 0.19921875,
      "learning_rate": 1.076e-05,
      "loss": 4.0049,
      "step": 1077
    },
    {
      "epoch": 6.911717495987159,
      "grad_norm": 0.474609375,
      "learning_rate": 1.077e-05,
      "loss": 4.003,
      "step": 1078
    },
    {
      "epoch": 6.918138041733547,
      "grad_norm": 0.43359375,
      "learning_rate": 1.0780000000000002e-05,
      "loss": 4.0075,
      "step": 1079
    },
    {
      "epoch": 6.924558587479936,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.079e-05,
      "loss": 4.0068,
      "step": 1080
    },
    {
      "epoch": 6.930979133226324,
      "grad_norm": 0.54296875,
      "learning_rate": 1.08e-05,
      "loss": 4.0014,
      "step": 1081
    },
    {
      "epoch": 6.937399678972713,
      "grad_norm": 0.28125,
      "learning_rate": 1.081e-05,
      "loss": 4.0052,
      "step": 1082
    },
    {
      "epoch": 6.943820224719101,
      "grad_norm": 0.177734375,
      "learning_rate": 1.0820000000000001e-05,
      "loss": 4.0078,
      "step": 1083
    },
    {
      "epoch": 6.95024077046549,
      "grad_norm": 0.74609375,
      "learning_rate": 1.083e-05,
      "loss": 4.0045,
      "step": 1084
    },
    {
      "epoch": 6.956661316211878,
      "grad_norm": 0.5625,
      "learning_rate": 1.084e-05,
      "loss": 4.0038,
      "step": 1085
    },
    {
      "epoch": 6.963081861958266,
      "grad_norm": 1.2109375,
      "learning_rate": 1.0850000000000001e-05,
      "loss": 4.0037,
      "step": 1086
    },
    {
      "epoch": 6.969502407704655,
      "grad_norm": 1.109375,
      "learning_rate": 1.0860000000000001e-05,
      "loss": 4.0069,
      "step": 1087
    },
    {
      "epoch": 6.975922953451043,
      "grad_norm": 2.46875,
      "learning_rate": 1.0870000000000001e-05,
      "loss": 4.006,
      "step": 1088
    },
    {
      "epoch": 6.982343499197432,
      "grad_norm": 2.53125,
      "learning_rate": 1.088e-05,
      "loss": 4.0039,
      "step": 1089
    },
    {
      "epoch": 6.98876404494382,
      "grad_norm": 0.6484375,
      "learning_rate": 1.089e-05,
      "loss": 4.0052,
      "step": 1090
    },
    {
      "epoch": 6.995184590690209,
      "grad_norm": 4.53125,
      "learning_rate": 1.09e-05,
      "loss": 4.0054,
      "step": 1091
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.546875,
      "learning_rate": 1.091e-05,
      "loss": 3.0077,
      "step": 1092
    },
    {
      "epoch": 7.006420545746389,
      "grad_norm": 3.390625,
      "learning_rate": 1.092e-05,
      "loss": 4.0072,
      "step": 1093
    },
    {
      "epoch": 7.012841091492777,
      "grad_norm": 2.921875,
      "learning_rate": 1.093e-05,
      "loss": 4.0086,
      "step": 1094
    },
    {
      "epoch": 7.019261637239166,
      "grad_norm": 3.421875,
      "learning_rate": 1.094e-05,
      "loss": 4.0105,
      "step": 1095
    },
    {
      "epoch": 7.025682182985554,
      "grad_norm": 2.84375,
      "learning_rate": 1.095e-05,
      "loss": 4.0075,
      "step": 1096
    },
    {
      "epoch": 7.032102728731942,
      "grad_norm": 2.265625,
      "learning_rate": 1.096e-05,
      "loss": 4.0114,
      "step": 1097
    },
    {
      "epoch": 7.03852327447833,
      "grad_norm": 2.59375,
      "learning_rate": 1.0970000000000002e-05,
      "loss": 4.0073,
      "step": 1098
    },
    {
      "epoch": 7.044943820224719,
      "grad_norm": 1.5703125,
      "learning_rate": 1.098e-05,
      "loss": 4.0032,
      "step": 1099
    },
    {
      "epoch": 7.051364365971107,
      "grad_norm": 4.125,
      "learning_rate": 1.099e-05,
      "loss": 4.0071,
      "step": 1100
    },
    {
      "epoch": 7.057784911717496,
      "grad_norm": 4.21875,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 4.01,
      "step": 1101
    },
    {
      "epoch": 7.064205457463885,
      "grad_norm": 3.546875,
      "learning_rate": 1.1010000000000001e-05,
      "loss": 4.0085,
      "step": 1102
    },
    {
      "epoch": 7.070626003210273,
      "grad_norm": 1.203125,
      "learning_rate": 1.1020000000000001e-05,
      "loss": 4.0014,
      "step": 1103
    },
    {
      "epoch": 7.077046548956662,
      "grad_norm": 1.3828125,
      "learning_rate": 1.103e-05,
      "loss": 4.0104,
      "step": 1104
    },
    {
      "epoch": 7.08346709470305,
      "grad_norm": 1.25,
      "learning_rate": 1.1040000000000001e-05,
      "loss": 4.0057,
      "step": 1105
    },
    {
      "epoch": 7.089887640449438,
      "grad_norm": 2.140625,
      "learning_rate": 1.1050000000000001e-05,
      "loss": 4.0034,
      "step": 1106
    },
    {
      "epoch": 7.096308186195826,
      "grad_norm": 2.53125,
      "learning_rate": 1.106e-05,
      "loss": 4.0049,
      "step": 1107
    },
    {
      "epoch": 7.102728731942215,
      "grad_norm": 1.34375,
      "learning_rate": 1.107e-05,
      "loss": 4.0021,
      "step": 1108
    },
    {
      "epoch": 7.109149277688603,
      "grad_norm": 4.1875,
      "learning_rate": 1.108e-05,
      "loss": 4.0097,
      "step": 1109
    },
    {
      "epoch": 7.115569823434992,
      "grad_norm": 4.71875,
      "learning_rate": 1.109e-05,
      "loss": 4.0192,
      "step": 1110
    },
    {
      "epoch": 7.121990369181381,
      "grad_norm": 4.25,
      "learning_rate": 1.11e-05,
      "loss": 4.0233,
      "step": 1111
    },
    {
      "epoch": 7.128410914927769,
      "grad_norm": 0.447265625,
      "learning_rate": 1.111e-05,
      "loss": 4.0035,
      "step": 1112
    },
    {
      "epoch": 7.134831460674158,
      "grad_norm": 1.84375,
      "learning_rate": 1.112e-05,
      "loss": 4.0027,
      "step": 1113
    },
    {
      "epoch": 7.141252006420546,
      "grad_norm": 1.703125,
      "learning_rate": 1.113e-05,
      "loss": 4.0061,
      "step": 1114
    },
    {
      "epoch": 7.147672552166934,
      "grad_norm": 0.5390625,
      "learning_rate": 1.114e-05,
      "loss": 3.9985,
      "step": 1115
    },
    {
      "epoch": 7.154093097913322,
      "grad_norm": 0.98828125,
      "learning_rate": 1.115e-05,
      "loss": 4.0024,
      "step": 1116
    },
    {
      "epoch": 7.160513643659711,
      "grad_norm": 0.482421875,
      "learning_rate": 1.1160000000000002e-05,
      "loss": 4.0057,
      "step": 1117
    },
    {
      "epoch": 7.166934189406099,
      "grad_norm": 1.1953125,
      "learning_rate": 1.117e-05,
      "loss": 4.0015,
      "step": 1118
    },
    {
      "epoch": 7.173354735152488,
      "grad_norm": 1.0078125,
      "learning_rate": 1.118e-05,
      "loss": 4.0053,
      "step": 1119
    },
    {
      "epoch": 7.179775280898877,
      "grad_norm": 0.87109375,
      "learning_rate": 1.1190000000000001e-05,
      "loss": 4.0044,
      "step": 1120
    },
    {
      "epoch": 7.186195826645265,
      "grad_norm": 1.1171875,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 4.0054,
      "step": 1121
    },
    {
      "epoch": 7.192616372391654,
      "grad_norm": 0.291015625,
      "learning_rate": 1.1210000000000001e-05,
      "loss": 4.0051,
      "step": 1122
    },
    {
      "epoch": 7.199036918138042,
      "grad_norm": 1.0703125,
      "learning_rate": 1.122e-05,
      "loss": 4.0016,
      "step": 1123
    },
    {
      "epoch": 7.20545746388443,
      "grad_norm": 0.87890625,
      "learning_rate": 1.1230000000000001e-05,
      "loss": 4.0034,
      "step": 1124
    },
    {
      "epoch": 7.211878009630818,
      "grad_norm": 1.140625,
      "learning_rate": 1.124e-05,
      "loss": 4.0095,
      "step": 1125
    },
    {
      "epoch": 7.218298555377207,
      "grad_norm": 1.1484375,
      "learning_rate": 1.125e-05,
      "loss": 4.0069,
      "step": 1126
    },
    {
      "epoch": 7.224719101123595,
      "grad_norm": 0.640625,
      "learning_rate": 1.126e-05,
      "loss": 4.0046,
      "step": 1127
    },
    {
      "epoch": 7.231139646869984,
      "grad_norm": 0.90625,
      "learning_rate": 1.127e-05,
      "loss": 4.0041,
      "step": 1128
    },
    {
      "epoch": 7.237560192616373,
      "grad_norm": 0.294921875,
      "learning_rate": 1.128e-05,
      "loss": 4.0087,
      "step": 1129
    },
    {
      "epoch": 7.243980738362761,
      "grad_norm": 0.7265625,
      "learning_rate": 1.129e-05,
      "loss": 4.0043,
      "step": 1130
    },
    {
      "epoch": 7.25040128410915,
      "grad_norm": 0.6796875,
      "learning_rate": 1.13e-05,
      "loss": 4.0102,
      "step": 1131
    },
    {
      "epoch": 7.256821829855538,
      "grad_norm": 0.330078125,
      "learning_rate": 1.1310000000000002e-05,
      "loss": 4.004,
      "step": 1132
    },
    {
      "epoch": 7.263242375601926,
      "grad_norm": 0.373046875,
      "learning_rate": 1.132e-05,
      "loss": 4.0029,
      "step": 1133
    },
    {
      "epoch": 7.269662921348314,
      "grad_norm": 0.443359375,
      "learning_rate": 1.133e-05,
      "loss": 4.008,
      "step": 1134
    },
    {
      "epoch": 7.276083467094703,
      "grad_norm": 0.1611328125,
      "learning_rate": 1.134e-05,
      "loss": 4.0039,
      "step": 1135
    },
    {
      "epoch": 7.282504012841091,
      "grad_norm": 0.54296875,
      "learning_rate": 1.1350000000000001e-05,
      "loss": 4.0096,
      "step": 1136
    },
    {
      "epoch": 7.28892455858748,
      "grad_norm": 0.40625,
      "learning_rate": 1.1360000000000001e-05,
      "loss": 4.0035,
      "step": 1137
    },
    {
      "epoch": 7.295345104333869,
      "grad_norm": 0.62109375,
      "learning_rate": 1.137e-05,
      "loss": 4.002,
      "step": 1138
    },
    {
      "epoch": 7.301765650080257,
      "grad_norm": 0.1630859375,
      "learning_rate": 1.1380000000000001e-05,
      "loss": 4.0019,
      "step": 1139
    },
    {
      "epoch": 7.308186195826646,
      "grad_norm": 0.353515625,
      "learning_rate": 1.1390000000000001e-05,
      "loss": 4.0045,
      "step": 1140
    },
    {
      "epoch": 7.314606741573034,
      "grad_norm": 0.310546875,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 4.006,
      "step": 1141
    },
    {
      "epoch": 7.321027287319422,
      "grad_norm": 0.21875,
      "learning_rate": 1.141e-05,
      "loss": 4.0087,
      "step": 1142
    },
    {
      "epoch": 7.32744783306581,
      "grad_norm": 0.130859375,
      "learning_rate": 1.142e-05,
      "loss": 4.0046,
      "step": 1143
    },
    {
      "epoch": 7.333868378812199,
      "grad_norm": 0.33984375,
      "learning_rate": 1.143e-05,
      "loss": 4.0063,
      "step": 1144
    },
    {
      "epoch": 7.340288924558587,
      "grad_norm": 0.169921875,
      "learning_rate": 1.144e-05,
      "loss": 4.0065,
      "step": 1145
    },
    {
      "epoch": 7.346709470304976,
      "grad_norm": 0.1474609375,
      "learning_rate": 1.145e-05,
      "loss": 4.0014,
      "step": 1146
    },
    {
      "epoch": 7.353130016051364,
      "grad_norm": 0.1279296875,
      "learning_rate": 1.146e-05,
      "loss": 4.0048,
      "step": 1147
    },
    {
      "epoch": 7.359550561797753,
      "grad_norm": 0.158203125,
      "learning_rate": 1.147e-05,
      "loss": 4.0033,
      "step": 1148
    },
    {
      "epoch": 7.365971107544142,
      "grad_norm": 0.287109375,
      "learning_rate": 1.148e-05,
      "loss": 4.0011,
      "step": 1149
    },
    {
      "epoch": 7.37239165329053,
      "grad_norm": 0.142578125,
      "learning_rate": 1.149e-05,
      "loss": 4.0006,
      "step": 1150
    },
    {
      "epoch": 7.378812199036918,
      "grad_norm": 0.1787109375,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 4.0012,
      "step": 1151
    },
    {
      "epoch": 7.385232744783306,
      "grad_norm": 0.30078125,
      "learning_rate": 1.151e-05,
      "loss": 4.0076,
      "step": 1152
    },
    {
      "epoch": 7.391653290529695,
      "grad_norm": 0.162109375,
      "learning_rate": 1.152e-05,
      "loss": 4.0054,
      "step": 1153
    },
    {
      "epoch": 7.398073836276083,
      "grad_norm": 0.1484375,
      "learning_rate": 1.153e-05,
      "loss": 4.002,
      "step": 1154
    },
    {
      "epoch": 7.404494382022472,
      "grad_norm": 0.142578125,
      "learning_rate": 1.1540000000000001e-05,
      "loss": 4.0049,
      "step": 1155
    },
    {
      "epoch": 7.41091492776886,
      "grad_norm": 0.15234375,
      "learning_rate": 1.1550000000000001e-05,
      "loss": 4.0065,
      "step": 1156
    },
    {
      "epoch": 7.417335473515249,
      "grad_norm": 0.140625,
      "learning_rate": 1.156e-05,
      "loss": 4.0025,
      "step": 1157
    },
    {
      "epoch": 7.423756019261638,
      "grad_norm": 0.162109375,
      "learning_rate": 1.1570000000000001e-05,
      "loss": 4.0032,
      "step": 1158
    },
    {
      "epoch": 7.430176565008026,
      "grad_norm": 0.2578125,
      "learning_rate": 1.1580000000000001e-05,
      "loss": 4.0041,
      "step": 1159
    },
    {
      "epoch": 7.436597110754414,
      "grad_norm": 0.142578125,
      "learning_rate": 1.159e-05,
      "loss": 4.0032,
      "step": 1160
    },
    {
      "epoch": 7.443017656500802,
      "grad_norm": 0.2109375,
      "learning_rate": 1.16e-05,
      "loss": 4.0064,
      "step": 1161
    },
    {
      "epoch": 7.449438202247191,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.161e-05,
      "loss": 4.0033,
      "step": 1162
    },
    {
      "epoch": 7.455858747993579,
      "grad_norm": 0.220703125,
      "learning_rate": 1.162e-05,
      "loss": 4.004,
      "step": 1163
    },
    {
      "epoch": 7.462279293739968,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.163e-05,
      "loss": 4.0033,
      "step": 1164
    },
    {
      "epoch": 7.468699839486356,
      "grad_norm": 0.53515625,
      "learning_rate": 1.164e-05,
      "loss": 4.0067,
      "step": 1165
    },
    {
      "epoch": 7.475120385232745,
      "grad_norm": 0.396484375,
      "learning_rate": 1.1650000000000002e-05,
      "loss": 4.0043,
      "step": 1166
    },
    {
      "epoch": 7.481540930979134,
      "grad_norm": 0.1328125,
      "learning_rate": 1.166e-05,
      "loss": 4.0016,
      "step": 1167
    },
    {
      "epoch": 7.487961476725522,
      "grad_norm": 0.3125,
      "learning_rate": 1.167e-05,
      "loss": 4.0053,
      "step": 1168
    },
    {
      "epoch": 7.49438202247191,
      "grad_norm": 0.255859375,
      "learning_rate": 1.168e-05,
      "loss": 4.0021,
      "step": 1169
    },
    {
      "epoch": 7.500802568218298,
      "grad_norm": 0.296875,
      "learning_rate": 1.1690000000000002e-05,
      "loss": 4.0021,
      "step": 1170
    },
    {
      "epoch": 7.507223113964687,
      "grad_norm": 0.294921875,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 4.0047,
      "step": 1171
    },
    {
      "epoch": 7.513643659711075,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.171e-05,
      "loss": 4.0036,
      "step": 1172
    },
    {
      "epoch": 7.520064205457464,
      "grad_norm": 0.357421875,
      "learning_rate": 1.172e-05,
      "loss": 4.0019,
      "step": 1173
    },
    {
      "epoch": 7.526484751203852,
      "grad_norm": 0.1796875,
      "learning_rate": 1.1730000000000001e-05,
      "loss": 4.0061,
      "step": 1174
    },
    {
      "epoch": 7.532905296950241,
      "grad_norm": 0.1630859375,
      "learning_rate": 1.1740000000000001e-05,
      "loss": 4.003,
      "step": 1175
    },
    {
      "epoch": 7.539325842696629,
      "grad_norm": 0.7734375,
      "learning_rate": 1.175e-05,
      "loss": 4.0044,
      "step": 1176
    },
    {
      "epoch": 7.545746388443018,
      "grad_norm": 0.1728515625,
      "learning_rate": 1.1760000000000001e-05,
      "loss": 4.006,
      "step": 1177
    },
    {
      "epoch": 7.552166934189406,
      "grad_norm": 0.1787109375,
      "learning_rate": 1.177e-05,
      "loss": 4.0059,
      "step": 1178
    },
    {
      "epoch": 7.558587479935794,
      "grad_norm": 0.3125,
      "learning_rate": 1.178e-05,
      "loss": 4.0046,
      "step": 1179
    },
    {
      "epoch": 7.565008025682183,
      "grad_norm": 0.275390625,
      "learning_rate": 1.179e-05,
      "loss": 4.0044,
      "step": 1180
    },
    {
      "epoch": 7.571428571428571,
      "grad_norm": 0.1669921875,
      "learning_rate": 1.18e-05,
      "loss": 3.9998,
      "step": 1181
    },
    {
      "epoch": 7.57784911717496,
      "grad_norm": 0.294921875,
      "learning_rate": 1.181e-05,
      "loss": 4.0089,
      "step": 1182
    },
    {
      "epoch": 7.584269662921348,
      "grad_norm": 0.19921875,
      "learning_rate": 1.182e-05,
      "loss": 4.0036,
      "step": 1183
    },
    {
      "epoch": 7.590690208667737,
      "grad_norm": 0.228515625,
      "learning_rate": 1.183e-05,
      "loss": 4.0024,
      "step": 1184
    },
    {
      "epoch": 7.597110754414125,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.1840000000000002e-05,
      "loss": 4.0057,
      "step": 1185
    },
    {
      "epoch": 7.603531300160514,
      "grad_norm": 1.78125,
      "learning_rate": 1.185e-05,
      "loss": 4.0026,
      "step": 1186
    },
    {
      "epoch": 7.609951845906902,
      "grad_norm": 0.2451171875,
      "learning_rate": 1.186e-05,
      "loss": 4.0099,
      "step": 1187
    },
    {
      "epoch": 7.61637239165329,
      "grad_norm": 0.142578125,
      "learning_rate": 1.187e-05,
      "loss": 4.0026,
      "step": 1188
    },
    {
      "epoch": 7.622792937399679,
      "grad_norm": 0.2734375,
      "learning_rate": 1.1880000000000001e-05,
      "loss": 4.0039,
      "step": 1189
    },
    {
      "epoch": 7.629213483146067,
      "grad_norm": 0.2333984375,
      "learning_rate": 1.1890000000000001e-05,
      "loss": 4.0062,
      "step": 1190
    },
    {
      "epoch": 7.635634028892456,
      "grad_norm": 0.1455078125,
      "learning_rate": 1.19e-05,
      "loss": 4.0057,
      "step": 1191
    },
    {
      "epoch": 7.642054574638844,
      "grad_norm": 0.2060546875,
      "learning_rate": 1.1910000000000001e-05,
      "loss": 4.0018,
      "step": 1192
    },
    {
      "epoch": 7.648475120385233,
      "grad_norm": 0.609375,
      "learning_rate": 1.1920000000000001e-05,
      "loss": 4.0065,
      "step": 1193
    },
    {
      "epoch": 7.654895666131621,
      "grad_norm": 0.1943359375,
      "learning_rate": 1.1930000000000001e-05,
      "loss": 4.0013,
      "step": 1194
    },
    {
      "epoch": 7.66131621187801,
      "grad_norm": 0.1533203125,
      "learning_rate": 1.1940000000000001e-05,
      "loss": 4.0021,
      "step": 1195
    },
    {
      "epoch": 7.667736757624398,
      "grad_norm": 0.189453125,
      "learning_rate": 1.195e-05,
      "loss": 4.0062,
      "step": 1196
    },
    {
      "epoch": 7.674157303370786,
      "grad_norm": 0.3203125,
      "learning_rate": 1.196e-05,
      "loss": 4.0002,
      "step": 1197
    },
    {
      "epoch": 7.680577849117175,
      "grad_norm": 0.40234375,
      "learning_rate": 1.197e-05,
      "loss": 4.0025,
      "step": 1198
    },
    {
      "epoch": 7.686998394863563,
      "grad_norm": 0.18359375,
      "learning_rate": 1.198e-05,
      "loss": 4.0095,
      "step": 1199
    },
    {
      "epoch": 7.693418940609952,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.199e-05,
      "loss": 4.0064,
      "step": 1200
    },
    {
      "epoch": 7.69983948635634,
      "grad_norm": 0.431640625,
      "learning_rate": 1.2e-05,
      "loss": 4.0082,
      "step": 1201
    },
    {
      "epoch": 7.706260032102729,
      "grad_norm": 0.1474609375,
      "learning_rate": 1.201e-05,
      "loss": 4.0045,
      "step": 1202
    },
    {
      "epoch": 7.712680577849117,
      "grad_norm": 0.236328125,
      "learning_rate": 1.202e-05,
      "loss": 4.0015,
      "step": 1203
    },
    {
      "epoch": 7.719101123595506,
      "grad_norm": 0.1484375,
      "learning_rate": 1.2030000000000002e-05,
      "loss": 4.001,
      "step": 1204
    },
    {
      "epoch": 7.725521669341894,
      "grad_norm": 0.59765625,
      "learning_rate": 1.204e-05,
      "loss": 4.0048,
      "step": 1205
    },
    {
      "epoch": 7.731942215088282,
      "grad_norm": 0.232421875,
      "learning_rate": 1.205e-05,
      "loss": 4.009,
      "step": 1206
    },
    {
      "epoch": 7.738362760834671,
      "grad_norm": 0.251953125,
      "learning_rate": 1.206e-05,
      "loss": 4.0008,
      "step": 1207
    },
    {
      "epoch": 7.744783306581059,
      "grad_norm": 0.390625,
      "learning_rate": 1.2070000000000001e-05,
      "loss": 4.01,
      "step": 1208
    },
    {
      "epoch": 7.751203852327448,
      "grad_norm": 0.248046875,
      "learning_rate": 1.2080000000000001e-05,
      "loss": 3.999,
      "step": 1209
    },
    {
      "epoch": 7.757624398073836,
      "grad_norm": 12.25,
      "learning_rate": 1.209e-05,
      "loss": 4.0309,
      "step": 1210
    },
    {
      "epoch": 7.764044943820225,
      "grad_norm": 0.224609375,
      "learning_rate": 1.2100000000000001e-05,
      "loss": 4.0024,
      "step": 1211
    },
    {
      "epoch": 7.770465489566613,
      "grad_norm": 0.236328125,
      "learning_rate": 1.2110000000000001e-05,
      "loss": 4.0025,
      "step": 1212
    },
    {
      "epoch": 7.776886035313002,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.2120000000000001e-05,
      "loss": 4.0019,
      "step": 1213
    },
    {
      "epoch": 7.78330658105939,
      "grad_norm": 12.625,
      "learning_rate": 1.213e-05,
      "loss": 4.0084,
      "step": 1214
    },
    {
      "epoch": 7.789727126805778,
      "grad_norm": 0.373046875,
      "learning_rate": 1.214e-05,
      "loss": 4.0017,
      "step": 1215
    },
    {
      "epoch": 7.796147672552167,
      "grad_norm": 0.36328125,
      "learning_rate": 1.215e-05,
      "loss": 4.0031,
      "step": 1216
    },
    {
      "epoch": 7.802568218298555,
      "grad_norm": 0.193359375,
      "learning_rate": 1.216e-05,
      "loss": 4.0036,
      "step": 1217
    },
    {
      "epoch": 7.808988764044944,
      "grad_norm": 0.169921875,
      "learning_rate": 1.217e-05,
      "loss": 4.0024,
      "step": 1218
    },
    {
      "epoch": 7.815409309791332,
      "grad_norm": 0.412109375,
      "learning_rate": 1.2180000000000002e-05,
      "loss": 4.0061,
      "step": 1219
    },
    {
      "epoch": 7.821829855537721,
      "grad_norm": 0.19140625,
      "learning_rate": 1.219e-05,
      "loss": 4.005,
      "step": 1220
    },
    {
      "epoch": 7.828250401284109,
      "grad_norm": 0.1552734375,
      "learning_rate": 1.22e-05,
      "loss": 4.0066,
      "step": 1221
    },
    {
      "epoch": 7.834670947030498,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.221e-05,
      "loss": 4.0052,
      "step": 1222
    },
    {
      "epoch": 7.841091492776886,
      "grad_norm": 0.1943359375,
      "learning_rate": 1.2220000000000002e-05,
      "loss": 4.0035,
      "step": 1223
    },
    {
      "epoch": 7.847512038523274,
      "grad_norm": 0.166015625,
      "learning_rate": 1.2230000000000001e-05,
      "loss": 4.0044,
      "step": 1224
    },
    {
      "epoch": 7.853932584269663,
      "grad_norm": 0.2578125,
      "learning_rate": 1.224e-05,
      "loss": 4.0134,
      "step": 1225
    },
    {
      "epoch": 7.860353130016051,
      "grad_norm": 0.162109375,
      "learning_rate": 1.225e-05,
      "loss": 4.002,
      "step": 1226
    },
    {
      "epoch": 7.86677367576244,
      "grad_norm": 0.41015625,
      "learning_rate": 1.2260000000000001e-05,
      "loss": 4.0051,
      "step": 1227
    },
    {
      "epoch": 7.873194221508828,
      "grad_norm": 0.2216796875,
      "learning_rate": 1.2270000000000001e-05,
      "loss": 4.0034,
      "step": 1228
    },
    {
      "epoch": 7.879614767255217,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.2280000000000001e-05,
      "loss": 4.0017,
      "step": 1229
    },
    {
      "epoch": 7.886035313001605,
      "grad_norm": 0.49609375,
      "learning_rate": 1.2290000000000001e-05,
      "loss": 4.0013,
      "step": 1230
    },
    {
      "epoch": 7.892455858747994,
      "grad_norm": 0.294921875,
      "learning_rate": 1.23e-05,
      "loss": 4.003,
      "step": 1231
    },
    {
      "epoch": 7.898876404494382,
      "grad_norm": 0.138671875,
      "learning_rate": 1.231e-05,
      "loss": 4.003,
      "step": 1232
    },
    {
      "epoch": 7.90529695024077,
      "grad_norm": 0.1982421875,
      "learning_rate": 1.232e-05,
      "loss": 4.0014,
      "step": 1233
    },
    {
      "epoch": 7.911717495987159,
      "grad_norm": 0.4296875,
      "learning_rate": 1.233e-05,
      "loss": 4.0054,
      "step": 1234
    },
    {
      "epoch": 7.918138041733547,
      "grad_norm": 0.466796875,
      "learning_rate": 1.234e-05,
      "loss": 4.0079,
      "step": 1235
    },
    {
      "epoch": 7.924558587479936,
      "grad_norm": 0.34765625,
      "learning_rate": 1.235e-05,
      "loss": 4.0071,
      "step": 1236
    },
    {
      "epoch": 7.930979133226324,
      "grad_norm": 0.3828125,
      "learning_rate": 1.236e-05,
      "loss": 4.0011,
      "step": 1237
    },
    {
      "epoch": 7.937399678972713,
      "grad_norm": 0.2099609375,
      "learning_rate": 1.2370000000000002e-05,
      "loss": 4.003,
      "step": 1238
    },
    {
      "epoch": 7.943820224719101,
      "grad_norm": 0.1875,
      "learning_rate": 1.238e-05,
      "loss": 4.0039,
      "step": 1239
    },
    {
      "epoch": 7.95024077046549,
      "grad_norm": 0.2734375,
      "learning_rate": 1.239e-05,
      "loss": 4.0066,
      "step": 1240
    },
    {
      "epoch": 7.956661316211878,
      "grad_norm": 0.41015625,
      "learning_rate": 1.24e-05,
      "loss": 4.005,
      "step": 1241
    },
    {
      "epoch": 7.963081861958266,
      "grad_norm": 0.169921875,
      "learning_rate": 1.2410000000000001e-05,
      "loss": 4.0076,
      "step": 1242
    },
    {
      "epoch": 7.969502407704655,
      "grad_norm": 0.203125,
      "learning_rate": 1.2420000000000001e-05,
      "loss": 4.0038,
      "step": 1243
    },
    {
      "epoch": 7.975922953451043,
      "grad_norm": 0.2314453125,
      "learning_rate": 1.243e-05,
      "loss": 4.0057,
      "step": 1244
    },
    {
      "epoch": 7.982343499197432,
      "grad_norm": 0.26953125,
      "learning_rate": 1.244e-05,
      "loss": 4.0069,
      "step": 1245
    },
    {
      "epoch": 7.98876404494382,
      "grad_norm": 0.1748046875,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 4.006,
      "step": 1246
    },
    {
      "epoch": 7.995184590690209,
      "grad_norm": 0.5859375,
      "learning_rate": 1.2460000000000001e-05,
      "loss": 4.0012,
      "step": 1247
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.5546875,
      "learning_rate": 1.2470000000000001e-05,
      "loss": 2.9977,
      "step": 1248
    },
    {
      "epoch": 8.006420545746389,
      "grad_norm": 0.54296875,
      "learning_rate": 1.248e-05,
      "loss": 4.0049,
      "step": 1249
    },
    {
      "epoch": 8.012841091492778,
      "grad_norm": 0.625,
      "learning_rate": 1.249e-05,
      "loss": 4.0104,
      "step": 1250
    },
    {
      "epoch": 8.019261637239165,
      "grad_norm": 0.373046875,
      "learning_rate": 1.25e-05,
      "loss": 4.0068,
      "step": 1251
    },
    {
      "epoch": 8.025682182985554,
      "grad_norm": 0.2333984375,
      "learning_rate": 1.2509999999999999e-05,
      "loss": 4.0083,
      "step": 1252
    },
    {
      "epoch": 8.032102728731942,
      "grad_norm": 0.416015625,
      "learning_rate": 1.252e-05,
      "loss": 4.0079,
      "step": 1253
    },
    {
      "epoch": 8.038523274478331,
      "grad_norm": 0.32421875,
      "learning_rate": 1.253e-05,
      "loss": 4.0031,
      "step": 1254
    },
    {
      "epoch": 8.044943820224718,
      "grad_norm": 0.1669921875,
      "learning_rate": 1.2540000000000002e-05,
      "loss": 4.0003,
      "step": 1255
    },
    {
      "epoch": 8.051364365971107,
      "grad_norm": 0.65234375,
      "learning_rate": 1.255e-05,
      "loss": 4.0054,
      "step": 1256
    },
    {
      "epoch": 8.057784911717496,
      "grad_norm": 0.353515625,
      "learning_rate": 1.256e-05,
      "loss": 4.0017,
      "step": 1257
    },
    {
      "epoch": 8.064205457463885,
      "grad_norm": 0.59765625,
      "learning_rate": 1.2570000000000002e-05,
      "loss": 4.0051,
      "step": 1258
    },
    {
      "epoch": 8.070626003210274,
      "grad_norm": 0.306640625,
      "learning_rate": 1.258e-05,
      "loss": 4.005,
      "step": 1259
    },
    {
      "epoch": 8.07704654895666,
      "grad_norm": 0.40234375,
      "learning_rate": 1.2590000000000001e-05,
      "loss": 4.0076,
      "step": 1260
    },
    {
      "epoch": 8.08346709470305,
      "grad_norm": 0.66015625,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 4.0018,
      "step": 1261
    },
    {
      "epoch": 8.089887640449438,
      "grad_norm": 1.8125,
      "learning_rate": 1.261e-05,
      "loss": 4.0041,
      "step": 1262
    },
    {
      "epoch": 8.096308186195827,
      "grad_norm": 0.36328125,
      "learning_rate": 1.2620000000000001e-05,
      "loss": 4.0031,
      "step": 1263
    },
    {
      "epoch": 8.102728731942214,
      "grad_norm": 0.4375,
      "learning_rate": 1.263e-05,
      "loss": 4.0068,
      "step": 1264
    },
    {
      "epoch": 8.109149277688603,
      "grad_norm": 0.291015625,
      "learning_rate": 1.2640000000000003e-05,
      "loss": 4.0041,
      "step": 1265
    },
    {
      "epoch": 8.115569823434992,
      "grad_norm": 1.0546875,
      "learning_rate": 1.2650000000000001e-05,
      "loss": 4.0051,
      "step": 1266
    },
    {
      "epoch": 8.12199036918138,
      "grad_norm": 0.1416015625,
      "learning_rate": 1.2659999999999999e-05,
      "loss": 4.0008,
      "step": 1267
    },
    {
      "epoch": 8.12841091492777,
      "grad_norm": 0.296875,
      "learning_rate": 1.267e-05,
      "loss": 3.9988,
      "step": 1268
    },
    {
      "epoch": 8.134831460674157,
      "grad_norm": 0.275390625,
      "learning_rate": 1.268e-05,
      "loss": 4.0076,
      "step": 1269
    },
    {
      "epoch": 8.141252006420546,
      "grad_norm": 0.1533203125,
      "learning_rate": 1.2690000000000002e-05,
      "loss": 4.0033,
      "step": 1270
    },
    {
      "epoch": 8.147672552166934,
      "grad_norm": 0.294921875,
      "learning_rate": 1.27e-05,
      "loss": 4.007,
      "step": 1271
    },
    {
      "epoch": 8.154093097913323,
      "grad_norm": 0.26171875,
      "learning_rate": 1.271e-05,
      "loss": 4.0051,
      "step": 1272
    },
    {
      "epoch": 8.16051364365971,
      "grad_norm": 0.380859375,
      "learning_rate": 1.2720000000000002e-05,
      "loss": 4.0051,
      "step": 1273
    },
    {
      "epoch": 8.1669341894061,
      "grad_norm": 0.4765625,
      "learning_rate": 1.273e-05,
      "loss": 4.0028,
      "step": 1274
    },
    {
      "epoch": 8.173354735152488,
      "grad_norm": 0.146484375,
      "learning_rate": 1.2740000000000002e-05,
      "loss": 4.0026,
      "step": 1275
    },
    {
      "epoch": 8.179775280898877,
      "grad_norm": 0.1982421875,
      "learning_rate": 1.2750000000000002e-05,
      "loss": 4.0028,
      "step": 1276
    },
    {
      "epoch": 8.186195826645266,
      "grad_norm": 0.306640625,
      "learning_rate": 1.276e-05,
      "loss": 4.0047,
      "step": 1277
    },
    {
      "epoch": 8.192616372391653,
      "grad_norm": 0.3125,
      "learning_rate": 1.2770000000000001e-05,
      "loss": 4.0034,
      "step": 1278
    },
    {
      "epoch": 8.199036918138042,
      "grad_norm": 0.162109375,
      "learning_rate": 1.278e-05,
      "loss": 4.0053,
      "step": 1279
    },
    {
      "epoch": 8.20545746388443,
      "grad_norm": 0.380859375,
      "learning_rate": 1.2790000000000001e-05,
      "loss": 4.0039,
      "step": 1280
    },
    {
      "epoch": 8.21187800963082,
      "grad_norm": 0.4609375,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 4.0079,
      "step": 1281
    },
    {
      "epoch": 8.218298555377206,
      "grad_norm": 0.458984375,
      "learning_rate": 1.281e-05,
      "loss": 4.004,
      "step": 1282
    },
    {
      "epoch": 8.224719101123595,
      "grad_norm": 0.51171875,
      "learning_rate": 1.2820000000000001e-05,
      "loss": 4.0052,
      "step": 1283
    },
    {
      "epoch": 8.231139646869984,
      "grad_norm": 0.1279296875,
      "learning_rate": 1.283e-05,
      "loss": 4.0023,
      "step": 1284
    },
    {
      "epoch": 8.237560192616373,
      "grad_norm": 0.55078125,
      "learning_rate": 1.2839999999999999e-05,
      "loss": 4.0027,
      "step": 1285
    },
    {
      "epoch": 8.243980738362762,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.285e-05,
      "loss": 4.0029,
      "step": 1286
    },
    {
      "epoch": 8.250401284109149,
      "grad_norm": 0.1337890625,
      "learning_rate": 1.286e-05,
      "loss": 4.0034,
      "step": 1287
    },
    {
      "epoch": 8.256821829855538,
      "grad_norm": 0.28125,
      "learning_rate": 1.2870000000000002e-05,
      "loss": 4.0042,
      "step": 1288
    },
    {
      "epoch": 8.263242375601926,
      "grad_norm": 0.384765625,
      "learning_rate": 1.288e-05,
      "loss": 4.0046,
      "step": 1289
    },
    {
      "epoch": 8.269662921348315,
      "grad_norm": 0.2490234375,
      "learning_rate": 1.2889999999999999e-05,
      "loss": 4.0031,
      "step": 1290
    },
    {
      "epoch": 8.276083467094702,
      "grad_norm": 1.0859375,
      "learning_rate": 1.29e-05,
      "loss": 4.0051,
      "step": 1291
    },
    {
      "epoch": 8.282504012841091,
      "grad_norm": 0.2470703125,
      "learning_rate": 1.291e-05,
      "loss": 4.005,
      "step": 1292
    },
    {
      "epoch": 8.28892455858748,
      "grad_norm": 0.1513671875,
      "learning_rate": 1.2920000000000002e-05,
      "loss": 4.0048,
      "step": 1293
    },
    {
      "epoch": 8.295345104333869,
      "grad_norm": 0.20703125,
      "learning_rate": 1.293e-05,
      "loss": 4.0092,
      "step": 1294
    },
    {
      "epoch": 8.301765650080256,
      "grad_norm": 0.400390625,
      "learning_rate": 1.294e-05,
      "loss": 4.0043,
      "step": 1295
    },
    {
      "epoch": 8.308186195826645,
      "grad_norm": 0.146484375,
      "learning_rate": 1.2950000000000001e-05,
      "loss": 3.9993,
      "step": 1296
    },
    {
      "epoch": 8.314606741573034,
      "grad_norm": 0.9140625,
      "learning_rate": 1.296e-05,
      "loss": 4.006,
      "step": 1297
    },
    {
      "epoch": 8.321027287319422,
      "grad_norm": 0.271484375,
      "learning_rate": 1.2970000000000001e-05,
      "loss": 4.0039,
      "step": 1298
    },
    {
      "epoch": 8.327447833065811,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.2980000000000001e-05,
      "loss": 4.0038,
      "step": 1299
    },
    {
      "epoch": 8.333868378812198,
      "grad_norm": 0.216796875,
      "learning_rate": 1.299e-05,
      "loss": 4.0064,
      "step": 1300
    }
  ],
  "logging_steps": 1,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 642,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.056432077193347e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}