{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.1068,
  "eval_steps": 500,
  "global_step": 8500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0001,
      "grad_norm": 1200.0023193359375,
      "learning_rate": 5.000000000000001e-07,
      "loss": 14.119,
      "step": 1
    },
    {
      "epoch": 0.0002,
      "grad_norm": 622.8350219726562,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 13.8646,
      "step": 2
    },
    {
      "epoch": 0.0003,
      "grad_norm": 1324.7479248046875,
      "learning_rate": 1.5e-06,
      "loss": 14.0779,
      "step": 3
    },
    {
      "epoch": 0.0004,
      "grad_norm": 2203.243408203125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 13.8076,
      "step": 4
    },
    {
      "epoch": 0.0005,
      "grad_norm": 710.7565307617188,
      "learning_rate": 2.5e-06,
      "loss": 19.1663,
      "step": 5
    },
    {
      "epoch": 0.0006,
      "grad_norm": 2382.641357421875,
      "learning_rate": 3e-06,
      "loss": 9.5343,
      "step": 6
    },
    {
      "epoch": 0.0007,
      "grad_norm": 621.6366577148438,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 14.9281,
      "step": 7
    },
    {
      "epoch": 0.0008,
      "grad_norm": 951.4312133789062,
      "learning_rate": 4.000000000000001e-06,
      "loss": 15.8061,
      "step": 8
    },
    {
      "epoch": 0.0009,
      "grad_norm": 658.730224609375,
      "learning_rate": 4.5e-06,
      "loss": 13.2833,
      "step": 9
    },
    {
      "epoch": 0.001,
      "grad_norm": 527.3458251953125,
      "learning_rate": 5e-06,
      "loss": 12.9623,
      "step": 10
    },
    {
      "epoch": 0.0011,
      "grad_norm": 519.3271484375,
      "learning_rate": 5.500000000000001e-06,
      "loss": 16.9994,
      "step": 11
    },
    {
      "epoch": 0.0012,
      "grad_norm": 672.4866333007812,
      "learning_rate": 6e-06,
      "loss": 13.1799,
      "step": 12
    },
    {
      "epoch": 0.0013,
      "grad_norm": 461.26812744140625,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 13.3611,
      "step": 13
    },
    {
      "epoch": 0.0014,
      "grad_norm": 530.612060546875,
      "learning_rate": 7.000000000000001e-06,
      "loss": 10.2872,
      "step": 14
    },
    {
      "epoch": 0.0015,
      "grad_norm": 11631.7060546875,
      "learning_rate": 7.5e-06,
      "loss": 12.5579,
      "step": 15
    },
    {
      "epoch": 0.0016,
      "grad_norm": 814.9800415039062,
      "learning_rate": 8.000000000000001e-06,
      "loss": 13.5092,
      "step": 16
    },
    {
      "epoch": 0.0017,
      "grad_norm": 591.564453125,
      "learning_rate": 8.500000000000002e-06,
      "loss": 10.4511,
      "step": 17
    },
    {
      "epoch": 0.0018,
      "grad_norm": 342.3058166503906,
      "learning_rate": 9e-06,
      "loss": 7.0515,
      "step": 18
    },
    {
      "epoch": 0.0019,
      "grad_norm": 396.5306091308594,
      "learning_rate": 9.5e-06,
      "loss": 9.8728,
      "step": 19
    },
    {
      "epoch": 0.002,
      "grad_norm": 1144.042236328125,
      "learning_rate": 1e-05,
      "loss": 10.6781,
      "step": 20
    },
    {
      "epoch": 0.0021,
      "grad_norm": 593.9048461914062,
      "learning_rate": 1.05e-05,
      "loss": 10.2028,
      "step": 21
    },
    {
      "epoch": 0.0022,
      "grad_norm": 353.4158630371094,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 8.2841,
      "step": 22
    },
    {
      "epoch": 0.0023,
      "grad_norm": 215.38671875,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 6.0089,
      "step": 23
    },
    {
      "epoch": 0.0024,
      "grad_norm": 476.2188415527344,
      "learning_rate": 1.2e-05,
      "loss": 7.9702,
      "step": 24
    },
    {
      "epoch": 0.0025,
      "grad_norm": 632.1690063476562,
      "learning_rate": 1.25e-05,
      "loss": 8.9473,
      "step": 25
    },
    {
      "epoch": 0.0026,
      "grad_norm": 390.53521728515625,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 7.7401,
      "step": 26
    },
    {
      "epoch": 0.0027,
      "grad_norm": 662.54833984375,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 8.1557,
      "step": 27
    },
    {
      "epoch": 0.0028,
      "grad_norm": 300.9270324707031,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 6.2397,
      "step": 28
    },
    {
      "epoch": 0.0029,
      "grad_norm": 452.31011962890625,
      "learning_rate": 1.45e-05,
      "loss": 8.6473,
      "step": 29
    },
    {
      "epoch": 0.003,
      "grad_norm": 255.45736694335938,
      "learning_rate": 1.5e-05,
      "loss": 6.2897,
      "step": 30
    },
    {
      "epoch": 0.0031,
      "grad_norm": 15456.908203125,
      "learning_rate": 1.55e-05,
      "loss": 6.8331,
      "step": 31
    },
    {
      "epoch": 0.0032,
      "grad_norm": 229.250732421875,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 7.0439,
      "step": 32
    },
    {
      "epoch": 0.0033,
      "grad_norm": 360.6927185058594,
      "learning_rate": 1.65e-05,
      "loss": 6.8334,
      "step": 33
    },
    {
      "epoch": 0.0034,
      "grad_norm": 203.5262908935547,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.8433,
      "step": 34
    },
    {
      "epoch": 0.0035,
      "grad_norm": 136.45298767089844,
      "learning_rate": 1.75e-05,
      "loss": 5.3895,
      "step": 35
    },
    {
      "epoch": 0.0036,
      "grad_norm": 155.35830688476562,
      "learning_rate": 1.8e-05,
      "loss": 6.5592,
      "step": 36
    },
    {
      "epoch": 0.0037,
      "grad_norm": 100.1277084350586,
      "learning_rate": 1.85e-05,
      "loss": 4.6016,
      "step": 37
    },
    {
      "epoch": 0.0038,
      "grad_norm": 362.4364929199219,
      "learning_rate": 1.9e-05,
      "loss": 5.8107,
      "step": 38
    },
    {
      "epoch": 0.0039,
      "grad_norm": 303.7940673828125,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 5.4166,
      "step": 39
    },
    {
      "epoch": 0.004,
      "grad_norm": 289.2499694824219,
      "learning_rate": 2e-05,
      "loss": 5.9465,
      "step": 40
    },
    {
      "epoch": 0.0041,
      "grad_norm": 188.70916748046875,
      "learning_rate": 2.05e-05,
      "loss": 4.7708,
      "step": 41
    },
    {
      "epoch": 0.0042,
      "grad_norm": 237.36611938476562,
      "learning_rate": 2.1e-05,
      "loss": 4.9795,
      "step": 42
    },
    {
      "epoch": 0.0043,
      "grad_norm": 530.0130615234375,
      "learning_rate": 2.15e-05,
      "loss": 4.6453,
      "step": 43
    },
    {
      "epoch": 0.0044,
      "grad_norm": 195.43161010742188,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 4.7322,
      "step": 44
    },
    {
      "epoch": 0.0045,
      "grad_norm": 223.3955841064453,
      "learning_rate": 2.25e-05,
      "loss": 4.8774,
      "step": 45
    },
    {
      "epoch": 0.0046,
      "grad_norm": 195.08828735351562,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 4.4239,
      "step": 46
    },
    {
      "epoch": 0.0047,
      "grad_norm": 201.23570251464844,
      "learning_rate": 2.35e-05,
      "loss": 5.3339,
      "step": 47
    },
    {
      "epoch": 0.0048,
      "grad_norm": 157.71995544433594,
      "learning_rate": 2.4e-05,
      "loss": 3.7473,
      "step": 48
    },
    {
      "epoch": 0.0049,
      "grad_norm": 68.49712371826172,
      "learning_rate": 2.45e-05,
      "loss": 3.5154,
      "step": 49
    },
    {
      "epoch": 0.005,
      "grad_norm": 137.03363037109375,
      "learning_rate": 2.5e-05,
      "loss": 3.6761,
      "step": 50
    },
    {
      "epoch": 0.0051,
      "grad_norm": 100.29153442382812,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 3.1779,
      "step": 51
    },
    {
      "epoch": 0.0052,
      "grad_norm": 164.61676025390625,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 3.6195,
      "step": 52
    },
    {
      "epoch": 0.0053,
      "grad_norm": 138.4666290283203,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 3.8459,
      "step": 53
    },
    {
      "epoch": 0.0054,
      "grad_norm": 430.7704772949219,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 3.5391,
      "step": 54
    },
    {
      "epoch": 0.0055,
      "grad_norm": 100.71952819824219,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 3.71,
      "step": 55
    },
    {
      "epoch": 0.0056,
      "grad_norm": 291.14154052734375,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 3.1142,
      "step": 56
    },
    {
      "epoch": 0.0057,
      "grad_norm": 89.535888671875,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 3.8865,
      "step": 57
    },
    {
      "epoch": 0.0058,
      "grad_norm": 49.601524353027344,
      "learning_rate": 2.9e-05,
      "loss": 2.595,
      "step": 58
    },
    {
      "epoch": 0.0059,
      "grad_norm": 83.9903793334961,
      "learning_rate": 2.95e-05,
      "loss": 2.9471,
      "step": 59
    },
    {
      "epoch": 0.006,
      "grad_norm": 139.70126342773438,
      "learning_rate": 3e-05,
      "loss": 3.74,
      "step": 60
    },
    {
      "epoch": 0.0061,
      "grad_norm": 86.51558685302734,
      "learning_rate": 3.05e-05,
      "loss": 3.801,
      "step": 61
    },
    {
      "epoch": 0.0062,
      "grad_norm": 76.30803680419922,
      "learning_rate": 3.1e-05,
      "loss": 2.6952,
      "step": 62
    },
    {
      "epoch": 0.0063,
      "grad_norm": 106.11151885986328,
      "learning_rate": 3.15e-05,
      "loss": 2.0604,
      "step": 63
    },
    {
      "epoch": 0.0064,
      "grad_norm": 55.28110122680664,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 2.7001,
      "step": 64
    },
    {
      "epoch": 0.0065,
      "grad_norm": 96.56853485107422,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 3.0387,
      "step": 65
    },
    {
      "epoch": 0.0066,
      "grad_norm": 54.827030181884766,
      "learning_rate": 3.3e-05,
      "loss": 1.9665,
      "step": 66
    },
    {
      "epoch": 0.0067,
      "grad_norm": 479.83013916015625,
      "learning_rate": 3.35e-05,
      "loss": 2.5752,
      "step": 67
    },
    {
      "epoch": 0.0068,
      "grad_norm": 59.37828063964844,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.9,
      "step": 68
    },
    {
      "epoch": 0.0069,
      "grad_norm": 185.78988647460938,
      "learning_rate": 3.45e-05,
      "loss": 2.8723,
      "step": 69
    },
    {
      "epoch": 0.007,
      "grad_norm": 43.122840881347656,
      "learning_rate": 3.5e-05,
      "loss": 2.9768,
      "step": 70
    },
    {
      "epoch": 0.0071,
      "grad_norm": 65.87982940673828,
      "learning_rate": 3.55e-05,
      "loss": 2.5826,
      "step": 71
    },
    {
      "epoch": 0.0072,
      "grad_norm": 46.65221405029297,
      "learning_rate": 3.6e-05,
      "loss": 2.0226,
      "step": 72
    },
    {
      "epoch": 0.0073,
      "grad_norm": 38.647438049316406,
      "learning_rate": 3.65e-05,
      "loss": 2.0231,
      "step": 73
    },
    {
      "epoch": 0.0074,
      "grad_norm": 51.9076042175293,
      "learning_rate": 3.7e-05,
      "loss": 1.7841,
      "step": 74
    },
    {
      "epoch": 0.0075,
      "grad_norm": 398.59429931640625,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.7743,
      "step": 75
    },
    {
      "epoch": 0.0076,
      "grad_norm": 54.934532165527344,
      "learning_rate": 3.8e-05,
      "loss": 1.9791,
      "step": 76
    },
    {
      "epoch": 0.0077,
      "grad_norm": 61.90827560424805,
      "learning_rate": 3.85e-05,
      "loss": 1.6385,
      "step": 77
    },
    {
      "epoch": 0.0078,
      "grad_norm": 86.92530059814453,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 2.23,
      "step": 78
    },
    {
      "epoch": 0.0079,
      "grad_norm": 32.42618942260742,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 1.6739,
      "step": 79
    },
    {
      "epoch": 0.008,
      "grad_norm": 43.01605224609375,
      "learning_rate": 4e-05,
      "loss": 2.2486,
      "step": 80
    },
    {
      "epoch": 0.0081,
      "grad_norm": 36.87302780151367,
      "learning_rate": 4.05e-05,
      "loss": 1.9447,
      "step": 81
    },
    {
      "epoch": 0.0082,
      "grad_norm": 108.80497741699219,
      "learning_rate": 4.1e-05,
      "loss": 1.1803,
      "step": 82
    },
    {
      "epoch": 0.0083,
      "grad_norm": 78.5074234008789,
      "learning_rate": 4.15e-05,
      "loss": 2.0255,
      "step": 83
    },
    {
      "epoch": 0.0084,
      "grad_norm": 126.53558349609375,
      "learning_rate": 4.2e-05,
      "loss": 1.3615,
      "step": 84
    },
    {
      "epoch": 0.0085,
      "grad_norm": 34.554813385009766,
      "learning_rate": 4.25e-05,
      "loss": 1.5286,
      "step": 85
    },
    {
      "epoch": 0.0086,
      "grad_norm": 58.276187896728516,
      "learning_rate": 4.3e-05,
      "loss": 1.9649,
      "step": 86
    },
    {
      "epoch": 0.0087,
      "grad_norm": 42.43537521362305,
      "learning_rate": 4.35e-05,
      "loss": 1.2026,
      "step": 87
    },
    {
      "epoch": 0.0088,
      "grad_norm": 56.498600006103516,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.489,
      "step": 88
    },
    {
      "epoch": 0.0089,
      "grad_norm": 188.74615478515625,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 1.4244,
      "step": 89
    },
    {
      "epoch": 0.009,
      "grad_norm": 31.258258819580078,
      "learning_rate": 4.5e-05,
      "loss": 1.6722,
      "step": 90
    },
    {
      "epoch": 0.0091,
      "grad_norm": 29.221803665161133,
      "learning_rate": 4.55e-05,
      "loss": 2.6922,
      "step": 91
    },
    {
      "epoch": 0.0092,
      "grad_norm": 37.086997985839844,
      "learning_rate": 4.600000000000001e-05,
      "loss": 2.122,
      "step": 92
    },
    {
      "epoch": 0.0093,
      "grad_norm": 67.60868072509766,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 2.0878,
      "step": 93
    },
    {
      "epoch": 0.0094,
      "grad_norm": 49.52315139770508,
      "learning_rate": 4.7e-05,
      "loss": 1.5706,
      "step": 94
    },
    {
      "epoch": 0.0095,
      "grad_norm": 58.37114715576172,
      "learning_rate": 4.75e-05,
      "loss": 2.0253,
      "step": 95
    },
    {
      "epoch": 0.0096,
      "grad_norm": 43.611080169677734,
      "learning_rate": 4.8e-05,
      "loss": 2.2332,
      "step": 96
    },
    {
      "epoch": 0.0097,
      "grad_norm": 53.272369384765625,
      "learning_rate": 4.85e-05,
      "loss": 1.47,
      "step": 97
    },
    {
      "epoch": 0.0098,
      "grad_norm": 31.558635711669922,
      "learning_rate": 4.9e-05,
      "loss": 0.9881,
      "step": 98
    },
    {
      "epoch": 0.0099,
      "grad_norm": 37.66632080078125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 1.4853,
      "step": 99
    },
    {
      "epoch": 0.01,
      "grad_norm": 44.606285095214844,
      "learning_rate": 5e-05,
      "loss": 1.8516,
      "step": 100
    },
    {
      "epoch": 0.0101,
      "grad_norm": 136.97853088378906,
      "learning_rate": 4.99949494949495e-05,
      "loss": 1.5311,
      "step": 101
    },
    {
      "epoch": 0.0102,
      "grad_norm": 22.425697326660156,
      "learning_rate": 4.9989898989898994e-05,
      "loss": 1.1485,
      "step": 102
    },
    {
      "epoch": 0.0103,
      "grad_norm": 28.628177642822266,
      "learning_rate": 4.998484848484849e-05,
      "loss": 1.2501,
      "step": 103
    },
    {
      "epoch": 0.0104,
      "grad_norm": 17.096609115600586,
      "learning_rate": 4.997979797979798e-05,
      "loss": 0.9571,
      "step": 104
    },
    {
      "epoch": 0.0105,
      "grad_norm": 79.11772918701172,
      "learning_rate": 4.997474747474748e-05,
      "loss": 2.6158,
      "step": 105
    },
    {
      "epoch": 0.0106,
      "grad_norm": 81.99260711669922,
      "learning_rate": 4.996969696969697e-05,
      "loss": 1.8117,
      "step": 106
    },
    {
      "epoch": 0.0107,
      "grad_norm": 53.0528450012207,
      "learning_rate": 4.996464646464647e-05,
      "loss": 2.0776,
      "step": 107
    },
    {
      "epoch": 0.0108,
      "grad_norm": 42.94725036621094,
      "learning_rate": 4.9959595959595964e-05,
      "loss": 1.285,
      "step": 108
    },
    {
      "epoch": 0.0109,
      "grad_norm": 29.207626342773438,
      "learning_rate": 4.995454545454546e-05,
      "loss": 2.049,
      "step": 109
    },
    {
      "epoch": 0.011,
      "grad_norm": 37.57393264770508,
      "learning_rate": 4.994949494949495e-05,
      "loss": 0.9042,
      "step": 110
    },
    {
      "epoch": 0.0111,
      "grad_norm": 32.10660171508789,
      "learning_rate": 4.994444444444445e-05,
      "loss": 1.4938,
      "step": 111
    },
    {
      "epoch": 0.0112,
      "grad_norm": 28.708343505859375,
      "learning_rate": 4.993939393939394e-05,
      "loss": 1.5244,
      "step": 112
    },
    {
      "epoch": 0.0113,
      "grad_norm": 40.67152786254883,
      "learning_rate": 4.993434343434344e-05,
      "loss": 2.7228,
      "step": 113
    },
    {
      "epoch": 0.0114,
      "grad_norm": 34.60795974731445,
      "learning_rate": 4.992929292929293e-05,
      "loss": 1.4832,
      "step": 114
    },
    {
      "epoch": 0.0115,
      "grad_norm": 18.434818267822266,
      "learning_rate": 4.992424242424243e-05,
      "loss": 1.3282,
      "step": 115
    },
    {
      "epoch": 0.0116,
      "grad_norm": 45.7252311706543,
      "learning_rate": 4.991919191919192e-05,
      "loss": 1.3122,
      "step": 116
    },
    {
      "epoch": 0.0117,
      "grad_norm": 22.563562393188477,
      "learning_rate": 4.991414141414142e-05,
      "loss": 1.2076,
      "step": 117
    },
    {
      "epoch": 0.0118,
      "grad_norm": 39.527339935302734,
      "learning_rate": 4.990909090909091e-05,
      "loss": 1.127,
      "step": 118
    },
    {
      "epoch": 0.0119,
      "grad_norm": 19.8275203704834,
      "learning_rate": 4.9904040404040406e-05,
      "loss": 0.7987,
      "step": 119
    },
    {
      "epoch": 0.012,
      "grad_norm": 17.92061996459961,
      "learning_rate": 4.98989898989899e-05,
      "loss": 1.1707,
      "step": 120
    },
    {
      "epoch": 0.0121,
      "grad_norm": 32.83548355102539,
      "learning_rate": 4.98939393939394e-05,
      "loss": 1.2299,
      "step": 121
    },
    {
      "epoch": 0.0122,
      "grad_norm": 26.857728958129883,
      "learning_rate": 4.9888888888888894e-05,
      "loss": 1.4244,
      "step": 122
    },
    {
      "epoch": 0.0123,
      "grad_norm": 23.411556243896484,
      "learning_rate": 4.988383838383839e-05,
      "loss": 1.2091,
      "step": 123
    },
    {
      "epoch": 0.0124,
      "grad_norm": 50.51976776123047,
      "learning_rate": 4.987878787878788e-05,
      "loss": 1.4219,
      "step": 124
    },
    {
      "epoch": 0.0125,
      "grad_norm": 36.07537841796875,
      "learning_rate": 4.9873737373737375e-05,
      "loss": 0.9768,
      "step": 125
    },
    {
      "epoch": 0.0126,
      "grad_norm": 30.927827835083008,
      "learning_rate": 4.986868686868687e-05,
      "loss": 1.1886,
      "step": 126
    },
    {
      "epoch": 0.0127,
      "grad_norm": 50.472957611083984,
      "learning_rate": 4.986363636363637e-05,
      "loss": 1.1986,
      "step": 127
    },
    {
      "epoch": 0.0128,
      "grad_norm": 19.249122619628906,
      "learning_rate": 4.985858585858586e-05,
      "loss": 0.8995,
      "step": 128
    },
    {
      "epoch": 0.0129,
      "grad_norm": 33.303375244140625,
      "learning_rate": 4.985353535353536e-05,
      "loss": 1.0028,
      "step": 129
    },
    {
      "epoch": 0.013,
      "grad_norm": 22.230316162109375,
      "learning_rate": 4.984848484848485e-05,
      "loss": 0.8178,
      "step": 130
    },
    {
      "epoch": 0.0131,
      "grad_norm": 45.19804000854492,
      "learning_rate": 4.9843434343434344e-05,
      "loss": 1.6245,
      "step": 131
    },
    {
      "epoch": 0.0132,
      "grad_norm": 37.1534538269043,
      "learning_rate": 4.983838383838384e-05,
      "loss": 1.044,
      "step": 132
    },
    {
      "epoch": 0.0133,
      "grad_norm": 20.091835021972656,
      "learning_rate": 4.9833333333333336e-05,
      "loss": 1.0191,
      "step": 133
    },
    {
      "epoch": 0.0134,
      "grad_norm": 18.180248260498047,
      "learning_rate": 4.982828282828283e-05,
      "loss": 0.5843,
      "step": 134
    },
    {
      "epoch": 0.0135,
      "grad_norm": 18.20668601989746,
      "learning_rate": 4.982323232323233e-05,
      "loss": 0.8532,
      "step": 135
    },
    {
      "epoch": 0.0136,
      "grad_norm": 12.808496475219727,
      "learning_rate": 4.981818181818182e-05,
      "loss": 0.6584,
      "step": 136
    },
    {
      "epoch": 0.0137,
      "grad_norm": 26.31954574584961,
      "learning_rate": 4.981313131313132e-05,
      "loss": 1.0545,
      "step": 137
    },
    {
      "epoch": 0.0138,
      "grad_norm": 35.347618103027344,
      "learning_rate": 4.980808080808081e-05,
      "loss": 0.7785,
      "step": 138
    },
    {
      "epoch": 0.0139,
      "grad_norm": 46.19343185424805,
      "learning_rate": 4.9803030303030305e-05,
      "loss": 1.274,
      "step": 139
    },
    {
      "epoch": 0.014,
      "grad_norm": 21.9566650390625,
      "learning_rate": 4.97979797979798e-05,
      "loss": 1.165,
      "step": 140
    },
    {
      "epoch": 0.0141,
      "grad_norm": 39.47710037231445,
      "learning_rate": 4.97929292929293e-05,
      "loss": 1.3045,
      "step": 141
    },
    {
      "epoch": 0.0142,
      "grad_norm": 24.425678253173828,
      "learning_rate": 4.9787878787878787e-05,
      "loss": 1.0872,
      "step": 142
    },
    {
      "epoch": 0.0143,
      "grad_norm": 22.48780059814453,
      "learning_rate": 4.978282828282829e-05,
      "loss": 1.1488,
      "step": 143
    },
    {
      "epoch": 0.0144,
      "grad_norm": 27.3831787109375,
      "learning_rate": 4.977777777777778e-05,
      "loss": 1.1538,
      "step": 144
    },
    {
      "epoch": 0.0145,
      "grad_norm": 35.39277648925781,
      "learning_rate": 4.9772727272727275e-05,
      "loss": 1.9485,
      "step": 145
    },
    {
      "epoch": 0.0146,
      "grad_norm": 19.116912841796875,
      "learning_rate": 4.976767676767677e-05,
      "loss": 1.1243,
      "step": 146
    },
    {
      "epoch": 0.0147,
      "grad_norm": 42.09323501586914,
      "learning_rate": 4.9762626262626267e-05,
      "loss": 1.2294,
      "step": 147
    },
    {
      "epoch": 0.0148,
      "grad_norm": 41.28529739379883,
      "learning_rate": 4.9757575757575756e-05,
      "loss": 1.4191,
      "step": 148
    },
    {
      "epoch": 0.0149,
      "grad_norm": 31.60830307006836,
      "learning_rate": 4.975252525252526e-05,
      "loss": 0.7183,
      "step": 149
    },
    {
      "epoch": 0.015,
      "grad_norm": 18.453824996948242,
      "learning_rate": 4.974747474747475e-05,
      "loss": 0.5816,
      "step": 150
    },
    {
      "epoch": 0.0151,
      "grad_norm": 32.060890197753906,
      "learning_rate": 4.9742424242424244e-05,
      "loss": 1.4178,
      "step": 151
    },
    {
      "epoch": 0.0152,
      "grad_norm": 17.44024085998535,
      "learning_rate": 4.973737373737374e-05,
      "loss": 0.7326,
      "step": 152
    },
    {
      "epoch": 0.0153,
      "grad_norm": 20.004423141479492,
      "learning_rate": 4.9732323232323236e-05,
      "loss": 1.0714,
      "step": 153
    },
    {
      "epoch": 0.0154,
      "grad_norm": 24.47506332397461,
      "learning_rate": 4.9727272727272725e-05,
      "loss": 1.7158,
      "step": 154
    },
    {
      "epoch": 0.0155,
      "grad_norm": 45.26461410522461,
      "learning_rate": 4.972222222222223e-05,
      "loss": 2.1165,
      "step": 155
    },
    {
      "epoch": 0.0156,
      "grad_norm": 21.99950408935547,
      "learning_rate": 4.971717171717172e-05,
      "loss": 1.1637,
      "step": 156
    },
    {
      "epoch": 0.0157,
      "grad_norm": 13.293519973754883,
      "learning_rate": 4.971212121212121e-05,
      "loss": 1.8121,
      "step": 157
    },
    {
      "epoch": 0.0158,
      "grad_norm": 27.29784393310547,
      "learning_rate": 4.970707070707071e-05,
      "loss": 0.9857,
      "step": 158
    },
    {
      "epoch": 0.0159,
      "grad_norm": 28.754610061645508,
      "learning_rate": 4.9702020202020205e-05,
      "loss": 1.469,
      "step": 159
    },
    {
      "epoch": 0.016,
      "grad_norm": 32.193321228027344,
      "learning_rate": 4.9696969696969694e-05,
      "loss": 1.422,
      "step": 160
    },
    {
      "epoch": 0.0161,
      "grad_norm": 23.36785316467285,
      "learning_rate": 4.96919191919192e-05,
      "loss": 0.7131,
      "step": 161
    },
    {
      "epoch": 0.0162,
      "grad_norm": 27.815889358520508,
      "learning_rate": 4.9686868686868686e-05,
      "loss": 0.6237,
      "step": 162
    },
    {
      "epoch": 0.0163,
      "grad_norm": 60.487388610839844,
      "learning_rate": 4.968181818181818e-05,
      "loss": 1.9885,
      "step": 163
    },
    {
      "epoch": 0.0164,
      "grad_norm": 26.063804626464844,
      "learning_rate": 4.967676767676768e-05,
      "loss": 0.8997,
      "step": 164
    },
    {
      "epoch": 0.0165,
      "grad_norm": 22.750244140625,
      "learning_rate": 4.9671717171717174e-05,
      "loss": 0.8805,
      "step": 165
    },
    {
      "epoch": 0.0166,
      "grad_norm": 73.63886260986328,
      "learning_rate": 4.966666666666667e-05,
      "loss": 1.0041,
      "step": 166
    },
    {
      "epoch": 0.0167,
      "grad_norm": 18.313190460205078,
      "learning_rate": 4.9661616161616166e-05,
      "loss": 0.9338,
      "step": 167
    },
    {
      "epoch": 0.0168,
      "grad_norm": 52.938560485839844,
      "learning_rate": 4.9656565656565655e-05,
      "loss": 1.4045,
      "step": 168
    },
    {
      "epoch": 0.0169,
      "grad_norm": 16.498266220092773,
      "learning_rate": 4.965151515151515e-05,
      "loss": 0.7122,
      "step": 169
    },
    {
      "epoch": 0.017,
      "grad_norm": 249.92642211914062,
      "learning_rate": 4.964646464646465e-05,
      "loss": 0.8493,
      "step": 170
    },
    {
      "epoch": 0.0171,
      "grad_norm": 19.559053421020508,
      "learning_rate": 4.964141414141414e-05,
      "loss": 1.4409,
      "step": 171
    },
    {
      "epoch": 0.0172,
      "grad_norm": 61.43571472167969,
      "learning_rate": 4.963636363636364e-05,
      "loss": 1.5982,
      "step": 172
    },
    {
      "epoch": 0.0173,
      "grad_norm": 15.27564525604248,
      "learning_rate": 4.9631313131313135e-05,
      "loss": 0.8294,
      "step": 173
    },
    {
      "epoch": 0.0174,
      "grad_norm": 19.138057708740234,
      "learning_rate": 4.9626262626262624e-05,
      "loss": 1.1057,
      "step": 174
    },
    {
      "epoch": 0.0175,
      "grad_norm": 20.625629425048828,
      "learning_rate": 4.962121212121213e-05,
      "loss": 0.8848,
      "step": 175
    },
    {
      "epoch": 0.0176,
      "grad_norm": 17.37908935546875,
      "learning_rate": 4.9616161616161616e-05,
      "loss": 1.1142,
      "step": 176
    },
    {
      "epoch": 0.0177,
      "grad_norm": 9.398472785949707,
      "learning_rate": 4.961111111111111e-05,
      "loss": 0.5209,
      "step": 177
    },
    {
      "epoch": 0.0178,
      "grad_norm": 16.033727645874023,
      "learning_rate": 4.960606060606061e-05,
      "loss": 0.7183,
      "step": 178
    },
    {
      "epoch": 0.0179,
      "grad_norm": 14.563871383666992,
      "learning_rate": 4.9601010101010104e-05,
      "loss": 1.4042,
      "step": 179
    },
    {
      "epoch": 0.018,
      "grad_norm": 58.65496063232422,
      "learning_rate": 4.9595959595959594e-05,
      "loss": 1.0551,
      "step": 180
    },
    {
      "epoch": 0.0181,
      "grad_norm": 64.78812408447266,
      "learning_rate": 4.9590909090909096e-05,
      "loss": 1.6846,
      "step": 181
    },
    {
      "epoch": 0.0182,
      "grad_norm": 19.850919723510742,
      "learning_rate": 4.9585858585858586e-05,
      "loss": 1.0862,
      "step": 182
    },
    {
      "epoch": 0.0183,
      "grad_norm": 15.983012199401855,
      "learning_rate": 4.958080808080808e-05,
      "loss": 0.6962,
      "step": 183
    },
    {
      "epoch": 0.0184,
      "grad_norm": 139.4163360595703,
      "learning_rate": 4.957575757575758e-05,
      "loss": 0.8121,
      "step": 184
    },
    {
      "epoch": 0.0185,
      "grad_norm": 29.277393341064453,
      "learning_rate": 4.9570707070707073e-05,
      "loss": 1.9143,
      "step": 185
    },
    {
      "epoch": 0.0186,
      "grad_norm": 46.36794662475586,
      "learning_rate": 4.956565656565657e-05,
      "loss": 1.6727,
      "step": 186
    },
    {
      "epoch": 0.0187,
      "grad_norm": 17.143814086914062,
      "learning_rate": 4.9560606060606065e-05,
      "loss": 1.2626,
      "step": 187
    },
    {
      "epoch": 0.0188,
      "grad_norm": 20.64394760131836,
      "learning_rate": 4.955555555555556e-05,
      "loss": 0.5357,
      "step": 188
    },
    {
      "epoch": 0.0189,
      "grad_norm": 23.12631607055664,
      "learning_rate": 4.955050505050505e-05,
      "loss": 1.1549,
      "step": 189
    },
    {
      "epoch": 0.019,
      "grad_norm": 21.69987678527832,
      "learning_rate": 4.9545454545454553e-05,
      "loss": 1.1196,
      "step": 190
    },
    {
      "epoch": 0.0191,
      "grad_norm": 15.465149879455566,
      "learning_rate": 4.954040404040404e-05,
      "loss": 0.8442,
      "step": 191
    },
    {
      "epoch": 0.0192,
      "grad_norm": 19.40150260925293,
      "learning_rate": 4.953535353535354e-05,
      "loss": 1.4818,
      "step": 192
    },
    {
      "epoch": 0.0193,
      "grad_norm": 45.28868865966797,
      "learning_rate": 4.9530303030303035e-05,
      "loss": 0.6941,
      "step": 193
    },
    {
      "epoch": 0.0194,
      "grad_norm": 31.20271873474121,
      "learning_rate": 4.952525252525253e-05,
      "loss": 0.9061,
      "step": 194
    },
    {
      "epoch": 0.0195,
      "grad_norm": 19.401716232299805,
      "learning_rate": 4.952020202020202e-05,
      "loss": 0.8751,
      "step": 195
    },
    {
      "epoch": 0.0196,
      "grad_norm": 21.05107879638672,
      "learning_rate": 4.951515151515152e-05,
      "loss": 0.8309,
      "step": 196
    },
    {
      "epoch": 0.0197,
      "grad_norm": 19.76274871826172,
      "learning_rate": 4.951010101010101e-05,
      "loss": 1.5215,
      "step": 197
    },
    {
      "epoch": 0.0198,
      "grad_norm": 60.32820129394531,
      "learning_rate": 4.950505050505051e-05,
      "loss": 0.3298,
      "step": 198
    },
    {
      "epoch": 0.0199,
      "grad_norm": 40.70603561401367,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 1.0566,
      "step": 199
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.866750717163086,
      "learning_rate": 4.94949494949495e-05,
      "loss": 0.4367,
      "step": 200
    },
    {
      "epoch": 0.0201,
      "grad_norm": 17.006671905517578,
      "learning_rate": 4.948989898989899e-05,
      "loss": 0.6704,
      "step": 201
    },
    {
      "epoch": 0.0202,
      "grad_norm": 16.066669464111328,
      "learning_rate": 4.948484848484849e-05,
      "loss": 0.6052,
      "step": 202
    },
    {
      "epoch": 0.0203,
      "grad_norm": 26.57935333251953,
      "learning_rate": 4.947979797979798e-05,
      "loss": 1.5021,
      "step": 203
    },
    {
      "epoch": 0.0204,
      "grad_norm": 27.08062744140625,
      "learning_rate": 4.947474747474748e-05,
      "loss": 1.0843,
      "step": 204
    },
    {
      "epoch": 0.0205,
      "grad_norm": 10.89648723602295,
      "learning_rate": 4.946969696969697e-05,
      "loss": 0.7753,
      "step": 205
    },
    {
      "epoch": 0.0206,
      "grad_norm": 30.760034561157227,
      "learning_rate": 4.946464646464647e-05,
      "loss": 1.321,
      "step": 206
    },
    {
      "epoch": 0.0207,
      "grad_norm": 22.139986038208008,
      "learning_rate": 4.945959595959596e-05,
      "loss": 0.5882,
      "step": 207
    },
    {
      "epoch": 0.0208,
      "grad_norm": 35.458824157714844,
      "learning_rate": 4.945454545454546e-05,
      "loss": 0.9911,
      "step": 208
    },
    {
      "epoch": 0.0209,
      "grad_norm": 11.132645606994629,
      "learning_rate": 4.944949494949495e-05,
      "loss": 0.7302,
      "step": 209
    },
    {
      "epoch": 0.021,
      "grad_norm": 8.548279762268066,
      "learning_rate": 4.9444444444444446e-05,
      "loss": 0.3699,
      "step": 210
    },
    {
      "epoch": 0.0211,
      "grad_norm": 14.16301441192627,
      "learning_rate": 4.943939393939394e-05,
      "loss": 0.8777,
      "step": 211
    },
    {
      "epoch": 0.0212,
      "grad_norm": 10.160534858703613,
      "learning_rate": 4.943434343434344e-05,
      "loss": 0.5285,
      "step": 212
    },
    {
      "epoch": 0.0213,
      "grad_norm": 12.006013870239258,
      "learning_rate": 4.9429292929292934e-05,
      "loss": 0.541,
      "step": 213
    },
    {
      "epoch": 0.0214,
      "grad_norm": 16.927885055541992,
      "learning_rate": 4.942424242424243e-05,
      "loss": 0.8746,
      "step": 214
    },
    {
      "epoch": 0.0215,
      "grad_norm": 20.354543685913086,
      "learning_rate": 4.941919191919192e-05,
      "loss": 0.6461,
      "step": 215
    },
    {
      "epoch": 0.0216,
      "grad_norm": 27.035537719726562,
      "learning_rate": 4.9414141414141415e-05,
      "loss": 1.2,
      "step": 216
    },
    {
      "epoch": 0.0217,
      "grad_norm": 12.59115982055664,
      "learning_rate": 4.940909090909091e-05,
      "loss": 0.7719,
      "step": 217
    },
    {
      "epoch": 0.0218,
      "grad_norm": 18.47043800354004,
      "learning_rate": 4.940404040404041e-05,
      "loss": 0.9711,
      "step": 218
    },
    {
      "epoch": 0.0219,
      "grad_norm": 20.3468017578125,
      "learning_rate": 4.93989898989899e-05,
      "loss": 1.0957,
      "step": 219
    },
    {
      "epoch": 0.022,
      "grad_norm": 10.201119422912598,
      "learning_rate": 4.93939393939394e-05,
      "loss": 0.5997,
      "step": 220
    },
    {
      "epoch": 0.0221,
      "grad_norm": 20.670448303222656,
      "learning_rate": 4.938888888888889e-05,
      "loss": 2.3696,
      "step": 221
    },
    {
      "epoch": 0.0222,
      "grad_norm": 9.15916633605957,
      "learning_rate": 4.9383838383838384e-05,
      "loss": 0.6258,
      "step": 222
    },
    {
      "epoch": 0.0223,
      "grad_norm": 16.741363525390625,
      "learning_rate": 4.937878787878788e-05,
      "loss": 0.7057,
      "step": 223
    },
    {
      "epoch": 0.0224,
      "grad_norm": 18.821931838989258,
      "learning_rate": 4.9373737373737376e-05,
      "loss": 1.2004,
      "step": 224
    },
    {
      "epoch": 0.0225,
      "grad_norm": 14.101298332214355,
      "learning_rate": 4.936868686868687e-05,
      "loss": 0.9066,
      "step": 225
    },
    {
      "epoch": 0.0226,
      "grad_norm": 16.5988712310791,
      "learning_rate": 4.936363636363637e-05,
      "loss": 1.3271,
      "step": 226
    },
    {
      "epoch": 0.0227,
      "grad_norm": 15.77312183380127,
      "learning_rate": 4.935858585858586e-05,
      "loss": 0.9334,
      "step": 227
    },
    {
      "epoch": 0.0228,
      "grad_norm": 31.8894100189209,
      "learning_rate": 4.935353535353536e-05,
      "loss": 1.4944,
      "step": 228
    },
    {
      "epoch": 0.0229,
      "grad_norm": 20.460468292236328,
      "learning_rate": 4.934848484848485e-05,
      "loss": 1.9892,
      "step": 229
    },
    {
      "epoch": 0.023,
      "grad_norm": 13.832818031311035,
      "learning_rate": 4.9343434343434346e-05,
      "loss": 0.8345,
      "step": 230
    },
    {
      "epoch": 0.0231,
      "grad_norm": 14.441542625427246,
      "learning_rate": 4.933838383838384e-05,
      "loss": 0.5648,
      "step": 231
    },
    {
      "epoch": 0.0232,
      "grad_norm": 18.907773971557617,
      "learning_rate": 4.933333333333334e-05,
      "loss": 0.5044,
      "step": 232
    },
    {
      "epoch": 0.0233,
      "grad_norm": 16.380857467651367,
      "learning_rate": 4.932828282828283e-05,
      "loss": 0.4363,
      "step": 233
    },
    {
      "epoch": 0.0234,
      "grad_norm": 12.36284065246582,
      "learning_rate": 4.932323232323233e-05,
      "loss": 0.8455,
      "step": 234
    },
    {
      "epoch": 0.0235,
      "grad_norm": 11.306267738342285,
      "learning_rate": 4.931818181818182e-05,
      "loss": 0.532,
      "step": 235
    },
    {
      "epoch": 0.0236,
      "grad_norm": 16.16178321838379,
      "learning_rate": 4.9313131313131315e-05,
      "loss": 0.7987,
      "step": 236
    },
    {
      "epoch": 0.0237,
      "grad_norm": 19.710914611816406,
      "learning_rate": 4.930808080808081e-05,
      "loss": 0.5562,
      "step": 237
    },
    {
      "epoch": 0.0238,
      "grad_norm": 29.080135345458984,
      "learning_rate": 4.930303030303031e-05,
      "loss": 1.2729,
      "step": 238
    },
    {
      "epoch": 0.0239,
      "grad_norm": 11.979623794555664,
      "learning_rate": 4.9297979797979796e-05,
      "loss": 1.0293,
      "step": 239
    },
    {
      "epoch": 0.024,
      "grad_norm": 15.626187324523926,
      "learning_rate": 4.92929292929293e-05,
      "loss": 0.6535,
      "step": 240
    },
    {
      "epoch": 0.0241,
      "grad_norm": 44.667816162109375,
      "learning_rate": 4.928787878787879e-05,
      "loss": 1.7181,
      "step": 241
    },
    {
      "epoch": 0.0242,
      "grad_norm": 19.63367462158203,
      "learning_rate": 4.9282828282828284e-05,
      "loss": 1.1674,
      "step": 242
    },
    {
      "epoch": 0.0243,
      "grad_norm": 20.7071590423584,
      "learning_rate": 4.927777777777778e-05,
      "loss": 1.3666,
      "step": 243
    },
    {
      "epoch": 0.0244,
      "grad_norm": 7.65720272064209,
      "learning_rate": 4.9272727272727276e-05,
      "loss": 0.5129,
      "step": 244
    },
    {
      "epoch": 0.0245,
      "grad_norm": 6.9388813972473145,
      "learning_rate": 4.9267676767676765e-05,
      "loss": 0.9246,
      "step": 245
    },
    {
      "epoch": 0.0246,
      "grad_norm": 15.36917781829834,
      "learning_rate": 4.926262626262627e-05,
      "loss": 1.1791,
      "step": 246
    },
    {
      "epoch": 0.0247,
      "grad_norm": 49.07154846191406,
      "learning_rate": 4.925757575757576e-05,
      "loss": 0.7816,
      "step": 247
    },
    {
      "epoch": 0.0248,
      "grad_norm": 9.148021697998047,
      "learning_rate": 4.925252525252525e-05,
      "loss": 0.5278,
      "step": 248
    },
    {
      "epoch": 0.0249,
      "grad_norm": 9.731310844421387,
      "learning_rate": 4.924747474747475e-05,
      "loss": 0.6986,
      "step": 249
    },
    {
      "epoch": 0.025,
      "grad_norm": 28.28838348388672,
      "learning_rate": 4.9242424242424245e-05,
      "loss": 1.4686,
      "step": 250
    },
    {
      "epoch": 0.0251,
      "grad_norm": 14.345922470092773,
      "learning_rate": 4.9237373737373734e-05,
      "loss": 1.1523,
      "step": 251
    },
    {
      "epoch": 0.0252,
      "grad_norm": 13.286799430847168,
      "learning_rate": 4.923232323232324e-05,
      "loss": 0.9124,
      "step": 252
    },
    {
      "epoch": 0.0253,
      "grad_norm": 17.285356521606445,
      "learning_rate": 4.9227272727272726e-05,
      "loss": 1.0465,
      "step": 253
    },
    {
      "epoch": 0.0254,
      "grad_norm": 10.533773422241211,
      "learning_rate": 4.922222222222222e-05,
      "loss": 0.6104,
      "step": 254
    },
    {
      "epoch": 0.0255,
      "grad_norm": 9.442602157592773,
      "learning_rate": 4.921717171717172e-05,
      "loss": 0.9623,
      "step": 255
    },
    {
      "epoch": 0.0256,
      "grad_norm": 10.018538475036621,
      "learning_rate": 4.9212121212121214e-05,
      "loss": 0.8409,
      "step": 256
    },
    {
      "epoch": 0.0257,
      "grad_norm": 21.055593490600586,
      "learning_rate": 4.920707070707071e-05,
      "loss": 0.875,
      "step": 257
    },
    {
      "epoch": 0.0258,
      "grad_norm": 203.1394805908203,
      "learning_rate": 4.9202020202020206e-05,
      "loss": 1.3795,
      "step": 258
    },
    {
      "epoch": 0.0259,
      "grad_norm": 7.641512870788574,
      "learning_rate": 4.9196969696969695e-05,
      "loss": 0.2561,
      "step": 259
    },
    {
      "epoch": 0.026,
      "grad_norm": 14.64362907409668,
      "learning_rate": 4.919191919191919e-05,
      "loss": 0.5102,
      "step": 260
    },
    {
      "epoch": 0.0261,
      "grad_norm": 44.99274444580078,
      "learning_rate": 4.918686868686869e-05,
      "loss": 0.9007,
      "step": 261
    },
    {
      "epoch": 0.0262,
      "grad_norm": 22.682571411132812,
      "learning_rate": 4.9181818181818183e-05,
      "loss": 0.714,
      "step": 262
    },
    {
      "epoch": 0.0263,
      "grad_norm": 12.68237590789795,
      "learning_rate": 4.917676767676768e-05,
      "loss": 1.2988,
      "step": 263
    },
    {
      "epoch": 0.0264,
      "grad_norm": 23.986665725708008,
      "learning_rate": 4.9171717171717175e-05,
      "loss": 0.5818,
      "step": 264
    },
    {
      "epoch": 0.0265,
      "grad_norm": 12.358917236328125,
      "learning_rate": 4.9166666666666665e-05,
      "loss": 1.1767,
      "step": 265
    },
    {
      "epoch": 0.0266,
      "grad_norm": 16.69634437561035,
      "learning_rate": 4.916161616161617e-05,
      "loss": 0.6266,
      "step": 266
    },
    {
      "epoch": 0.0267,
      "grad_norm": 13.057369232177734,
      "learning_rate": 4.9156565656565657e-05,
      "loss": 0.9181,
      "step": 267
    },
    {
      "epoch": 0.0268,
      "grad_norm": 8.068781852722168,
      "learning_rate": 4.915151515151515e-05,
      "loss": 0.5648,
      "step": 268
    },
    {
      "epoch": 0.0269,
      "grad_norm": 11.44855785369873,
      "learning_rate": 4.914646464646465e-05,
      "loss": 1.4061,
      "step": 269
    },
    {
      "epoch": 0.027,
      "grad_norm": 21.385562896728516,
      "learning_rate": 4.9141414141414145e-05,
      "loss": 0.9678,
      "step": 270
    },
    {
      "epoch": 0.0271,
      "grad_norm": 8.333712577819824,
      "learning_rate": 4.913636363636364e-05,
      "loss": 0.3391,
      "step": 271
    },
    {
      "epoch": 0.0272,
      "grad_norm": 17.07071876525879,
      "learning_rate": 4.9131313131313137e-05,
      "loss": 0.6171,
      "step": 272
    },
    {
      "epoch": 0.0273,
      "grad_norm": 9.711565971374512,
      "learning_rate": 4.912626262626263e-05,
      "loss": 0.3702,
      "step": 273
    },
    {
      "epoch": 0.0274,
      "grad_norm": 8.19430160522461,
      "learning_rate": 4.912121212121212e-05,
      "loss": 0.8769,
      "step": 274
    },
    {
      "epoch": 0.0275,
      "grad_norm": 12.115705490112305,
      "learning_rate": 4.9116161616161625e-05,
      "loss": 1.3439,
      "step": 275
    },
    {
      "epoch": 0.0276,
      "grad_norm": 15.138751983642578,
      "learning_rate": 4.9111111111111114e-05,
      "loss": 0.5337,
      "step": 276
    },
    {
      "epoch": 0.0277,
      "grad_norm": 18.212078094482422,
      "learning_rate": 4.910606060606061e-05,
      "loss": 1.2825,
      "step": 277
    },
    {
      "epoch": 0.0278,
      "grad_norm": 11.690363883972168,
      "learning_rate": 4.9101010101010106e-05,
      "loss": 0.8469,
      "step": 278
    },
    {
      "epoch": 0.0279,
      "grad_norm": 11.738852500915527,
      "learning_rate": 4.90959595959596e-05,
      "loss": 1.4203,
      "step": 279
    },
    {
      "epoch": 0.028,
      "grad_norm": 11.041299819946289,
      "learning_rate": 4.909090909090909e-05,
      "loss": 0.8039,
      "step": 280
    },
    {
      "epoch": 0.0281,
      "grad_norm": 18.324243545532227,
      "learning_rate": 4.9085858585858594e-05,
      "loss": 0.8116,
      "step": 281
    },
    {
      "epoch": 0.0282,
      "grad_norm": 23.416894912719727,
      "learning_rate": 4.908080808080808e-05,
      "loss": 0.6987,
      "step": 282
    },
    {
      "epoch": 0.0283,
      "grad_norm": 11.011762619018555,
      "learning_rate": 4.907575757575758e-05,
      "loss": 0.8327,
      "step": 283
    },
    {
      "epoch": 0.0284,
      "grad_norm": 12.292765617370605,
      "learning_rate": 4.9070707070707075e-05,
      "loss": 0.9062,
      "step": 284
    },
    {
      "epoch": 0.0285,
      "grad_norm": 11.701115608215332,
      "learning_rate": 4.906565656565657e-05,
      "loss": 0.9207,
      "step": 285
    },
    {
      "epoch": 0.0286,
      "grad_norm": 30.01003646850586,
      "learning_rate": 4.906060606060606e-05,
      "loss": 1.1441,
      "step": 286
    },
    {
      "epoch": 0.0287,
      "grad_norm": 20.930044174194336,
      "learning_rate": 4.905555555555556e-05,
      "loss": 0.7437,
      "step": 287
    },
    {
      "epoch": 0.0288,
      "grad_norm": 17.954282760620117,
      "learning_rate": 4.905050505050505e-05,
      "loss": 1.5891,
      "step": 288
    },
    {
      "epoch": 0.0289,
      "grad_norm": 13.728120803833008,
      "learning_rate": 4.904545454545455e-05,
      "loss": 0.8892,
      "step": 289
    },
    {
      "epoch": 0.029,
      "grad_norm": 29.378421783447266,
      "learning_rate": 4.9040404040404044e-05,
      "loss": 0.7753,
      "step": 290
    },
    {
      "epoch": 0.0291,
      "grad_norm": 13.874923706054688,
      "learning_rate": 4.903535353535354e-05,
      "loss": 1.8361,
      "step": 291
    },
    {
      "epoch": 0.0292,
      "grad_norm": 10.979392051696777,
      "learning_rate": 4.903030303030303e-05,
      "loss": 0.9742,
      "step": 292
    },
    {
      "epoch": 0.0293,
      "grad_norm": 11.921304702758789,
      "learning_rate": 4.902525252525253e-05,
      "loss": 1.5685,
      "step": 293
    },
    {
      "epoch": 0.0294,
      "grad_norm": 13.257525444030762,
      "learning_rate": 4.902020202020202e-05,
      "loss": 1.3165,
      "step": 294
    },
    {
      "epoch": 0.0295,
      "grad_norm": 13.973977088928223,
      "learning_rate": 4.901515151515152e-05,
      "loss": 0.7221,
      "step": 295
    },
    {
      "epoch": 0.0296,
      "grad_norm": 12.32652473449707,
      "learning_rate": 4.901010101010101e-05,
      "loss": 0.9399,
      "step": 296
    },
    {
      "epoch": 0.0297,
      "grad_norm": 11.097368240356445,
      "learning_rate": 4.900505050505051e-05,
      "loss": 0.6389,
      "step": 297
    },
    {
      "epoch": 0.0298,
      "grad_norm": 12.060166358947754,
      "learning_rate": 4.9e-05,
      "loss": 0.6109,
      "step": 298
    },
    {
      "epoch": 0.0299,
      "grad_norm": 15.481499671936035,
      "learning_rate": 4.89949494949495e-05,
      "loss": 1.6851,
      "step": 299
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.807249069213867,
      "learning_rate": 4.898989898989899e-05,
      "loss": 0.5515,
      "step": 300
    },
    {
      "epoch": 0.0301,
      "grad_norm": 9.842585563659668,
      "learning_rate": 4.8984848484848486e-05,
      "loss": 0.7478,
      "step": 301
    },
    {
      "epoch": 0.0302,
      "grad_norm": 13.695808410644531,
      "learning_rate": 4.897979797979798e-05,
      "loss": 1.0963,
      "step": 302
    },
    {
      "epoch": 0.0303,
      "grad_norm": 14.065753936767578,
      "learning_rate": 4.897474747474748e-05,
      "loss": 1.4774,
      "step": 303
    },
    {
      "epoch": 0.0304,
      "grad_norm": 92.40540313720703,
      "learning_rate": 4.896969696969697e-05,
      "loss": 1.2215,
      "step": 304
    },
    {
      "epoch": 0.0305,
      "grad_norm": 8.312875747680664,
      "learning_rate": 4.896464646464647e-05,
      "loss": 0.6106,
      "step": 305
    },
    {
      "epoch": 0.0306,
      "grad_norm": 23.882278442382812,
      "learning_rate": 4.895959595959596e-05,
      "loss": 1.2691,
      "step": 306
    },
    {
      "epoch": 0.0307,
      "grad_norm": 12.021655082702637,
      "learning_rate": 4.8954545454545456e-05,
      "loss": 0.7145,
      "step": 307
    },
    {
      "epoch": 0.0308,
      "grad_norm": 6.766305923461914,
      "learning_rate": 4.894949494949495e-05,
      "loss": 1.0836,
      "step": 308
    },
    {
      "epoch": 0.0309,
      "grad_norm": 13.398940086364746,
      "learning_rate": 4.894444444444445e-05,
      "loss": 0.5714,
      "step": 309
    },
    {
      "epoch": 0.031,
      "grad_norm": 13.5343017578125,
      "learning_rate": 4.8939393939393944e-05,
      "loss": 1.1556,
      "step": 310
    },
    {
      "epoch": 0.0311,
      "grad_norm": 20.2145938873291,
      "learning_rate": 4.893434343434344e-05,
      "loss": 0.9826,
      "step": 311
    },
    {
      "epoch": 0.0312,
      "grad_norm": 8.160256385803223,
      "learning_rate": 4.892929292929293e-05,
      "loss": 1.0281,
      "step": 312
    },
    {
      "epoch": 0.0313,
      "grad_norm": 13.952061653137207,
      "learning_rate": 4.8924242424242425e-05,
      "loss": 1.1316,
      "step": 313
    },
    {
      "epoch": 0.0314,
      "grad_norm": 6.801046371459961,
      "learning_rate": 4.891919191919192e-05,
      "loss": 0.6036,
      "step": 314
    },
    {
      "epoch": 0.0315,
      "grad_norm": 16.27368927001953,
      "learning_rate": 4.891414141414142e-05,
      "loss": 1.3686,
      "step": 315
    },
    {
      "epoch": 0.0316,
      "grad_norm": 31.0072021484375,
      "learning_rate": 4.890909090909091e-05,
      "loss": 0.3179,
      "step": 316
    },
    {
      "epoch": 0.0317,
      "grad_norm": 6.993695259094238,
      "learning_rate": 4.890404040404041e-05,
      "loss": 0.8467,
      "step": 317
    },
    {
      "epoch": 0.0318,
      "grad_norm": 15.915569305419922,
      "learning_rate": 4.88989898989899e-05,
      "loss": 0.6704,
      "step": 318
    },
    {
      "epoch": 0.0319,
      "grad_norm": 239.8886260986328,
      "learning_rate": 4.88939393939394e-05,
      "loss": 0.5967,
      "step": 319
    },
    {
      "epoch": 0.032,
      "grad_norm": 7.528474807739258,
      "learning_rate": 4.888888888888889e-05,
      "loss": 0.392,
      "step": 320
    },
    {
      "epoch": 0.0321,
      "grad_norm": 11.318697929382324,
      "learning_rate": 4.8883838383838386e-05,
      "loss": 1.1082,
      "step": 321
    },
    {
      "epoch": 0.0322,
      "grad_norm": 9.048145294189453,
      "learning_rate": 4.887878787878788e-05,
      "loss": 0.6736,
      "step": 322
    },
    {
      "epoch": 0.0323,
      "grad_norm": 15.707635879516602,
      "learning_rate": 4.887373737373738e-05,
      "loss": 0.9181,
      "step": 323
    },
    {
      "epoch": 0.0324,
      "grad_norm": 12.557865142822266,
      "learning_rate": 4.886868686868687e-05,
      "loss": 0.29,
      "step": 324
    },
    {
      "epoch": 0.0325,
      "grad_norm": 9.12903881072998,
      "learning_rate": 4.886363636363637e-05,
      "loss": 0.7495,
      "step": 325
    },
    {
      "epoch": 0.0326,
      "grad_norm": 8.758808135986328,
      "learning_rate": 4.885858585858586e-05,
      "loss": 0.6794,
      "step": 326
    },
    {
      "epoch": 0.0327,
      "grad_norm": 12.683980941772461,
      "learning_rate": 4.8853535353535355e-05,
      "loss": 0.9999,
      "step": 327
    },
    {
      "epoch": 0.0328,
      "grad_norm": 14.612350463867188,
      "learning_rate": 4.884848484848485e-05,
      "loss": 0.3855,
      "step": 328
    },
    {
      "epoch": 0.0329,
      "grad_norm": 21.392486572265625,
      "learning_rate": 4.884343434343435e-05,
      "loss": 1.2688,
      "step": 329
    },
    {
      "epoch": 0.033,
      "grad_norm": 15.868525505065918,
      "learning_rate": 4.8838383838383836e-05,
      "loss": 0.5956,
      "step": 330
    },
    {
      "epoch": 0.0331,
      "grad_norm": 7.198798179626465,
      "learning_rate": 4.883333333333334e-05,
      "loss": 0.5013,
      "step": 331
    },
    {
      "epoch": 0.0332,
      "grad_norm": 19.240337371826172,
      "learning_rate": 4.882828282828283e-05,
      "loss": 0.4128,
      "step": 332
    },
    {
      "epoch": 0.0333,
      "grad_norm": 20.93824005126953,
      "learning_rate": 4.8823232323232324e-05,
      "loss": 1.0867,
      "step": 333
    },
    {
      "epoch": 0.0334,
      "grad_norm": 14.670286178588867,
      "learning_rate": 4.881818181818182e-05,
      "loss": 1.3094,
      "step": 334
    },
    {
      "epoch": 0.0335,
      "grad_norm": 14.4096097946167,
      "learning_rate": 4.8813131313131316e-05,
      "loss": 1.4552,
      "step": 335
    },
    {
      "epoch": 0.0336,
      "grad_norm": 8.753094673156738,
      "learning_rate": 4.8808080808080805e-05,
      "loss": 0.517,
      "step": 336
    },
    {
      "epoch": 0.0337,
      "grad_norm": 9.008380889892578,
      "learning_rate": 4.880303030303031e-05,
      "loss": 0.2669,
      "step": 337
    },
    {
      "epoch": 0.0338,
      "grad_norm": 7.411258697509766,
      "learning_rate": 4.87979797979798e-05,
      "loss": 0.5625,
      "step": 338
    },
    {
      "epoch": 0.0339,
      "grad_norm": 36.413265228271484,
      "learning_rate": 4.879292929292929e-05,
      "loss": 0.9905,
      "step": 339
    },
    {
      "epoch": 0.034,
      "grad_norm": 15.959607124328613,
      "learning_rate": 4.878787878787879e-05,
      "loss": 0.8361,
      "step": 340
    },
    {
      "epoch": 0.0341,
      "grad_norm": 11.098933219909668,
      "learning_rate": 4.8782828282828285e-05,
      "loss": 0.6048,
      "step": 341
    },
    {
      "epoch": 0.0342,
      "grad_norm": 17.87739372253418,
      "learning_rate": 4.8777777777777775e-05,
      "loss": 0.2568,
      "step": 342
    },
    {
      "epoch": 0.0343,
      "grad_norm": 14.206648826599121,
      "learning_rate": 4.877272727272728e-05,
      "loss": 0.8438,
      "step": 343
    },
    {
      "epoch": 0.0344,
      "grad_norm": 15.068122863769531,
      "learning_rate": 4.8767676767676767e-05,
      "loss": 1.3168,
      "step": 344
    },
    {
      "epoch": 0.0345,
      "grad_norm": 9.560630798339844,
      "learning_rate": 4.876262626262626e-05,
      "loss": 1.02,
      "step": 345
    },
    {
      "epoch": 0.0346,
      "grad_norm": 6.160396575927734,
      "learning_rate": 4.875757575757576e-05,
      "loss": 1.2475,
      "step": 346
    },
    {
      "epoch": 0.0347,
      "grad_norm": 10.107789993286133,
      "learning_rate": 4.8752525252525255e-05,
      "loss": 0.5989,
      "step": 347
    },
    {
      "epoch": 0.0348,
      "grad_norm": 18.293609619140625,
      "learning_rate": 4.874747474747475e-05,
      "loss": 1.2099,
      "step": 348
    },
    {
      "epoch": 0.0349,
      "grad_norm": 7.829981327056885,
      "learning_rate": 4.8742424242424246e-05,
      "loss": 1.1347,
      "step": 349
    },
    {
      "epoch": 0.035,
      "grad_norm": 8.208759307861328,
      "learning_rate": 4.8737373737373736e-05,
      "loss": 0.608,
      "step": 350
    },
    {
      "epoch": 0.0351,
      "grad_norm": 18.28854751586914,
      "learning_rate": 4.873232323232323e-05,
      "loss": 1.0756,
      "step": 351
    },
    {
      "epoch": 0.0352,
      "grad_norm": 12.385089874267578,
      "learning_rate": 4.872727272727273e-05,
      "loss": 1.2539,
      "step": 352
    },
    {
      "epoch": 0.0353,
      "grad_norm": 7.9065117835998535,
      "learning_rate": 4.8722222222222224e-05,
      "loss": 0.5372,
      "step": 353
    },
    {
      "epoch": 0.0354,
      "grad_norm": 12.812912940979004,
      "learning_rate": 4.871717171717172e-05,
      "loss": 0.6893,
      "step": 354
    },
    {
      "epoch": 0.0355,
      "grad_norm": 6.888930320739746,
      "learning_rate": 4.8712121212121216e-05,
      "loss": 0.6978,
      "step": 355
    },
    {
      "epoch": 0.0356,
      "grad_norm": 14.430351257324219,
      "learning_rate": 4.8707070707070705e-05,
      "loss": 0.965,
      "step": 356
    },
    {
      "epoch": 0.0357,
      "grad_norm": 7.433613300323486,
      "learning_rate": 4.870202020202021e-05,
      "loss": 0.3785,
      "step": 357
    },
    {
      "epoch": 0.0358,
      "grad_norm": 7.567981243133545,
      "learning_rate": 4.8696969696969704e-05,
      "loss": 0.8996,
      "step": 358
    },
    {
      "epoch": 0.0359,
      "grad_norm": 17.922311782836914,
      "learning_rate": 4.869191919191919e-05,
      "loss": 0.6871,
      "step": 359
    },
    {
      "epoch": 0.036,
      "grad_norm": 11.450801849365234,
      "learning_rate": 4.868686868686869e-05,
      "loss": 0.7456,
      "step": 360
    },
    {
      "epoch": 0.0361,
      "grad_norm": 10.464077949523926,
      "learning_rate": 4.8681818181818185e-05,
      "loss": 0.7888,
      "step": 361
    },
    {
      "epoch": 0.0362,
      "grad_norm": 19.166343688964844,
      "learning_rate": 4.867676767676768e-05,
      "loss": 1.2915,
      "step": 362
    },
    {
      "epoch": 0.0363,
      "grad_norm": 11.414459228515625,
      "learning_rate": 4.867171717171718e-05,
      "loss": 0.559,
      "step": 363
    },
    {
      "epoch": 0.0364,
      "grad_norm": 10.059521675109863,
      "learning_rate": 4.866666666666667e-05,
      "loss": 0.7673,
      "step": 364
    },
    {
      "epoch": 0.0365,
      "grad_norm": 26.71124839782715,
      "learning_rate": 4.866161616161616e-05,
      "loss": 0.7282,
      "step": 365
    },
    {
      "epoch": 0.0366,
      "grad_norm": 19.962120056152344,
      "learning_rate": 4.865656565656566e-05,
      "loss": 0.7094,
      "step": 366
    },
    {
      "epoch": 0.0367,
      "grad_norm": 28.424602508544922,
      "learning_rate": 4.8651515151515154e-05,
      "loss": 0.9619,
      "step": 367
    },
    {
      "epoch": 0.0368,
      "grad_norm": 12.629812240600586,
      "learning_rate": 4.864646464646465e-05,
      "loss": 0.8908,
      "step": 368
    },
    {
      "epoch": 0.0369,
      "grad_norm": 20.333589553833008,
      "learning_rate": 4.8641414141414146e-05,
      "loss": 0.4261,
      "step": 369
    },
    {
      "epoch": 0.037,
      "grad_norm": 14.529850006103516,
      "learning_rate": 4.863636363636364e-05,
      "loss": 0.636,
      "step": 370
    },
    {
      "epoch": 0.0371,
      "grad_norm": 9.732831954956055,
      "learning_rate": 4.863131313131313e-05,
      "loss": 0.9047,
      "step": 371
    },
    {
      "epoch": 0.0372,
      "grad_norm": 13.312891006469727,
      "learning_rate": 4.8626262626262634e-05,
      "loss": 1.0083,
      "step": 372
    },
    {
      "epoch": 0.0373,
      "grad_norm": 6.541841506958008,
      "learning_rate": 4.862121212121212e-05,
      "loss": 0.7656,
      "step": 373
    },
    {
      "epoch": 0.0374,
      "grad_norm": 18.138639450073242,
      "learning_rate": 4.861616161616162e-05,
      "loss": 1.2074,
      "step": 374
    },
    {
      "epoch": 0.0375,
      "grad_norm": 13.19892692565918,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.1779,
      "step": 375
    },
    {
      "epoch": 0.0376,
      "grad_norm": 10.944900512695312,
      "learning_rate": 4.860606060606061e-05,
      "loss": 1.4522,
      "step": 376
    },
    {
      "epoch": 0.0377,
      "grad_norm": 10.266851425170898,
      "learning_rate": 4.86010101010101e-05,
      "loss": 0.9194,
      "step": 377
    },
    {
      "epoch": 0.0378,
      "grad_norm": 11.814465522766113,
      "learning_rate": 4.85959595959596e-05,
      "loss": 1.1232,
      "step": 378
    },
    {
      "epoch": 0.0379,
      "grad_norm": 21.817598342895508,
      "learning_rate": 4.859090909090909e-05,
      "loss": 0.8054,
      "step": 379
    },
    {
      "epoch": 0.038,
      "grad_norm": 13.01474380493164,
      "learning_rate": 4.858585858585859e-05,
      "loss": 0.4494,
      "step": 380
    },
    {
      "epoch": 0.0381,
      "grad_norm": 31.210464477539062,
      "learning_rate": 4.8580808080808084e-05,
      "loss": 0.2753,
      "step": 381
    },
    {
      "epoch": 0.0382,
      "grad_norm": 13.225306510925293,
      "learning_rate": 4.857575757575758e-05,
      "loss": 1.1849,
      "step": 382
    },
    {
      "epoch": 0.0383,
      "grad_norm": 15.374393463134766,
      "learning_rate": 4.857070707070707e-05,
      "loss": 0.4956,
      "step": 383
    },
    {
      "epoch": 0.0384,
      "grad_norm": 6.718681335449219,
      "learning_rate": 4.856565656565657e-05,
      "loss": 0.2611,
      "step": 384
    },
    {
      "epoch": 0.0385,
      "grad_norm": 10.369194984436035,
      "learning_rate": 4.856060606060606e-05,
      "loss": 0.5714,
      "step": 385
    },
    {
      "epoch": 0.0386,
      "grad_norm": 7.143906116485596,
      "learning_rate": 4.855555555555556e-05,
      "loss": 1.0328,
      "step": 386
    },
    {
      "epoch": 0.0387,
      "grad_norm": 4.076488971710205,
      "learning_rate": 4.8550505050505053e-05,
      "loss": 0.4252,
      "step": 387
    },
    {
      "epoch": 0.0388,
      "grad_norm": 17.2523136138916,
      "learning_rate": 4.854545454545455e-05,
      "loss": 0.6196,
      "step": 388
    },
    {
      "epoch": 0.0389,
      "grad_norm": 11.443089485168457,
      "learning_rate": 4.854040404040404e-05,
      "loss": 0.558,
      "step": 389
    },
    {
      "epoch": 0.039,
      "grad_norm": 7.043203353881836,
      "learning_rate": 4.853535353535354e-05,
      "loss": 0.4216,
      "step": 390
    },
    {
      "epoch": 0.0391,
      "grad_norm": 20.377119064331055,
      "learning_rate": 4.853030303030303e-05,
      "loss": 0.5879,
      "step": 391
    },
    {
      "epoch": 0.0392,
      "grad_norm": 14.212441444396973,
      "learning_rate": 4.852525252525253e-05,
      "loss": 0.9374,
      "step": 392
    },
    {
      "epoch": 0.0393,
      "grad_norm": 19.898862838745117,
      "learning_rate": 4.852020202020202e-05,
      "loss": 0.8806,
      "step": 393
    },
    {
      "epoch": 0.0394,
      "grad_norm": 37.54584884643555,
      "learning_rate": 4.851515151515152e-05,
      "loss": 0.4911,
      "step": 394
    },
    {
      "epoch": 0.0395,
      "grad_norm": 14.015514373779297,
      "learning_rate": 4.851010101010101e-05,
      "loss": 1.3867,
      "step": 395
    },
    {
      "epoch": 0.0396,
      "grad_norm": 11.118803977966309,
      "learning_rate": 4.850505050505051e-05,
      "loss": 0.8367,
      "step": 396
    },
    {
      "epoch": 0.0397,
      "grad_norm": 7.441770076751709,
      "learning_rate": 4.85e-05,
      "loss": 1.1458,
      "step": 397
    },
    {
      "epoch": 0.0398,
      "grad_norm": 12.461899757385254,
      "learning_rate": 4.8494949494949496e-05,
      "loss": 0.6276,
      "step": 398
    },
    {
      "epoch": 0.0399,
      "grad_norm": 13.511693000793457,
      "learning_rate": 4.848989898989899e-05,
      "loss": 1.1515,
      "step": 399
    },
    {
      "epoch": 0.04,
      "grad_norm": 23.601058959960938,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.7651,
      "step": 400
    },
    {
      "epoch": 0.0401,
      "grad_norm": 5.617293834686279,
      "learning_rate": 4.8479797979797984e-05,
      "loss": 0.5046,
      "step": 401
    },
    {
      "epoch": 0.0402,
      "grad_norm": 8.16679859161377,
      "learning_rate": 4.847474747474748e-05,
      "loss": 1.161,
      "step": 402
    },
    {
      "epoch": 0.0403,
      "grad_norm": 9.897177696228027,
      "learning_rate": 4.846969696969697e-05,
      "loss": 0.4938,
      "step": 403
    },
    {
      "epoch": 0.0404,
      "grad_norm": 14.34158706665039,
      "learning_rate": 4.8464646464646465e-05,
      "loss": 1.3358,
      "step": 404
    },
    {
      "epoch": 0.0405,
      "grad_norm": 10.053960800170898,
      "learning_rate": 4.845959595959596e-05,
      "loss": 0.6779,
      "step": 405
    },
    {
      "epoch": 0.0406,
      "grad_norm": 17.740015029907227,
      "learning_rate": 4.845454545454546e-05,
      "loss": 1.0468,
      "step": 406
    },
    {
      "epoch": 0.0407,
      "grad_norm": 10.370110511779785,
      "learning_rate": 4.844949494949495e-05,
      "loss": 1.3736,
      "step": 407
    },
    {
      "epoch": 0.0408,
      "grad_norm": 34.03911590576172,
      "learning_rate": 4.844444444444445e-05,
      "loss": 2.4821,
      "step": 408
    },
    {
      "epoch": 0.0409,
      "grad_norm": 15.810975074768066,
      "learning_rate": 4.843939393939394e-05,
      "loss": 0.6964,
      "step": 409
    },
    {
      "epoch": 0.041,
      "grad_norm": 22.661298751831055,
      "learning_rate": 4.843434343434344e-05,
      "loss": 1.4005,
      "step": 410
    },
    {
      "epoch": 0.0411,
      "grad_norm": 12.591663360595703,
      "learning_rate": 4.842929292929293e-05,
      "loss": 1.0643,
      "step": 411
    },
    {
      "epoch": 0.0412,
      "grad_norm": 13.346220016479492,
      "learning_rate": 4.8424242424242426e-05,
      "loss": 1.0117,
      "step": 412
    },
    {
      "epoch": 0.0413,
      "grad_norm": 18.14653968811035,
      "learning_rate": 4.841919191919192e-05,
      "loss": 1.3417,
      "step": 413
    },
    {
      "epoch": 0.0414,
      "grad_norm": 12.649237632751465,
      "learning_rate": 4.841414141414142e-05,
      "loss": 1.4863,
      "step": 414
    },
    {
      "epoch": 0.0415,
      "grad_norm": 21.664125442504883,
      "learning_rate": 4.840909090909091e-05,
      "loss": 1.4603,
      "step": 415
    },
    {
      "epoch": 0.0416,
      "grad_norm": 10.049198150634766,
      "learning_rate": 4.840404040404041e-05,
      "loss": 0.3523,
      "step": 416
    },
    {
      "epoch": 0.0417,
      "grad_norm": 10.647835731506348,
      "learning_rate": 4.83989898989899e-05,
      "loss": 0.5765,
      "step": 417
    },
    {
      "epoch": 0.0418,
      "grad_norm": 14.153108596801758,
      "learning_rate": 4.8393939393939395e-05,
      "loss": 0.8817,
      "step": 418
    },
    {
      "epoch": 0.0419,
      "grad_norm": 8.155879974365234,
      "learning_rate": 4.838888888888889e-05,
      "loss": 0.8779,
      "step": 419
    },
    {
      "epoch": 0.042,
      "grad_norm": 16.285446166992188,
      "learning_rate": 4.838383838383839e-05,
      "loss": 1.5017,
      "step": 420
    },
    {
      "epoch": 0.0421,
      "grad_norm": 13.782051086425781,
      "learning_rate": 4.8378787878787876e-05,
      "loss": 0.9785,
      "step": 421
    },
    {
      "epoch": 0.0422,
      "grad_norm": 8.297294616699219,
      "learning_rate": 4.837373737373738e-05,
      "loss": 0.5833,
      "step": 422
    },
    {
      "epoch": 0.0423,
      "grad_norm": 8.140225410461426,
      "learning_rate": 4.836868686868687e-05,
      "loss": 0.3891,
      "step": 423
    },
    {
      "epoch": 0.0424,
      "grad_norm": 12.897626876831055,
      "learning_rate": 4.8363636363636364e-05,
      "loss": 0.5859,
      "step": 424
    },
    {
      "epoch": 0.0425,
      "grad_norm": 8.002643585205078,
      "learning_rate": 4.835858585858586e-05,
      "loss": 1.2898,
      "step": 425
    },
    {
      "epoch": 0.0426,
      "grad_norm": 6.414963245391846,
      "learning_rate": 4.8353535353535356e-05,
      "loss": 0.8543,
      "step": 426
    },
    {
      "epoch": 0.0427,
      "grad_norm": 15.286041259765625,
      "learning_rate": 4.8348484848484846e-05,
      "loss": 1.5794,
      "step": 427
    },
    {
      "epoch": 0.0428,
      "grad_norm": 7.241785049438477,
      "learning_rate": 4.834343434343435e-05,
      "loss": 0.8121,
      "step": 428
    },
    {
      "epoch": 0.0429,
      "grad_norm": 17.483112335205078,
      "learning_rate": 4.833838383838384e-05,
      "loss": 0.6508,
      "step": 429
    },
    {
      "epoch": 0.043,
      "grad_norm": 7.844998359680176,
      "learning_rate": 4.8333333333333334e-05,
      "loss": 1.0334,
      "step": 430
    },
    {
      "epoch": 0.0431,
      "grad_norm": 28.124364852905273,
      "learning_rate": 4.832828282828283e-05,
      "loss": 0.9534,
      "step": 431
    },
    {
      "epoch": 0.0432,
      "grad_norm": 24.504833221435547,
      "learning_rate": 4.8323232323232326e-05,
      "loss": 0.8051,
      "step": 432
    },
    {
      "epoch": 0.0433,
      "grad_norm": 7.719658851623535,
      "learning_rate": 4.8318181818181815e-05,
      "loss": 0.4189,
      "step": 433
    },
    {
      "epoch": 0.0434,
      "grad_norm": 7.990880489349365,
      "learning_rate": 4.831313131313132e-05,
      "loss": 0.5447,
      "step": 434
    },
    {
      "epoch": 0.0435,
      "grad_norm": 9.898496627807617,
      "learning_rate": 4.830808080808081e-05,
      "loss": 0.5401,
      "step": 435
    },
    {
      "epoch": 0.0436,
      "grad_norm": 6.576588153839111,
      "learning_rate": 4.83030303030303e-05,
      "loss": 0.3878,
      "step": 436
    },
    {
      "epoch": 0.0437,
      "grad_norm": 8.307194709777832,
      "learning_rate": 4.82979797979798e-05,
      "loss": 0.6891,
      "step": 437
    },
    {
      "epoch": 0.0438,
      "grad_norm": 10.640518188476562,
      "learning_rate": 4.8292929292929295e-05,
      "loss": 0.4814,
      "step": 438
    },
    {
      "epoch": 0.0439,
      "grad_norm": 12.610625267028809,
      "learning_rate": 4.828787878787879e-05,
      "loss": 0.4904,
      "step": 439
    },
    {
      "epoch": 0.044,
      "grad_norm": 12.677209854125977,
      "learning_rate": 4.828282828282829e-05,
      "loss": 0.4039,
      "step": 440
    },
    {
      "epoch": 0.0441,
      "grad_norm": 33.20845413208008,
      "learning_rate": 4.8277777777777776e-05,
      "loss": 1.2058,
      "step": 441
    },
    {
      "epoch": 0.0442,
      "grad_norm": 8.6856050491333,
      "learning_rate": 4.827272727272727e-05,
      "loss": 0.8692,
      "step": 442
    },
    {
      "epoch": 0.0443,
      "grad_norm": 7.953062534332275,
      "learning_rate": 4.8267676767676775e-05,
      "loss": 0.7384,
      "step": 443
    },
    {
      "epoch": 0.0444,
      "grad_norm": 9.17030143737793,
      "learning_rate": 4.8262626262626264e-05,
      "loss": 0.9808,
      "step": 444
    },
    {
      "epoch": 0.0445,
      "grad_norm": 8.517611503601074,
      "learning_rate": 4.825757575757576e-05,
      "loss": 0.6245,
      "step": 445
    },
    {
      "epoch": 0.0446,
      "grad_norm": 17.104707717895508,
      "learning_rate": 4.8252525252525256e-05,
      "loss": 0.5952,
      "step": 446
    },
    {
      "epoch": 0.0447,
      "grad_norm": 34.37602996826172,
      "learning_rate": 4.824747474747475e-05,
      "loss": 0.6299,
      "step": 447
    },
    {
      "epoch": 0.0448,
      "grad_norm": 60.394920349121094,
      "learning_rate": 4.824242424242425e-05,
      "loss": 1.1893,
      "step": 448
    },
    {
      "epoch": 0.0449,
      "grad_norm": 10.541866302490234,
      "learning_rate": 4.8237373737373744e-05,
      "loss": 0.5926,
      "step": 449
    },
    {
      "epoch": 0.045,
      "grad_norm": 9.046228408813477,
      "learning_rate": 4.823232323232323e-05,
      "loss": 0.7712,
      "step": 450
    },
    {
      "epoch": 0.0451,
      "grad_norm": 16.005712509155273,
      "learning_rate": 4.822727272727273e-05,
      "loss": 0.8696,
      "step": 451
    },
    {
      "epoch": 0.0452,
      "grad_norm": 29.7133731842041,
      "learning_rate": 4.8222222222222225e-05,
      "loss": 0.9574,
      "step": 452
    },
    {
      "epoch": 0.0453,
      "grad_norm": 13.506728172302246,
      "learning_rate": 4.821717171717172e-05,
      "loss": 0.3234,
      "step": 453
    },
    {
      "epoch": 0.0454,
      "grad_norm": 13.250686645507812,
      "learning_rate": 4.821212121212122e-05,
      "loss": 0.8468,
      "step": 454
    },
    {
      "epoch": 0.0455,
      "grad_norm": 10.870772361755371,
      "learning_rate": 4.820707070707071e-05,
      "loss": 0.909,
      "step": 455
    },
    {
      "epoch": 0.0456,
      "grad_norm": 8.403849601745605,
      "learning_rate": 4.82020202020202e-05,
      "loss": 0.6264,
      "step": 456
    },
    {
      "epoch": 0.0457,
      "grad_norm": 9.790478706359863,
      "learning_rate": 4.81969696969697e-05,
      "loss": 1.3182,
      "step": 457
    },
    {
      "epoch": 0.0458,
      "grad_norm": 6.866161346435547,
      "learning_rate": 4.8191919191919194e-05,
      "loss": 0.7229,
      "step": 458
    },
    {
      "epoch": 0.0459,
      "grad_norm": 10.686247825622559,
      "learning_rate": 4.818686868686869e-05,
      "loss": 0.6584,
      "step": 459
    },
    {
      "epoch": 0.046,
      "grad_norm": 12.36663818359375,
      "learning_rate": 4.8181818181818186e-05,
      "loss": 0.7243,
      "step": 460
    },
    {
      "epoch": 0.0461,
      "grad_norm": 29.45331573486328,
      "learning_rate": 4.817676767676768e-05,
      "loss": 0.7155,
      "step": 461
    },
    {
      "epoch": 0.0462,
      "grad_norm": 9.957222938537598,
      "learning_rate": 4.817171717171717e-05,
      "loss": 1.035,
      "step": 462
    },
    {
      "epoch": 0.0463,
      "grad_norm": 9.982279777526855,
      "learning_rate": 4.8166666666666674e-05,
      "loss": 0.7029,
      "step": 463
    },
    {
      "epoch": 0.0464,
      "grad_norm": 8.960149765014648,
      "learning_rate": 4.8161616161616163e-05,
      "loss": 0.6482,
      "step": 464
    },
    {
      "epoch": 0.0465,
      "grad_norm": 17.250402450561523,
      "learning_rate": 4.815656565656566e-05,
      "loss": 0.6139,
      "step": 465
    },
    {
      "epoch": 0.0466,
      "grad_norm": 9.218216896057129,
      "learning_rate": 4.8151515151515155e-05,
      "loss": 0.4765,
      "step": 466
    },
    {
      "epoch": 0.0467,
      "grad_norm": 16.008926391601562,
      "learning_rate": 4.814646464646465e-05,
      "loss": 0.4798,
      "step": 467
    },
    {
      "epoch": 0.0468,
      "grad_norm": 10.026300430297852,
      "learning_rate": 4.814141414141414e-05,
      "loss": 0.4937,
      "step": 468
    },
    {
      "epoch": 0.0469,
      "grad_norm": 8.620448112487793,
      "learning_rate": 4.813636363636364e-05,
      "loss": 1.1495,
      "step": 469
    },
    {
      "epoch": 0.047,
      "grad_norm": 11.26388168334961,
      "learning_rate": 4.813131313131313e-05,
      "loss": 0.3594,
      "step": 470
    },
    {
      "epoch": 0.0471,
      "grad_norm": 11.656169891357422,
      "learning_rate": 4.812626262626263e-05,
      "loss": 1.0631,
      "step": 471
    },
    {
      "epoch": 0.0472,
      "grad_norm": 20.822431564331055,
      "learning_rate": 4.8121212121212125e-05,
      "loss": 0.5391,
      "step": 472
    },
    {
      "epoch": 0.0473,
      "grad_norm": 16.706493377685547,
      "learning_rate": 4.811616161616162e-05,
      "loss": 1.2127,
      "step": 473
    },
    {
      "epoch": 0.0474,
      "grad_norm": 13.5263090133667,
      "learning_rate": 4.811111111111111e-05,
      "loss": 0.529,
      "step": 474
    },
    {
      "epoch": 0.0475,
      "grad_norm": 18.68608856201172,
      "learning_rate": 4.810606060606061e-05,
      "loss": 0.6907,
      "step": 475
    },
    {
      "epoch": 0.0476,
      "grad_norm": 21.239158630371094,
      "learning_rate": 4.81010101010101e-05,
      "loss": 0.5721,
      "step": 476
    },
    {
      "epoch": 0.0477,
      "grad_norm": 8.555880546569824,
      "learning_rate": 4.80959595959596e-05,
      "loss": 0.7194,
      "step": 477
    },
    {
      "epoch": 0.0478,
      "grad_norm": 8.348388671875,
      "learning_rate": 4.8090909090909094e-05,
      "loss": 0.543,
      "step": 478
    },
    {
      "epoch": 0.0479,
      "grad_norm": 13.713888168334961,
      "learning_rate": 4.808585858585859e-05,
      "loss": 0.5408,
      "step": 479
    },
    {
      "epoch": 0.048,
      "grad_norm": 29.810054779052734,
      "learning_rate": 4.808080808080808e-05,
      "loss": 0.6116,
      "step": 480
    },
    {
      "epoch": 0.0481,
      "grad_norm": 9.750988960266113,
      "learning_rate": 4.807575757575758e-05,
      "loss": 0.5767,
      "step": 481
    },
    {
      "epoch": 0.0482,
      "grad_norm": 61.696598052978516,
      "learning_rate": 4.807070707070707e-05,
      "loss": 0.6656,
      "step": 482
    },
    {
      "epoch": 0.0483,
      "grad_norm": 23.422903060913086,
      "learning_rate": 4.806565656565657e-05,
      "loss": 1.1034,
      "step": 483
    },
    {
      "epoch": 0.0484,
      "grad_norm": 11.548199653625488,
      "learning_rate": 4.806060606060606e-05,
      "loss": 0.7642,
      "step": 484
    },
    {
      "epoch": 0.0485,
      "grad_norm": 7.780558109283447,
      "learning_rate": 4.805555555555556e-05,
      "loss": 0.7172,
      "step": 485
    },
    {
      "epoch": 0.0486,
      "grad_norm": 5.460698127746582,
      "learning_rate": 4.805050505050505e-05,
      "loss": 0.4797,
      "step": 486
    },
    {
      "epoch": 0.0487,
      "grad_norm": 9.899198532104492,
      "learning_rate": 4.804545454545455e-05,
      "loss": 0.5858,
      "step": 487
    },
    {
      "epoch": 0.0488,
      "grad_norm": 25.51073455810547,
      "learning_rate": 4.804040404040404e-05,
      "loss": 0.7125,
      "step": 488
    },
    {
      "epoch": 0.0489,
      "grad_norm": 22.650386810302734,
      "learning_rate": 4.8035353535353536e-05,
      "loss": 0.8319,
      "step": 489
    },
    {
      "epoch": 0.049,
      "grad_norm": 9.198136329650879,
      "learning_rate": 4.803030303030303e-05,
      "loss": 0.4875,
      "step": 490
    },
    {
      "epoch": 0.0491,
      "grad_norm": 12.245604515075684,
      "learning_rate": 4.802525252525253e-05,
      "loss": 0.6316,
      "step": 491
    },
    {
      "epoch": 0.0492,
      "grad_norm": 15.670899391174316,
      "learning_rate": 4.8020202020202024e-05,
      "loss": 1.1094,
      "step": 492
    },
    {
      "epoch": 0.0493,
      "grad_norm": 14.815546989440918,
      "learning_rate": 4.801515151515152e-05,
      "loss": 1.6366,
      "step": 493
    },
    {
      "epoch": 0.0494,
      "grad_norm": 10.705636024475098,
      "learning_rate": 4.801010101010101e-05,
      "loss": 0.6209,
      "step": 494
    },
    {
      "epoch": 0.0495,
      "grad_norm": 5.984363079071045,
      "learning_rate": 4.8005050505050505e-05,
      "loss": 1.3406,
      "step": 495
    },
    {
      "epoch": 0.0496,
      "grad_norm": 17.70397186279297,
      "learning_rate": 4.8e-05,
      "loss": 0.9204,
      "step": 496
    },
    {
      "epoch": 0.0497,
      "grad_norm": 21.12493133544922,
      "learning_rate": 4.79949494949495e-05,
      "loss": 1.3471,
      "step": 497
    },
    {
      "epoch": 0.0498,
      "grad_norm": 17.487226486206055,
      "learning_rate": 4.798989898989899e-05,
      "loss": 1.2869,
      "step": 498
    },
    {
      "epoch": 0.0499,
      "grad_norm": 16.75757598876953,
      "learning_rate": 4.798484848484849e-05,
      "loss": 1.0526,
      "step": 499
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.328012466430664,
      "learning_rate": 4.797979797979798e-05,
      "loss": 1.2136,
      "step": 500
    },
    {
      "epoch": 0.0501,
      "grad_norm": 17.00678062438965,
      "learning_rate": 4.797474747474748e-05,
      "loss": 0.9174,
      "step": 501
    },
    {
      "epoch": 0.0502,
      "grad_norm": 13.656867027282715,
      "learning_rate": 4.796969696969697e-05,
      "loss": 0.761,
      "step": 502
    },
    {
      "epoch": 0.0503,
      "grad_norm": 10.717095375061035,
      "learning_rate": 4.7964646464646466e-05,
      "loss": 0.7384,
      "step": 503
    },
    {
      "epoch": 0.0504,
      "grad_norm": 16.668622970581055,
      "learning_rate": 4.795959595959596e-05,
      "loss": 0.815,
      "step": 504
    },
    {
      "epoch": 0.0505,
      "grad_norm": 13.69119644165039,
      "learning_rate": 4.795454545454546e-05,
      "loss": 0.5949,
      "step": 505
    },
    {
      "epoch": 0.0506,
      "grad_norm": 11.282341003417969,
      "learning_rate": 4.794949494949495e-05,
      "loss": 0.8142,
      "step": 506
    },
    {
      "epoch": 0.0507,
      "grad_norm": 6.222525596618652,
      "learning_rate": 4.794444444444445e-05,
      "loss": 0.5861,
      "step": 507
    },
    {
      "epoch": 0.0508,
      "grad_norm": 7.643500328063965,
      "learning_rate": 4.793939393939394e-05,
      "loss": 0.3149,
      "step": 508
    },
    {
      "epoch": 0.0509,
      "grad_norm": 9.985461235046387,
      "learning_rate": 4.7934343434343436e-05,
      "loss": 0.3595,
      "step": 509
    },
    {
      "epoch": 0.051,
      "grad_norm": 17.415267944335938,
      "learning_rate": 4.792929292929293e-05,
      "loss": 0.3823,
      "step": 510
    },
    {
      "epoch": 0.0511,
      "grad_norm": 8.167945861816406,
      "learning_rate": 4.792424242424243e-05,
      "loss": 0.6358,
      "step": 511
    },
    {
      "epoch": 0.0512,
      "grad_norm": 18.5587100982666,
      "learning_rate": 4.791919191919192e-05,
      "loss": 1.4154,
      "step": 512
    },
    {
      "epoch": 0.0513,
      "grad_norm": 12.890840530395508,
      "learning_rate": 4.791414141414142e-05,
      "loss": 0.5699,
      "step": 513
    },
    {
      "epoch": 0.0514,
      "grad_norm": 9.874274253845215,
      "learning_rate": 4.790909090909091e-05,
      "loss": 0.7811,
      "step": 514
    },
    {
      "epoch": 0.0515,
      "grad_norm": 9.55707836151123,
      "learning_rate": 4.7904040404040405e-05,
      "loss": 0.9996,
      "step": 515
    },
    {
      "epoch": 0.0516,
      "grad_norm": 15.88961124420166,
      "learning_rate": 4.78989898989899e-05,
      "loss": 0.6487,
      "step": 516
    },
    {
      "epoch": 0.0517,
      "grad_norm": 6.5389180183410645,
      "learning_rate": 4.78939393939394e-05,
      "loss": 0.5045,
      "step": 517
    },
    {
      "epoch": 0.0518,
      "grad_norm": 13.842283248901367,
      "learning_rate": 4.7888888888888886e-05,
      "loss": 0.756,
      "step": 518
    },
    {
      "epoch": 0.0519,
      "grad_norm": 11.713705062866211,
      "learning_rate": 4.788383838383839e-05,
      "loss": 0.7266,
      "step": 519
    },
    {
      "epoch": 0.052,
      "grad_norm": 18.057937622070312,
      "learning_rate": 4.787878787878788e-05,
      "loss": 1.1145,
      "step": 520
    },
    {
      "epoch": 0.0521,
      "grad_norm": 14.976628303527832,
      "learning_rate": 4.7873737373737374e-05,
      "loss": 0.5535,
      "step": 521
    },
    {
      "epoch": 0.0522,
      "grad_norm": 9.787546157836914,
      "learning_rate": 4.786868686868687e-05,
      "loss": 0.5118,
      "step": 522
    },
    {
      "epoch": 0.0523,
      "grad_norm": 7.052227973937988,
      "learning_rate": 4.7863636363636366e-05,
      "loss": 0.8466,
      "step": 523
    },
    {
      "epoch": 0.0524,
      "grad_norm": 12.500731468200684,
      "learning_rate": 4.7858585858585855e-05,
      "loss": 2.3732,
      "step": 524
    },
    {
      "epoch": 0.0525,
      "grad_norm": 6.117061138153076,
      "learning_rate": 4.785353535353536e-05,
      "loss": 0.8191,
      "step": 525
    },
    {
      "epoch": 0.0526,
      "grad_norm": 21.544170379638672,
      "learning_rate": 4.784848484848485e-05,
      "loss": 1.4121,
      "step": 526
    },
    {
      "epoch": 0.0527,
      "grad_norm": 17.770788192749023,
      "learning_rate": 4.784343434343434e-05,
      "loss": 0.8991,
      "step": 527
    },
    {
      "epoch": 0.0528,
      "grad_norm": 17.298330307006836,
      "learning_rate": 4.7838383838383846e-05,
      "loss": 0.8517,
      "step": 528
    },
    {
      "epoch": 0.0529,
      "grad_norm": 23.78373146057129,
      "learning_rate": 4.7833333333333335e-05,
      "loss": 0.429,
      "step": 529
    },
    {
      "epoch": 0.053,
      "grad_norm": 7.000113487243652,
      "learning_rate": 4.782828282828283e-05,
      "loss": 0.9779,
      "step": 530
    },
    {
      "epoch": 0.0531,
      "grad_norm": 14.874780654907227,
      "learning_rate": 4.782323232323233e-05,
      "loss": 0.9736,
      "step": 531
    },
    {
      "epoch": 0.0532,
      "grad_norm": 12.712928771972656,
      "learning_rate": 4.781818181818182e-05,
      "loss": 1.2141,
      "step": 532
    },
    {
      "epoch": 0.0533,
      "grad_norm": 157.283935546875,
      "learning_rate": 4.781313131313131e-05,
      "loss": 0.1849,
      "step": 533
    },
    {
      "epoch": 0.0534,
      "grad_norm": 12.13798713684082,
      "learning_rate": 4.7808080808080815e-05,
      "loss": 1.0029,
      "step": 534
    },
    {
      "epoch": 0.0535,
      "grad_norm": 8.694860458374023,
      "learning_rate": 4.7803030303030304e-05,
      "loss": 0.6827,
      "step": 535
    },
    {
      "epoch": 0.0536,
      "grad_norm": 9.920144081115723,
      "learning_rate": 4.77979797979798e-05,
      "loss": 0.5001,
      "step": 536
    },
    {
      "epoch": 0.0537,
      "grad_norm": 18.132814407348633,
      "learning_rate": 4.7792929292929296e-05,
      "loss": 1.4923,
      "step": 537
    },
    {
      "epoch": 0.0538,
      "grad_norm": 13.660527229309082,
      "learning_rate": 4.778787878787879e-05,
      "loss": 0.9071,
      "step": 538
    },
    {
      "epoch": 0.0539,
      "grad_norm": 43.71529769897461,
      "learning_rate": 4.778282828282828e-05,
      "loss": 0.6314,
      "step": 539
    },
    {
      "epoch": 0.054,
      "grad_norm": 6.684488773345947,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 0.4099,
      "step": 540
    },
    {
      "epoch": 0.0541,
      "grad_norm": 9.68802261352539,
      "learning_rate": 4.777272727272727e-05,
      "loss": 1.2339,
      "step": 541
    },
    {
      "epoch": 0.0542,
      "grad_norm": 5.818267822265625,
      "learning_rate": 4.776767676767677e-05,
      "loss": 2.0167,
      "step": 542
    },
    {
      "epoch": 0.0543,
      "grad_norm": 13.396550178527832,
      "learning_rate": 4.7762626262626265e-05,
      "loss": 0.8111,
      "step": 543
    },
    {
      "epoch": 0.0544,
      "grad_norm": 6.180233478546143,
      "learning_rate": 4.775757575757576e-05,
      "loss": 0.8333,
      "step": 544
    },
    {
      "epoch": 0.0545,
      "grad_norm": 7.569714546203613,
      "learning_rate": 4.775252525252526e-05,
      "loss": 1.0088,
      "step": 545
    },
    {
      "epoch": 0.0546,
      "grad_norm": 10.76653003692627,
      "learning_rate": 4.774747474747475e-05,
      "loss": 0.9361,
      "step": 546
    },
    {
      "epoch": 0.0547,
      "grad_norm": 9.213648796081543,
      "learning_rate": 4.774242424242424e-05,
      "loss": 0.726,
      "step": 547
    },
    {
      "epoch": 0.0548,
      "grad_norm": 12.07512092590332,
      "learning_rate": 4.773737373737374e-05,
      "loss": 0.8165,
      "step": 548
    },
    {
      "epoch": 0.0549,
      "grad_norm": 6.379137992858887,
      "learning_rate": 4.7732323232323234e-05,
      "loss": 0.413,
      "step": 549
    },
    {
      "epoch": 0.055,
      "grad_norm": 8.558244705200195,
      "learning_rate": 4.772727272727273e-05,
      "loss": 0.3789,
      "step": 550
    },
    {
      "epoch": 0.0551,
      "grad_norm": 11.15512752532959,
      "learning_rate": 4.7722222222222226e-05,
      "loss": 0.6544,
      "step": 551
    },
    {
      "epoch": 0.0552,
      "grad_norm": 6.666895389556885,
      "learning_rate": 4.771717171717172e-05,
      "loss": 0.4832,
      "step": 552
    },
    {
      "epoch": 0.0553,
      "grad_norm": 9.668054580688477,
      "learning_rate": 4.771212121212121e-05,
      "loss": 0.5093,
      "step": 553
    },
    {
      "epoch": 0.0554,
      "grad_norm": 15.73165225982666,
      "learning_rate": 4.7707070707070714e-05,
      "loss": 0.5958,
      "step": 554
    },
    {
      "epoch": 0.0555,
      "grad_norm": 11.919795036315918,
      "learning_rate": 4.7702020202020204e-05,
      "loss": 0.6412,
      "step": 555
    },
    {
      "epoch": 0.0556,
      "grad_norm": 11.729137420654297,
      "learning_rate": 4.76969696969697e-05,
      "loss": 1.167,
      "step": 556
    },
    {
      "epoch": 0.0557,
      "grad_norm": 13.204307556152344,
      "learning_rate": 4.7691919191919196e-05,
      "loss": 0.5571,
      "step": 557
    },
    {
      "epoch": 0.0558,
      "grad_norm": 12.772867202758789,
      "learning_rate": 4.768686868686869e-05,
      "loss": 0.7087,
      "step": 558
    },
    {
      "epoch": 0.0559,
      "grad_norm": 19.61652946472168,
      "learning_rate": 4.768181818181818e-05,
      "loss": 0.4207,
      "step": 559
    },
    {
      "epoch": 0.056,
      "grad_norm": 19.650575637817383,
      "learning_rate": 4.7676767676767684e-05,
      "loss": 1.1546,
      "step": 560
    },
    {
      "epoch": 0.0561,
      "grad_norm": 8.32691764831543,
      "learning_rate": 4.767171717171717e-05,
      "loss": 0.7897,
      "step": 561
    },
    {
      "epoch": 0.0562,
      "grad_norm": 6.780239105224609,
      "learning_rate": 4.766666666666667e-05,
      "loss": 0.7995,
      "step": 562
    },
    {
      "epoch": 0.0563,
      "grad_norm": 11.454874992370605,
      "learning_rate": 4.7661616161616165e-05,
      "loss": 1.0677,
      "step": 563
    },
    {
      "epoch": 0.0564,
      "grad_norm": 12.181625366210938,
      "learning_rate": 4.765656565656566e-05,
      "loss": 0.4911,
      "step": 564
    },
    {
      "epoch": 0.0565,
      "grad_norm": 13.204384803771973,
      "learning_rate": 4.765151515151515e-05,
      "loss": 0.8083,
      "step": 565
    },
    {
      "epoch": 0.0566,
      "grad_norm": 7.024822235107422,
      "learning_rate": 4.764646464646465e-05,
      "loss": 0.7372,
      "step": 566
    },
    {
      "epoch": 0.0567,
      "grad_norm": 9.412765502929688,
      "learning_rate": 4.764141414141414e-05,
      "loss": 1.0473,
      "step": 567
    },
    {
      "epoch": 0.0568,
      "grad_norm": 11.275609016418457,
      "learning_rate": 4.763636363636364e-05,
      "loss": 0.9539,
      "step": 568
    },
    {
      "epoch": 0.0569,
      "grad_norm": 16.782033920288086,
      "learning_rate": 4.7631313131313134e-05,
      "loss": 0.8618,
      "step": 569
    },
    {
      "epoch": 0.057,
      "grad_norm": 6.9136857986450195,
      "learning_rate": 4.762626262626263e-05,
      "loss": 0.4988,
      "step": 570
    },
    {
      "epoch": 0.0571,
      "grad_norm": 31.620676040649414,
      "learning_rate": 4.762121212121212e-05,
      "loss": 0.8344,
      "step": 571
    },
    {
      "epoch": 0.0572,
      "grad_norm": 21.954177856445312,
      "learning_rate": 4.761616161616162e-05,
      "loss": 0.4662,
      "step": 572
    },
    {
      "epoch": 0.0573,
      "grad_norm": 15.426892280578613,
      "learning_rate": 4.761111111111111e-05,
      "loss": 1.6782,
      "step": 573
    },
    {
      "epoch": 0.0574,
      "grad_norm": 13.740044593811035,
      "learning_rate": 4.760606060606061e-05,
      "loss": 0.9636,
      "step": 574
    },
    {
      "epoch": 0.0575,
      "grad_norm": 7.591161251068115,
      "learning_rate": 4.76010101010101e-05,
      "loss": 0.4966,
      "step": 575
    },
    {
      "epoch": 0.0576,
      "grad_norm": 6.801815986633301,
      "learning_rate": 4.75959595959596e-05,
      "loss": 0.4343,
      "step": 576
    },
    {
      "epoch": 0.0577,
      "grad_norm": 11.44105339050293,
      "learning_rate": 4.759090909090909e-05,
      "loss": 1.1007,
      "step": 577
    },
    {
      "epoch": 0.0578,
      "grad_norm": 7.257315158843994,
      "learning_rate": 4.758585858585859e-05,
      "loss": 0.4426,
      "step": 578
    },
    {
      "epoch": 0.0579,
      "grad_norm": 13.316791534423828,
      "learning_rate": 4.758080808080808e-05,
      "loss": 0.8249,
      "step": 579
    },
    {
      "epoch": 0.058,
      "grad_norm": 14.513383865356445,
      "learning_rate": 4.7575757575757576e-05,
      "loss": 0.7038,
      "step": 580
    },
    {
      "epoch": 0.0581,
      "grad_norm": 11.767633438110352,
      "learning_rate": 4.757070707070707e-05,
      "loss": 0.8575,
      "step": 581
    },
    {
      "epoch": 0.0582,
      "grad_norm": 7.295989513397217,
      "learning_rate": 4.756565656565657e-05,
      "loss": 0.451,
      "step": 582
    },
    {
      "epoch": 0.0583,
      "grad_norm": 31.647180557250977,
      "learning_rate": 4.7560606060606064e-05,
      "loss": 1.2049,
      "step": 583
    },
    {
      "epoch": 0.0584,
      "grad_norm": 7.396601676940918,
      "learning_rate": 4.755555555555556e-05,
      "loss": 0.4248,
      "step": 584
    },
    {
      "epoch": 0.0585,
      "grad_norm": 13.179383277893066,
      "learning_rate": 4.755050505050505e-05,
      "loss": 1.6189,
      "step": 585
    },
    {
      "epoch": 0.0586,
      "grad_norm": 10.316752433776855,
      "learning_rate": 4.7545454545454545e-05,
      "loss": 0.5359,
      "step": 586
    },
    {
      "epoch": 0.0587,
      "grad_norm": 13.060739517211914,
      "learning_rate": 4.754040404040404e-05,
      "loss": 0.5362,
      "step": 587
    },
    {
      "epoch": 0.0588,
      "grad_norm": 22.149280548095703,
      "learning_rate": 4.753535353535354e-05,
      "loss": 0.5258,
      "step": 588
    },
    {
      "epoch": 0.0589,
      "grad_norm": 10.577301979064941,
      "learning_rate": 4.7530303030303033e-05,
      "loss": 0.775,
      "step": 589
    },
    {
      "epoch": 0.059,
      "grad_norm": 24.235462188720703,
      "learning_rate": 4.752525252525253e-05,
      "loss": 0.6519,
      "step": 590
    },
    {
      "epoch": 0.0591,
      "grad_norm": 10.091059684753418,
      "learning_rate": 4.752020202020202e-05,
      "loss": 0.8549,
      "step": 591
    },
    {
      "epoch": 0.0592,
      "grad_norm": 12.989404678344727,
      "learning_rate": 4.751515151515152e-05,
      "loss": 0.3291,
      "step": 592
    },
    {
      "epoch": 0.0593,
      "grad_norm": 21.599058151245117,
      "learning_rate": 4.751010101010101e-05,
      "loss": 0.9889,
      "step": 593
    },
    {
      "epoch": 0.0594,
      "grad_norm": 22.59168815612793,
      "learning_rate": 4.750505050505051e-05,
      "loss": 1.5124,
      "step": 594
    },
    {
      "epoch": 0.0595,
      "grad_norm": 11.120415687561035,
      "learning_rate": 4.75e-05,
      "loss": 0.6384,
      "step": 595
    },
    {
      "epoch": 0.0596,
      "grad_norm": 20.90540313720703,
      "learning_rate": 4.74949494949495e-05,
      "loss": 0.4959,
      "step": 596
    },
    {
      "epoch": 0.0597,
      "grad_norm": 12.164955139160156,
      "learning_rate": 4.748989898989899e-05,
      "loss": 0.9579,
      "step": 597
    },
    {
      "epoch": 0.0598,
      "grad_norm": 7.543600082397461,
      "learning_rate": 4.748484848484849e-05,
      "loss": 0.8209,
      "step": 598
    },
    {
      "epoch": 0.0599,
      "grad_norm": 10.233712196350098,
      "learning_rate": 4.747979797979798e-05,
      "loss": 0.6137,
      "step": 599
    },
    {
      "epoch": 0.06,
      "grad_norm": 22.6064395904541,
      "learning_rate": 4.7474747474747476e-05,
      "loss": 0.5566,
      "step": 600
    },
    {
      "epoch": 0.0601,
      "grad_norm": 7.197295188903809,
      "learning_rate": 4.746969696969697e-05,
      "loss": 0.7037,
      "step": 601
    },
    {
      "epoch": 0.0602,
      "grad_norm": 7.893517971038818,
      "learning_rate": 4.746464646464647e-05,
      "loss": 0.5522,
      "step": 602
    },
    {
      "epoch": 0.0603,
      "grad_norm": 8.366425514221191,
      "learning_rate": 4.745959595959596e-05,
      "loss": 0.6237,
      "step": 603
    },
    {
      "epoch": 0.0604,
      "grad_norm": 7.006853103637695,
      "learning_rate": 4.745454545454546e-05,
      "loss": 0.4059,
      "step": 604
    },
    {
      "epoch": 0.0605,
      "grad_norm": 11.613106727600098,
      "learning_rate": 4.744949494949495e-05,
      "loss": 1.1968,
      "step": 605
    },
    {
      "epoch": 0.0606,
      "grad_norm": 9.174874305725098,
      "learning_rate": 4.7444444444444445e-05,
      "loss": 0.4083,
      "step": 606
    },
    {
      "epoch": 0.0607,
      "grad_norm": 42.756439208984375,
      "learning_rate": 4.743939393939394e-05,
      "loss": 1.1492,
      "step": 607
    },
    {
      "epoch": 0.0608,
      "grad_norm": 5.780830383300781,
      "learning_rate": 4.743434343434344e-05,
      "loss": 0.4781,
      "step": 608
    },
    {
      "epoch": 0.0609,
      "grad_norm": 14.298049926757812,
      "learning_rate": 4.7429292929292926e-05,
      "loss": 1.0038,
      "step": 609
    },
    {
      "epoch": 0.061,
      "grad_norm": 6.508335113525391,
      "learning_rate": 4.742424242424243e-05,
      "loss": 0.6454,
      "step": 610
    },
    {
      "epoch": 0.0611,
      "grad_norm": 11.635771751403809,
      "learning_rate": 4.741919191919192e-05,
      "loss": 0.7262,
      "step": 611
    },
    {
      "epoch": 0.0612,
      "grad_norm": 14.164590835571289,
      "learning_rate": 4.7414141414141414e-05,
      "loss": 0.4201,
      "step": 612
    },
    {
      "epoch": 0.0613,
      "grad_norm": 16.382678985595703,
      "learning_rate": 4.740909090909092e-05,
      "loss": 0.3881,
      "step": 613
    },
    {
      "epoch": 0.0614,
      "grad_norm": 7.79246711730957,
      "learning_rate": 4.7404040404040406e-05,
      "loss": 0.4381,
      "step": 614
    },
    {
      "epoch": 0.0615,
      "grad_norm": 8.691168785095215,
      "learning_rate": 4.73989898989899e-05,
      "loss": 0.3693,
      "step": 615
    },
    {
      "epoch": 0.0616,
      "grad_norm": 12.245211601257324,
      "learning_rate": 4.73939393939394e-05,
      "loss": 0.5765,
      "step": 616
    },
    {
      "epoch": 0.0617,
      "grad_norm": 27.451051712036133,
      "learning_rate": 4.7388888888888894e-05,
      "loss": 0.3884,
      "step": 617
    },
    {
      "epoch": 0.0618,
      "grad_norm": 14.757604598999023,
      "learning_rate": 4.738383838383838e-05,
      "loss": 0.48,
      "step": 618
    },
    {
      "epoch": 0.0619,
      "grad_norm": 6.650001525878906,
      "learning_rate": 4.7378787878787886e-05,
      "loss": 0.3436,
      "step": 619
    },
    {
      "epoch": 0.062,
      "grad_norm": 8.026239395141602,
      "learning_rate": 4.7373737373737375e-05,
      "loss": 0.9777,
      "step": 620
    },
    {
      "epoch": 0.0621,
      "grad_norm": 17.74627685546875,
      "learning_rate": 4.736868686868687e-05,
      "loss": 0.6417,
      "step": 621
    },
    {
      "epoch": 0.0622,
      "grad_norm": 5.392073631286621,
      "learning_rate": 4.736363636363637e-05,
      "loss": 0.751,
      "step": 622
    },
    {
      "epoch": 0.0623,
      "grad_norm": 8.19266128540039,
      "learning_rate": 4.735858585858586e-05,
      "loss": 0.4423,
      "step": 623
    },
    {
      "epoch": 0.0624,
      "grad_norm": 96.91624450683594,
      "learning_rate": 4.735353535353535e-05,
      "loss": 0.9238,
      "step": 624
    },
    {
      "epoch": 0.0625,
      "grad_norm": 27.404258728027344,
      "learning_rate": 4.7348484848484855e-05,
      "loss": 0.4144,
      "step": 625
    },
    {
      "epoch": 0.0626,
      "grad_norm": 7.842190742492676,
      "learning_rate": 4.7343434343434344e-05,
      "loss": 0.5475,
      "step": 626
    },
    {
      "epoch": 0.0627,
      "grad_norm": 7.412627220153809,
      "learning_rate": 4.733838383838384e-05,
      "loss": 0.3878,
      "step": 627
    },
    {
      "epoch": 0.0628,
      "grad_norm": 15.507277488708496,
      "learning_rate": 4.7333333333333336e-05,
      "loss": 0.1,
      "step": 628
    },
    {
      "epoch": 0.0629,
      "grad_norm": 5.0873003005981445,
      "learning_rate": 4.732828282828283e-05,
      "loss": 0.2396,
      "step": 629
    },
    {
      "epoch": 0.063,
      "grad_norm": 11.989870071411133,
      "learning_rate": 4.732323232323232e-05,
      "loss": 0.5671,
      "step": 630
    },
    {
      "epoch": 0.0631,
      "grad_norm": 9.290095329284668,
      "learning_rate": 4.7318181818181824e-05,
      "loss": 0.6286,
      "step": 631
    },
    {
      "epoch": 0.0632,
      "grad_norm": 7.587223052978516,
      "learning_rate": 4.7313131313131314e-05,
      "loss": 0.2387,
      "step": 632
    },
    {
      "epoch": 0.0633,
      "grad_norm": 11.302098274230957,
      "learning_rate": 4.730808080808081e-05,
      "loss": 0.5626,
      "step": 633
    },
    {
      "epoch": 0.0634,
      "grad_norm": 7.648301601409912,
      "learning_rate": 4.7303030303030306e-05,
      "loss": 0.2807,
      "step": 634
    },
    {
      "epoch": 0.0635,
      "grad_norm": 13.879380226135254,
      "learning_rate": 4.72979797979798e-05,
      "loss": 0.4435,
      "step": 635
    },
    {
      "epoch": 0.0636,
      "grad_norm": 5.6118998527526855,
      "learning_rate": 4.72929292929293e-05,
      "loss": 0.4426,
      "step": 636
    },
    {
      "epoch": 0.0637,
      "grad_norm": 11.586562156677246,
      "learning_rate": 4.7287878787878794e-05,
      "loss": 0.2029,
      "step": 637
    },
    {
      "epoch": 0.0638,
      "grad_norm": 3.7730977535247803,
      "learning_rate": 4.728282828282828e-05,
      "loss": 0.0995,
      "step": 638
    },
    {
      "epoch": 0.0639,
      "grad_norm": 2.619572162628174,
      "learning_rate": 4.727777777777778e-05,
      "loss": 0.0703,
      "step": 639
    },
    {
      "epoch": 0.064,
      "grad_norm": 6.645572185516357,
      "learning_rate": 4.7272727272727275e-05,
      "loss": 0.3372,
      "step": 640
    },
    {
      "epoch": 0.0641,
      "grad_norm": 15.155945777893066,
      "learning_rate": 4.726767676767677e-05,
      "loss": 0.4715,
      "step": 641
    },
    {
      "epoch": 0.0642,
      "grad_norm": 9.443462371826172,
      "learning_rate": 4.726262626262627e-05,
      "loss": 0.3082,
      "step": 642
    },
    {
      "epoch": 0.0643,
      "grad_norm": 10.958251953125,
      "learning_rate": 4.725757575757576e-05,
      "loss": 0.9581,
      "step": 643
    },
    {
      "epoch": 0.0644,
      "grad_norm": 8.747466087341309,
      "learning_rate": 4.725252525252525e-05,
      "loss": 0.5137,
      "step": 644
    },
    {
      "epoch": 0.0645,
      "grad_norm": 9.261674880981445,
      "learning_rate": 4.7247474747474755e-05,
      "loss": 0.6771,
      "step": 645
    },
    {
      "epoch": 0.0646,
      "grad_norm": 4.0851922035217285,
      "learning_rate": 4.7242424242424244e-05,
      "loss": 0.1428,
      "step": 646
    },
    {
      "epoch": 0.0647,
      "grad_norm": 5.984081745147705,
      "learning_rate": 4.723737373737374e-05,
      "loss": 0.7303,
      "step": 647
    },
    {
      "epoch": 0.0648,
      "grad_norm": 7.86775541305542,
      "learning_rate": 4.7232323232323236e-05,
      "loss": 0.1995,
      "step": 648
    },
    {
      "epoch": 0.0649,
      "grad_norm": 4.814087867736816,
      "learning_rate": 4.722727272727273e-05,
      "loss": 0.3161,
      "step": 649
    },
    {
      "epoch": 0.065,
      "grad_norm": 4.86448335647583,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.1751,
      "step": 650
    },
    {
      "epoch": 0.0651,
      "grad_norm": 20.645355224609375,
      "learning_rate": 4.7217171717171724e-05,
      "loss": 0.4188,
      "step": 651
    },
    {
      "epoch": 0.0652,
      "grad_norm": 10.306366920471191,
      "learning_rate": 4.721212121212121e-05,
      "loss": 1.477,
      "step": 652
    },
    {
      "epoch": 0.0653,
      "grad_norm": 78.95001220703125,
      "learning_rate": 4.720707070707071e-05,
      "loss": 1.6967,
      "step": 653
    },
    {
      "epoch": 0.0654,
      "grad_norm": 31.779699325561523,
      "learning_rate": 4.7202020202020205e-05,
      "loss": 0.8375,
      "step": 654
    },
    {
      "epoch": 0.0655,
      "grad_norm": 15.083161354064941,
      "learning_rate": 4.71969696969697e-05,
      "loss": 0.6722,
      "step": 655
    },
    {
      "epoch": 0.0656,
      "grad_norm": 9.035998344421387,
      "learning_rate": 4.719191919191919e-05,
      "loss": 0.4292,
      "step": 656
    },
    {
      "epoch": 0.0657,
      "grad_norm": 12.7191743850708,
      "learning_rate": 4.718686868686869e-05,
      "loss": 1.0542,
      "step": 657
    },
    {
      "epoch": 0.0658,
      "grad_norm": 10.864739418029785,
      "learning_rate": 4.718181818181818e-05,
      "loss": 0.2294,
      "step": 658
    },
    {
      "epoch": 0.0659,
      "grad_norm": 9.637469291687012,
      "learning_rate": 4.717676767676768e-05,
      "loss": 0.6032,
      "step": 659
    },
    {
      "epoch": 0.066,
      "grad_norm": 7.577165126800537,
      "learning_rate": 4.7171717171717174e-05,
      "loss": 0.0488,
      "step": 660
    },
    {
      "epoch": 0.0661,
      "grad_norm": 10.12633228302002,
      "learning_rate": 4.716666666666667e-05,
      "loss": 0.6231,
      "step": 661
    },
    {
      "epoch": 0.0662,
      "grad_norm": 12.298242568969727,
      "learning_rate": 4.716161616161616e-05,
      "loss": 0.1611,
      "step": 662
    },
    {
      "epoch": 0.0663,
      "grad_norm": 9.494952201843262,
      "learning_rate": 4.715656565656566e-05,
      "loss": 0.5961,
      "step": 663
    },
    {
      "epoch": 0.0664,
      "grad_norm": 8.113798141479492,
      "learning_rate": 4.715151515151515e-05,
      "loss": 0.2538,
      "step": 664
    },
    {
      "epoch": 0.0665,
      "grad_norm": 11.252359390258789,
      "learning_rate": 4.714646464646465e-05,
      "loss": 0.5797,
      "step": 665
    },
    {
      "epoch": 0.0666,
      "grad_norm": 14.233952522277832,
      "learning_rate": 4.7141414141414143e-05,
      "loss": 1.0949,
      "step": 666
    },
    {
      "epoch": 0.0667,
      "grad_norm": 5.563956260681152,
      "learning_rate": 4.713636363636364e-05,
      "loss": 0.2733,
      "step": 667
    },
    {
      "epoch": 0.0668,
      "grad_norm": 4.358007907867432,
      "learning_rate": 4.713131313131313e-05,
      "loss": 0.3264,
      "step": 668
    },
    {
      "epoch": 0.0669,
      "grad_norm": 10.478852272033691,
      "learning_rate": 4.712626262626263e-05,
      "loss": 0.8221,
      "step": 669
    },
    {
      "epoch": 0.067,
      "grad_norm": 4.1840500831604,
      "learning_rate": 4.712121212121212e-05,
      "loss": 0.1092,
      "step": 670
    },
    {
      "epoch": 0.0671,
      "grad_norm": 6.47021484375,
      "learning_rate": 4.7116161616161617e-05,
      "loss": 0.7028,
      "step": 671
    },
    {
      "epoch": 0.0672,
      "grad_norm": 8.5515775680542,
      "learning_rate": 4.711111111111111e-05,
      "loss": 0.3048,
      "step": 672
    },
    {
      "epoch": 0.0673,
      "grad_norm": 23.477113723754883,
      "learning_rate": 4.710606060606061e-05,
      "loss": 0.5892,
      "step": 673
    },
    {
      "epoch": 0.0674,
      "grad_norm": 12.882564544677734,
      "learning_rate": 4.7101010101010105e-05,
      "loss": 0.2702,
      "step": 674
    },
    {
      "epoch": 0.0675,
      "grad_norm": 5.846103191375732,
      "learning_rate": 4.70959595959596e-05,
      "loss": 0.4743,
      "step": 675
    },
    {
      "epoch": 0.0676,
      "grad_norm": 25.20199966430664,
      "learning_rate": 4.709090909090909e-05,
      "loss": 0.6628,
      "step": 676
    },
    {
      "epoch": 0.0677,
      "grad_norm": 23.499977111816406,
      "learning_rate": 4.7085858585858586e-05,
      "loss": 0.5637,
      "step": 677
    },
    {
      "epoch": 0.0678,
      "grad_norm": 18.081689834594727,
      "learning_rate": 4.708080808080808e-05,
      "loss": 0.5265,
      "step": 678
    },
    {
      "epoch": 0.0679,
      "grad_norm": 3.70007586479187,
      "learning_rate": 4.707575757575758e-05,
      "loss": 0.0876,
      "step": 679
    },
    {
      "epoch": 0.068,
      "grad_norm": 7.008825302124023,
      "learning_rate": 4.7070707070707074e-05,
      "loss": 0.3307,
      "step": 680
    },
    {
      "epoch": 0.0681,
      "grad_norm": 4.308728218078613,
      "learning_rate": 4.706565656565657e-05,
      "loss": 0.1039,
      "step": 681
    },
    {
      "epoch": 0.0682,
      "grad_norm": 6.793579578399658,
      "learning_rate": 4.706060606060606e-05,
      "loss": 0.1301,
      "step": 682
    },
    {
      "epoch": 0.0683,
      "grad_norm": 12.7081937789917,
      "learning_rate": 4.7055555555555555e-05,
      "loss": 1.8409,
      "step": 683
    },
    {
      "epoch": 0.0684,
      "grad_norm": 14.072186470031738,
      "learning_rate": 4.705050505050505e-05,
      "loss": 0.5917,
      "step": 684
    },
    {
      "epoch": 0.0685,
      "grad_norm": 8.23597240447998,
      "learning_rate": 4.704545454545455e-05,
      "loss": 0.4657,
      "step": 685
    },
    {
      "epoch": 0.0686,
      "grad_norm": 15.613850593566895,
      "learning_rate": 4.704040404040404e-05,
      "loss": 0.3272,
      "step": 686
    },
    {
      "epoch": 0.0687,
      "grad_norm": 8.98124885559082,
      "learning_rate": 4.703535353535354e-05,
      "loss": 0.3486,
      "step": 687
    },
    {
      "epoch": 0.0688,
      "grad_norm": 4.472506046295166,
      "learning_rate": 4.703030303030303e-05,
      "loss": 0.5444,
      "step": 688
    },
    {
      "epoch": 0.0689,
      "grad_norm": 9.48814582824707,
      "learning_rate": 4.702525252525253e-05,
      "loss": 1.3112,
      "step": 689
    },
    {
      "epoch": 0.069,
      "grad_norm": 11.752342224121094,
      "learning_rate": 4.702020202020202e-05,
      "loss": 0.4246,
      "step": 690
    },
    {
      "epoch": 0.0691,
      "grad_norm": 17.31199836730957,
      "learning_rate": 4.7015151515151516e-05,
      "loss": 0.2394,
      "step": 691
    },
    {
      "epoch": 0.0692,
      "grad_norm": 89.7054443359375,
      "learning_rate": 4.701010101010101e-05,
      "loss": 0.1399,
      "step": 692
    },
    {
      "epoch": 0.0693,
      "grad_norm": 3.6308093070983887,
      "learning_rate": 4.700505050505051e-05,
      "loss": 0.0972,
      "step": 693
    },
    {
      "epoch": 0.0694,
      "grad_norm": 10.235424995422363,
      "learning_rate": 4.7e-05,
      "loss": 0.4011,
      "step": 694
    },
    {
      "epoch": 0.0695,
      "grad_norm": 7.753368377685547,
      "learning_rate": 4.69949494949495e-05,
      "loss": 0.7507,
      "step": 695
    },
    {
      "epoch": 0.0696,
      "grad_norm": 9.682744026184082,
      "learning_rate": 4.698989898989899e-05,
      "loss": 0.4172,
      "step": 696
    },
    {
      "epoch": 0.0697,
      "grad_norm": 5.181023120880127,
      "learning_rate": 4.6984848484848485e-05,
      "loss": 0.1858,
      "step": 697
    },
    {
      "epoch": 0.0698,
      "grad_norm": 3.355999231338501,
      "learning_rate": 4.697979797979799e-05,
      "loss": 0.1182,
      "step": 698
    },
    {
      "epoch": 0.0699,
      "grad_norm": 31.387754440307617,
      "learning_rate": 4.697474747474748e-05,
      "loss": 0.2369,
      "step": 699
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.348132133483887,
      "learning_rate": 4.696969696969697e-05,
      "loss": 0.7568,
      "step": 700
    },
    {
      "epoch": 0.0701,
      "grad_norm": 9.773565292358398,
      "learning_rate": 4.696464646464647e-05,
      "loss": 0.7279,
      "step": 701
    },
    {
      "epoch": 0.0702,
      "grad_norm": 5.424749374389648,
      "learning_rate": 4.6959595959595965e-05,
      "loss": 0.2144,
      "step": 702
    },
    {
      "epoch": 0.0703,
      "grad_norm": 9.352611541748047,
      "learning_rate": 4.6954545454545454e-05,
      "loss": 0.6518,
      "step": 703
    },
    {
      "epoch": 0.0704,
      "grad_norm": 15.857589721679688,
      "learning_rate": 4.694949494949496e-05,
      "loss": 1.3922,
      "step": 704
    },
    {
      "epoch": 0.0705,
      "grad_norm": 6.549785614013672,
      "learning_rate": 4.6944444444444446e-05,
      "loss": 0.3476,
      "step": 705
    },
    {
      "epoch": 0.0706,
      "grad_norm": 23.261369705200195,
      "learning_rate": 4.693939393939394e-05,
      "loss": 0.7879,
      "step": 706
    },
    {
      "epoch": 0.0707,
      "grad_norm": 7.095074653625488,
      "learning_rate": 4.693434343434344e-05,
      "loss": 0.2666,
      "step": 707
    },
    {
      "epoch": 0.0708,
      "grad_norm": 7.623691082000732,
      "learning_rate": 4.6929292929292934e-05,
      "loss": 0.3642,
      "step": 708
    },
    {
      "epoch": 0.0709,
      "grad_norm": 7.598694324493408,
      "learning_rate": 4.6924242424242424e-05,
      "loss": 0.1486,
      "step": 709
    },
    {
      "epoch": 0.071,
      "grad_norm": 21.409324645996094,
      "learning_rate": 4.6919191919191926e-05,
      "loss": 0.1543,
      "step": 710
    },
    {
      "epoch": 0.0711,
      "grad_norm": 5.484747886657715,
      "learning_rate": 4.6914141414141416e-05,
      "loss": 0.6561,
      "step": 711
    },
    {
      "epoch": 0.0712,
      "grad_norm": 42.256866455078125,
      "learning_rate": 4.690909090909091e-05,
      "loss": 0.598,
      "step": 712
    },
    {
      "epoch": 0.0713,
      "grad_norm": 19.961584091186523,
      "learning_rate": 4.690404040404041e-05,
      "loss": 0.7561,
      "step": 713
    },
    {
      "epoch": 0.0714,
      "grad_norm": 8.112367630004883,
      "learning_rate": 4.6898989898989903e-05,
      "loss": 0.8252,
      "step": 714
    },
    {
      "epoch": 0.0715,
      "grad_norm": 10.9951171875,
      "learning_rate": 4.689393939393939e-05,
      "loss": 1.3281,
      "step": 715
    },
    {
      "epoch": 0.0716,
      "grad_norm": 11.957502365112305,
      "learning_rate": 4.6888888888888895e-05,
      "loss": 0.1443,
      "step": 716
    },
    {
      "epoch": 0.0717,
      "grad_norm": 9.994376182556152,
      "learning_rate": 4.6883838383838385e-05,
      "loss": 0.0654,
      "step": 717
    },
    {
      "epoch": 0.0718,
      "grad_norm": 4.544734001159668,
      "learning_rate": 4.687878787878788e-05,
      "loss": 0.5792,
      "step": 718
    },
    {
      "epoch": 0.0719,
      "grad_norm": 25.42876625061035,
      "learning_rate": 4.687373737373738e-05,
      "loss": 0.5857,
      "step": 719
    },
    {
      "epoch": 0.072,
      "grad_norm": 11.348254203796387,
      "learning_rate": 4.686868686868687e-05,
      "loss": 1.0385,
      "step": 720
    },
    {
      "epoch": 0.0721,
      "grad_norm": 8.52822208404541,
      "learning_rate": 4.686363636363636e-05,
      "loss": 0.3257,
      "step": 721
    },
    {
      "epoch": 0.0722,
      "grad_norm": 6.158051013946533,
      "learning_rate": 4.6858585858585865e-05,
      "loss": 0.3983,
      "step": 722
    },
    {
      "epoch": 0.0723,
      "grad_norm": 10.068512916564941,
      "learning_rate": 4.6853535353535354e-05,
      "loss": 0.1222,
      "step": 723
    },
    {
      "epoch": 0.0724,
      "grad_norm": 2.9226181507110596,
      "learning_rate": 4.684848484848485e-05,
      "loss": 0.4994,
      "step": 724
    },
    {
      "epoch": 0.0725,
      "grad_norm": 2.174476146697998,
      "learning_rate": 4.6843434343434346e-05,
      "loss": 0.0616,
      "step": 725
    },
    {
      "epoch": 0.0726,
      "grad_norm": 8.270079612731934,
      "learning_rate": 4.683838383838384e-05,
      "loss": 0.8031,
      "step": 726
    },
    {
      "epoch": 0.0727,
      "grad_norm": 11.392613410949707,
      "learning_rate": 4.683333333333334e-05,
      "loss": 0.3132,
      "step": 727
    },
    {
      "epoch": 0.0728,
      "grad_norm": 167.62567138671875,
      "learning_rate": 4.6828282828282834e-05,
      "loss": 0.3023,
      "step": 728
    },
    {
      "epoch": 0.0729,
      "grad_norm": 15.943645477294922,
      "learning_rate": 4.682323232323232e-05,
      "loss": 0.4041,
      "step": 729
    },
    {
      "epoch": 0.073,
      "grad_norm": 10.7491455078125,
      "learning_rate": 4.681818181818182e-05,
      "loss": 0.3188,
      "step": 730
    },
    {
      "epoch": 0.0731,
      "grad_norm": 5.413421630859375,
      "learning_rate": 4.6813131313131315e-05,
      "loss": 0.7286,
      "step": 731
    },
    {
      "epoch": 0.0732,
      "grad_norm": 6.692235946655273,
      "learning_rate": 4.680808080808081e-05,
      "loss": 0.0824,
      "step": 732
    },
    {
      "epoch": 0.0733,
      "grad_norm": 8.524211883544922,
      "learning_rate": 4.680303030303031e-05,
      "loss": 0.3482,
      "step": 733
    },
    {
      "epoch": 0.0734,
      "grad_norm": 10.18047046661377,
      "learning_rate": 4.67979797979798e-05,
      "loss": 0.3748,
      "step": 734
    },
    {
      "epoch": 0.0735,
      "grad_norm": 10.826290130615234,
      "learning_rate": 4.679292929292929e-05,
      "loss": 0.6384,
      "step": 735
    },
    {
      "epoch": 0.0736,
      "grad_norm": 8.935458183288574,
      "learning_rate": 4.6787878787878795e-05,
      "loss": 0.4909,
      "step": 736
    },
    {
      "epoch": 0.0737,
      "grad_norm": 26.529464721679688,
      "learning_rate": 4.6782828282828284e-05,
      "loss": 0.4468,
      "step": 737
    },
    {
      "epoch": 0.0738,
      "grad_norm": 5.679102420806885,
      "learning_rate": 4.677777777777778e-05,
      "loss": 0.6113,
      "step": 738
    },
    {
      "epoch": 0.0739,
      "grad_norm": 4.459352493286133,
      "learning_rate": 4.6772727272727276e-05,
      "loss": 0.109,
      "step": 739
    },
    {
      "epoch": 0.074,
      "grad_norm": 4.487341403961182,
      "learning_rate": 4.676767676767677e-05,
      "loss": 0.0972,
      "step": 740
    },
    {
      "epoch": 0.0741,
      "grad_norm": 14.358076095581055,
      "learning_rate": 4.676262626262626e-05,
      "loss": 0.9366,
      "step": 741
    },
    {
      "epoch": 0.0742,
      "grad_norm": 7.084780216217041,
      "learning_rate": 4.6757575757575764e-05,
      "loss": 0.3887,
      "step": 742
    },
    {
      "epoch": 0.0743,
      "grad_norm": 11.160077095031738,
      "learning_rate": 4.675252525252525e-05,
      "loss": 0.3986,
      "step": 743
    },
    {
      "epoch": 0.0744,
      "grad_norm": 59.25438690185547,
      "learning_rate": 4.674747474747475e-05,
      "loss": 0.9321,
      "step": 744
    },
    {
      "epoch": 0.0745,
      "grad_norm": 19.67375373840332,
      "learning_rate": 4.6742424242424245e-05,
      "loss": 0.4438,
      "step": 745
    },
    {
      "epoch": 0.0746,
      "grad_norm": 6.453327178955078,
      "learning_rate": 4.673737373737374e-05,
      "loss": 0.4201,
      "step": 746
    },
    {
      "epoch": 0.0747,
      "grad_norm": 20.192138671875,
      "learning_rate": 4.673232323232323e-05,
      "loss": 0.179,
      "step": 747
    },
    {
      "epoch": 0.0748,
      "grad_norm": 6.784066200256348,
      "learning_rate": 4.672727272727273e-05,
      "loss": 0.9114,
      "step": 748
    },
    {
      "epoch": 0.0749,
      "grad_norm": 2.7038817405700684,
      "learning_rate": 4.672222222222222e-05,
      "loss": 0.0737,
      "step": 749
    },
    {
      "epoch": 0.075,
      "grad_norm": 6.5852155685424805,
      "learning_rate": 4.671717171717172e-05,
      "loss": 0.4542,
      "step": 750
    },
    {
      "epoch": 0.0751,
      "grad_norm": 7.611302375793457,
      "learning_rate": 4.6712121212121214e-05,
      "loss": 0.4161,
      "step": 751
    },
    {
      "epoch": 0.0752,
      "grad_norm": 4.451181888580322,
      "learning_rate": 4.670707070707071e-05,
      "loss": 0.1472,
      "step": 752
    },
    {
      "epoch": 0.0753,
      "grad_norm": 11.076878547668457,
      "learning_rate": 4.67020202020202e-05,
      "loss": 0.2727,
      "step": 753
    },
    {
      "epoch": 0.0754,
      "grad_norm": 7.021873950958252,
      "learning_rate": 4.66969696969697e-05,
      "loss": 0.2698,
      "step": 754
    },
    {
      "epoch": 0.0755,
      "grad_norm": 6.479084491729736,
      "learning_rate": 4.669191919191919e-05,
      "loss": 0.1572,
      "step": 755
    },
    {
      "epoch": 0.0756,
      "grad_norm": 11.082287788391113,
      "learning_rate": 4.668686868686869e-05,
      "loss": 0.6095,
      "step": 756
    },
    {
      "epoch": 0.0757,
      "grad_norm": 23.908411026000977,
      "learning_rate": 4.6681818181818184e-05,
      "loss": 0.2006,
      "step": 757
    },
    {
      "epoch": 0.0758,
      "grad_norm": 13.192843437194824,
      "learning_rate": 4.667676767676768e-05,
      "loss": 0.8294,
      "step": 758
    },
    {
      "epoch": 0.0759,
      "grad_norm": 8.812220573425293,
      "learning_rate": 4.667171717171717e-05,
      "loss": 0.7161,
      "step": 759
    },
    {
      "epoch": 0.076,
      "grad_norm": 8.20221996307373,
      "learning_rate": 4.666666666666667e-05,
      "loss": 1.1038,
      "step": 760
    },
    {
      "epoch": 0.0761,
      "grad_norm": 5.019773960113525,
      "learning_rate": 4.666161616161616e-05,
      "loss": 0.8062,
      "step": 761
    },
    {
      "epoch": 0.0762,
      "grad_norm": 13.652153968811035,
      "learning_rate": 4.665656565656566e-05,
      "loss": 0.2762,
      "step": 762
    },
    {
      "epoch": 0.0763,
      "grad_norm": 7.196718215942383,
      "learning_rate": 4.665151515151515e-05,
      "loss": 0.1551,
      "step": 763
    },
    {
      "epoch": 0.0764,
      "grad_norm": 11.245386123657227,
      "learning_rate": 4.664646464646465e-05,
      "loss": 0.604,
      "step": 764
    },
    {
      "epoch": 0.0765,
      "grad_norm": 8.428146362304688,
      "learning_rate": 4.6641414141414145e-05,
      "loss": 0.9993,
      "step": 765
    },
    {
      "epoch": 0.0766,
      "grad_norm": 64.93014526367188,
      "learning_rate": 4.663636363636364e-05,
      "loss": 0.3076,
      "step": 766
    },
    {
      "epoch": 0.0767,
      "grad_norm": 7.382200717926025,
      "learning_rate": 4.663131313131313e-05,
      "loss": 0.2001,
      "step": 767
    },
    {
      "epoch": 0.0768,
      "grad_norm": 6.617574691772461,
      "learning_rate": 4.6626262626262626e-05,
      "loss": 0.3458,
      "step": 768
    },
    {
      "epoch": 0.0769,
      "grad_norm": 6.887572765350342,
      "learning_rate": 4.662121212121212e-05,
      "loss": 0.149,
      "step": 769
    },
    {
      "epoch": 0.077,
      "grad_norm": 4.664615631103516,
      "learning_rate": 4.661616161616162e-05,
      "loss": 0.1893,
      "step": 770
    },
    {
      "epoch": 0.0771,
      "grad_norm": 7.255354881286621,
      "learning_rate": 4.6611111111111114e-05,
      "loss": 0.5117,
      "step": 771
    },
    {
      "epoch": 0.0772,
      "grad_norm": 6.716618537902832,
      "learning_rate": 4.660606060606061e-05,
      "loss": 0.1452,
      "step": 772
    },
    {
      "epoch": 0.0773,
      "grad_norm": 24.375770568847656,
      "learning_rate": 4.66010101010101e-05,
      "loss": 0.8571,
      "step": 773
    },
    {
      "epoch": 0.0774,
      "grad_norm": 11.463606834411621,
      "learning_rate": 4.6595959595959595e-05,
      "loss": 0.3914,
      "step": 774
    },
    {
      "epoch": 0.0775,
      "grad_norm": 6.243062973022461,
      "learning_rate": 4.659090909090909e-05,
      "loss": 0.2765,
      "step": 775
    },
    {
      "epoch": 0.0776,
      "grad_norm": 13.712824821472168,
      "learning_rate": 4.658585858585859e-05,
      "loss": 0.5558,
      "step": 776
    },
    {
      "epoch": 0.0777,
      "grad_norm": 8.9948148727417,
      "learning_rate": 4.658080808080808e-05,
      "loss": 0.7895,
      "step": 777
    },
    {
      "epoch": 0.0778,
      "grad_norm": 11.676502227783203,
      "learning_rate": 4.657575757575758e-05,
      "loss": 0.946,
      "step": 778
    },
    {
      "epoch": 0.0779,
      "grad_norm": 13.762018203735352,
      "learning_rate": 4.657070707070707e-05,
      "loss": 0.3589,
      "step": 779
    },
    {
      "epoch": 0.078,
      "grad_norm": 6.962002754211426,
      "learning_rate": 4.656565656565657e-05,
      "loss": 0.2416,
      "step": 780
    },
    {
      "epoch": 0.0781,
      "grad_norm": 9.308279037475586,
      "learning_rate": 4.656060606060606e-05,
      "loss": 1.1139,
      "step": 781
    },
    {
      "epoch": 0.0782,
      "grad_norm": 6.2414231300354,
      "learning_rate": 4.6555555555555556e-05,
      "loss": 0.8344,
      "step": 782
    },
    {
      "epoch": 0.0783,
      "grad_norm": 9.760634422302246,
      "learning_rate": 4.655050505050505e-05,
      "loss": 0.5663,
      "step": 783
    },
    {
      "epoch": 0.0784,
      "grad_norm": 3.9554941654205322,
      "learning_rate": 4.654545454545455e-05,
      "loss": 0.2936,
      "step": 784
    },
    {
      "epoch": 0.0785,
      "grad_norm": 10.270462036132812,
      "learning_rate": 4.6540404040404044e-05,
      "loss": 0.1633,
      "step": 785
    },
    {
      "epoch": 0.0786,
      "grad_norm": 8.158556938171387,
      "learning_rate": 4.653535353535354e-05,
      "loss": 0.3595,
      "step": 786
    },
    {
      "epoch": 0.0787,
      "grad_norm": 5.659685134887695,
      "learning_rate": 4.6530303030303036e-05,
      "loss": 0.1063,
      "step": 787
    },
    {
      "epoch": 0.0788,
      "grad_norm": 5.968666076660156,
      "learning_rate": 4.6525252525252525e-05,
      "loss": 0.2716,
      "step": 788
    },
    {
      "epoch": 0.0789,
      "grad_norm": 13.935894966125488,
      "learning_rate": 4.652020202020203e-05,
      "loss": 0.6332,
      "step": 789
    },
    {
      "epoch": 0.079,
      "grad_norm": 13.683603286743164,
      "learning_rate": 4.651515151515152e-05,
      "loss": 0.3596,
      "step": 790
    },
    {
      "epoch": 0.0791,
      "grad_norm": 7.470770835876465,
      "learning_rate": 4.6510101010101013e-05,
      "loss": 0.1887,
      "step": 791
    },
    {
      "epoch": 0.0792,
      "grad_norm": 7.539287090301514,
      "learning_rate": 4.650505050505051e-05,
      "loss": 0.2136,
      "step": 792
    },
    {
      "epoch": 0.0793,
      "grad_norm": 28.53262710571289,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.8211,
      "step": 793
    },
    {
      "epoch": 0.0794,
      "grad_norm": 7.007681846618652,
      "learning_rate": 4.6494949494949495e-05,
      "loss": 0.2608,
      "step": 794
    },
    {
      "epoch": 0.0795,
      "grad_norm": 40.51974105834961,
      "learning_rate": 4.6489898989899e-05,
      "loss": 0.53,
      "step": 795
    },
    {
      "epoch": 0.0796,
      "grad_norm": 4.46160364151001,
      "learning_rate": 4.6484848484848487e-05,
      "loss": 0.1038,
      "step": 796
    },
    {
      "epoch": 0.0797,
      "grad_norm": 9.563056945800781,
      "learning_rate": 4.647979797979798e-05,
      "loss": 0.4582,
      "step": 797
    },
    {
      "epoch": 0.0798,
      "grad_norm": 10.842552185058594,
      "learning_rate": 4.647474747474748e-05,
      "loss": 0.2908,
      "step": 798
    },
    {
      "epoch": 0.0799,
      "grad_norm": 11.072558403015137,
      "learning_rate": 4.6469696969696975e-05,
      "loss": 0.5421,
      "step": 799
    },
    {
      "epoch": 0.08,
      "grad_norm": 6.61069917678833,
      "learning_rate": 4.6464646464646464e-05,
      "loss": 0.2228,
      "step": 800
    },
    {
      "epoch": 0.0801,
      "grad_norm": 10.150572776794434,
      "learning_rate": 4.6459595959595967e-05,
      "loss": 0.6927,
      "step": 801
    },
    {
      "epoch": 0.0802,
      "grad_norm": 17.814241409301758,
      "learning_rate": 4.6454545454545456e-05,
      "loss": 1.8196,
      "step": 802
    },
    {
      "epoch": 0.0803,
      "grad_norm": 11.95235538482666,
      "learning_rate": 4.644949494949495e-05,
      "loss": 0.6518,
      "step": 803
    },
    {
      "epoch": 0.0804,
      "grad_norm": 32.18625259399414,
      "learning_rate": 4.644444444444445e-05,
      "loss": 0.5419,
      "step": 804
    },
    {
      "epoch": 0.0805,
      "grad_norm": 13.166687965393066,
      "learning_rate": 4.6439393939393944e-05,
      "loss": 0.6816,
      "step": 805
    },
    {
      "epoch": 0.0806,
      "grad_norm": 6.814924716949463,
      "learning_rate": 4.643434343434343e-05,
      "loss": 0.1363,
      "step": 806
    },
    {
      "epoch": 0.0807,
      "grad_norm": 5.961161136627197,
      "learning_rate": 4.6429292929292936e-05,
      "loss": 0.1612,
      "step": 807
    },
    {
      "epoch": 0.0808,
      "grad_norm": 18.62491798400879,
      "learning_rate": 4.6424242424242425e-05,
      "loss": 0.947,
      "step": 808
    },
    {
      "epoch": 0.0809,
      "grad_norm": 15.02493953704834,
      "learning_rate": 4.641919191919192e-05,
      "loss": 0.125,
      "step": 809
    },
    {
      "epoch": 0.081,
      "grad_norm": 7.916865825653076,
      "learning_rate": 4.641414141414142e-05,
      "loss": 0.1478,
      "step": 810
    },
    {
      "epoch": 0.0811,
      "grad_norm": 7.595586776733398,
      "learning_rate": 4.640909090909091e-05,
      "loss": 0.437,
      "step": 811
    },
    {
      "epoch": 0.0812,
      "grad_norm": 9.839394569396973,
      "learning_rate": 4.64040404040404e-05,
      "loss": 0.89,
      "step": 812
    },
    {
      "epoch": 0.0813,
      "grad_norm": 9.488465309143066,
      "learning_rate": 4.6398989898989905e-05,
      "loss": 0.6115,
      "step": 813
    },
    {
      "epoch": 0.0814,
      "grad_norm": 9.0413818359375,
      "learning_rate": 4.6393939393939394e-05,
      "loss": 0.5661,
      "step": 814
    },
    {
      "epoch": 0.0815,
      "grad_norm": 8.749642372131348,
      "learning_rate": 4.638888888888889e-05,
      "loss": 0.4166,
      "step": 815
    },
    {
      "epoch": 0.0816,
      "grad_norm": 7.9378228187561035,
      "learning_rate": 4.6383838383838386e-05,
      "loss": 0.2146,
      "step": 816
    },
    {
      "epoch": 0.0817,
      "grad_norm": 7.307866096496582,
      "learning_rate": 4.637878787878788e-05,
      "loss": 0.9113,
      "step": 817
    },
    {
      "epoch": 0.0818,
      "grad_norm": 7.408169269561768,
      "learning_rate": 4.637373737373738e-05,
      "loss": 0.2913,
      "step": 818
    },
    {
      "epoch": 0.0819,
      "grad_norm": 3.1109116077423096,
      "learning_rate": 4.6368686868686874e-05,
      "loss": 0.1224,
      "step": 819
    },
    {
      "epoch": 0.082,
      "grad_norm": 3.8772356510162354,
      "learning_rate": 4.636363636363636e-05,
      "loss": 0.0662,
      "step": 820
    },
    {
      "epoch": 0.0821,
      "grad_norm": 9.783671379089355,
      "learning_rate": 4.635858585858586e-05,
      "loss": 0.4311,
      "step": 821
    },
    {
      "epoch": 0.0822,
      "grad_norm": 9.744071006774902,
      "learning_rate": 4.6353535353535355e-05,
      "loss": 0.0717,
      "step": 822
    },
    {
      "epoch": 0.0823,
      "grad_norm": 6.802423477172852,
      "learning_rate": 4.634848484848485e-05,
      "loss": 0.7831,
      "step": 823
    },
    {
      "epoch": 0.0824,
      "grad_norm": 7.148281097412109,
      "learning_rate": 4.634343434343435e-05,
      "loss": 1.1459,
      "step": 824
    },
    {
      "epoch": 0.0825,
      "grad_norm": 8.734697341918945,
      "learning_rate": 4.633838383838384e-05,
      "loss": 0.7342,
      "step": 825
    },
    {
      "epoch": 0.0826,
      "grad_norm": 4.75167179107666,
      "learning_rate": 4.633333333333333e-05,
      "loss": 0.163,
      "step": 826
    },
    {
      "epoch": 0.0827,
      "grad_norm": 8.761204719543457,
      "learning_rate": 4.6328282828282835e-05,
      "loss": 0.434,
      "step": 827
    },
    {
      "epoch": 0.0828,
      "grad_norm": 9.490363121032715,
      "learning_rate": 4.6323232323232324e-05,
      "loss": 0.3698,
      "step": 828
    },
    {
      "epoch": 0.0829,
      "grad_norm": 10.310127258300781,
      "learning_rate": 4.631818181818182e-05,
      "loss": 0.6811,
      "step": 829
    },
    {
      "epoch": 0.083,
      "grad_norm": 4.639791011810303,
      "learning_rate": 4.6313131313131316e-05,
      "loss": 0.065,
      "step": 830
    },
    {
      "epoch": 0.0831,
      "grad_norm": 8.464225769042969,
      "learning_rate": 4.630808080808081e-05,
      "loss": 0.5378,
      "step": 831
    },
    {
      "epoch": 0.0832,
      "grad_norm": 10.934346199035645,
      "learning_rate": 4.63030303030303e-05,
      "loss": 0.3056,
      "step": 832
    },
    {
      "epoch": 0.0833,
      "grad_norm": 5.811197280883789,
      "learning_rate": 4.6297979797979804e-05,
      "loss": 0.289,
      "step": 833
    },
    {
      "epoch": 0.0834,
      "grad_norm": 10.626136779785156,
      "learning_rate": 4.6292929292929294e-05,
      "loss": 0.8072,
      "step": 834
    },
    {
      "epoch": 0.0835,
      "grad_norm": 7.5146284103393555,
      "learning_rate": 4.628787878787879e-05,
      "loss": 0.1699,
      "step": 835
    },
    {
      "epoch": 0.0836,
      "grad_norm": 7.499359130859375,
      "learning_rate": 4.6282828282828286e-05,
      "loss": 0.5189,
      "step": 836
    },
    {
      "epoch": 0.0837,
      "grad_norm": 6.173251152038574,
      "learning_rate": 4.627777777777778e-05,
      "loss": 0.2117,
      "step": 837
    },
    {
      "epoch": 0.0838,
      "grad_norm": 5.570885181427002,
      "learning_rate": 4.627272727272727e-05,
      "loss": 0.2083,
      "step": 838
    },
    {
      "epoch": 0.0839,
      "grad_norm": 9.94780158996582,
      "learning_rate": 4.6267676767676774e-05,
      "loss": 0.3538,
      "step": 839
    },
    {
      "epoch": 0.084,
      "grad_norm": 2.95841121673584,
      "learning_rate": 4.626262626262626e-05,
      "loss": 0.1185,
      "step": 840
    },
    {
      "epoch": 0.0841,
      "grad_norm": 4.791739463806152,
      "learning_rate": 4.625757575757576e-05,
      "loss": 0.4348,
      "step": 841
    },
    {
      "epoch": 0.0842,
      "grad_norm": 2.917983055114746,
      "learning_rate": 4.6252525252525255e-05,
      "loss": 0.5639,
      "step": 842
    },
    {
      "epoch": 0.0843,
      "grad_norm": 10.25067138671875,
      "learning_rate": 4.624747474747475e-05,
      "loss": 0.4756,
      "step": 843
    },
    {
      "epoch": 0.0844,
      "grad_norm": 6.659273147583008,
      "learning_rate": 4.624242424242424e-05,
      "loss": 0.4624,
      "step": 844
    },
    {
      "epoch": 0.0845,
      "grad_norm": 7.313775062561035,
      "learning_rate": 4.623737373737374e-05,
      "loss": 0.2124,
      "step": 845
    },
    {
      "epoch": 0.0846,
      "grad_norm": 8.237592697143555,
      "learning_rate": 4.623232323232323e-05,
      "loss": 0.3514,
      "step": 846
    },
    {
      "epoch": 0.0847,
      "grad_norm": 9.92939567565918,
      "learning_rate": 4.622727272727273e-05,
      "loss": 0.195,
      "step": 847
    },
    {
      "epoch": 0.0848,
      "grad_norm": 8.48379135131836,
      "learning_rate": 4.6222222222222224e-05,
      "loss": 0.3916,
      "step": 848
    },
    {
      "epoch": 0.0849,
      "grad_norm": 6.645566940307617,
      "learning_rate": 4.621717171717172e-05,
      "loss": 0.5841,
      "step": 849
    },
    {
      "epoch": 0.085,
      "grad_norm": 15.608076095581055,
      "learning_rate": 4.621212121212121e-05,
      "loss": 0.6767,
      "step": 850
    },
    {
      "epoch": 0.0851,
      "grad_norm": 4.901000499725342,
      "learning_rate": 4.620707070707071e-05,
      "loss": 0.6087,
      "step": 851
    },
    {
      "epoch": 0.0852,
      "grad_norm": 13.747392654418945,
      "learning_rate": 4.62020202020202e-05,
      "loss": 0.3423,
      "step": 852
    },
    {
      "epoch": 0.0853,
      "grad_norm": 4.5394110679626465,
      "learning_rate": 4.61969696969697e-05,
      "loss": 0.6583,
      "step": 853
    },
    {
      "epoch": 0.0854,
      "grad_norm": 7.889669418334961,
      "learning_rate": 4.619191919191919e-05,
      "loss": 0.3442,
      "step": 854
    },
    {
      "epoch": 0.0855,
      "grad_norm": 11.735001564025879,
      "learning_rate": 4.618686868686869e-05,
      "loss": 1.0466,
      "step": 855
    },
    {
      "epoch": 0.0856,
      "grad_norm": 5.307590007781982,
      "learning_rate": 4.618181818181818e-05,
      "loss": 0.3218,
      "step": 856
    },
    {
      "epoch": 0.0857,
      "grad_norm": 8.648876190185547,
      "learning_rate": 4.617676767676768e-05,
      "loss": 0.7431,
      "step": 857
    },
    {
      "epoch": 0.0858,
      "grad_norm": 5.3228936195373535,
      "learning_rate": 4.617171717171717e-05,
      "loss": 0.3576,
      "step": 858
    },
    {
      "epoch": 0.0859,
      "grad_norm": 5.599617958068848,
      "learning_rate": 4.6166666666666666e-05,
      "loss": 0.5279,
      "step": 859
    },
    {
      "epoch": 0.086,
      "grad_norm": 4.471961498260498,
      "learning_rate": 4.616161616161616e-05,
      "loss": 0.1579,
      "step": 860
    },
    {
      "epoch": 0.0861,
      "grad_norm": 17.58660125732422,
      "learning_rate": 4.615656565656566e-05,
      "loss": 0.3717,
      "step": 861
    },
    {
      "epoch": 0.0862,
      "grad_norm": 3.9882967472076416,
      "learning_rate": 4.6151515151515154e-05,
      "loss": 0.0793,
      "step": 862
    },
    {
      "epoch": 0.0863,
      "grad_norm": 6.3740973472595215,
      "learning_rate": 4.614646464646465e-05,
      "loss": 0.1201,
      "step": 863
    },
    {
      "epoch": 0.0864,
      "grad_norm": 4.920657634735107,
      "learning_rate": 4.614141414141414e-05,
      "loss": 0.2156,
      "step": 864
    },
    {
      "epoch": 0.0865,
      "grad_norm": 9.198040008544922,
      "learning_rate": 4.6136363636363635e-05,
      "loss": 0.857,
      "step": 865
    },
    {
      "epoch": 0.0866,
      "grad_norm": 2.049314498901367,
      "learning_rate": 4.613131313131313e-05,
      "loss": 0.0382,
      "step": 866
    },
    {
      "epoch": 0.0867,
      "grad_norm": 10.62370491027832,
      "learning_rate": 4.612626262626263e-05,
      "loss": 0.4609,
      "step": 867
    },
    {
      "epoch": 0.0868,
      "grad_norm": 5.664398670196533,
      "learning_rate": 4.612121212121212e-05,
      "loss": 0.2681,
      "step": 868
    },
    {
      "epoch": 0.0869,
      "grad_norm": 5.17221736907959,
      "learning_rate": 4.611616161616162e-05,
      "loss": 0.2411,
      "step": 869
    },
    {
      "epoch": 0.087,
      "grad_norm": 11.490407943725586,
      "learning_rate": 4.6111111111111115e-05,
      "loss": 0.2023,
      "step": 870
    },
    {
      "epoch": 0.0871,
      "grad_norm": 14.976919174194336,
      "learning_rate": 4.610606060606061e-05,
      "loss": 0.8691,
      "step": 871
    },
    {
      "epoch": 0.0872,
      "grad_norm": 10.168030738830566,
      "learning_rate": 4.610101010101011e-05,
      "loss": 0.0441,
      "step": 872
    },
    {
      "epoch": 0.0873,
      "grad_norm": 6.1635050773620605,
      "learning_rate": 4.6095959595959597e-05,
      "loss": 0.314,
      "step": 873
    },
    {
      "epoch": 0.0874,
      "grad_norm": 29.25715446472168,
      "learning_rate": 4.609090909090909e-05,
      "loss": 0.7108,
      "step": 874
    },
    {
      "epoch": 0.0875,
      "grad_norm": 9.411470413208008,
      "learning_rate": 4.608585858585859e-05,
      "loss": 0.9901,
      "step": 875
    },
    {
      "epoch": 0.0876,
      "grad_norm": 6.977803707122803,
      "learning_rate": 4.6080808080808085e-05,
      "loss": 0.2301,
      "step": 876
    },
    {
      "epoch": 0.0877,
      "grad_norm": 17.31676483154297,
      "learning_rate": 4.607575757575758e-05,
      "loss": 0.5699,
      "step": 877
    },
    {
      "epoch": 0.0878,
      "grad_norm": 7.8542704582214355,
      "learning_rate": 4.6070707070707077e-05,
      "loss": 0.3534,
      "step": 878
    },
    {
      "epoch": 0.0879,
      "grad_norm": 12.752866744995117,
      "learning_rate": 4.6065656565656566e-05,
      "loss": 0.4414,
      "step": 879
    },
    {
      "epoch": 0.088,
      "grad_norm": 9.152913093566895,
      "learning_rate": 4.606060606060607e-05,
      "loss": 0.4156,
      "step": 880
    },
    {
      "epoch": 0.0881,
      "grad_norm": 4.936456680297852,
      "learning_rate": 4.605555555555556e-05,
      "loss": 0.2453,
      "step": 881
    },
    {
      "epoch": 0.0882,
      "grad_norm": 4.874292373657227,
      "learning_rate": 4.6050505050505054e-05,
      "loss": 0.5912,
      "step": 882
    },
    {
      "epoch": 0.0883,
      "grad_norm": 6.6156535148620605,
      "learning_rate": 4.604545454545455e-05,
      "loss": 0.1882,
      "step": 883
    },
    {
      "epoch": 0.0884,
      "grad_norm": 10.35744857788086,
      "learning_rate": 4.6040404040404046e-05,
      "loss": 0.8125,
      "step": 884
    },
    {
      "epoch": 0.0885,
      "grad_norm": 15.084107398986816,
      "learning_rate": 4.6035353535353535e-05,
      "loss": 0.9432,
      "step": 885
    },
    {
      "epoch": 0.0886,
      "grad_norm": 3.3917102813720703,
      "learning_rate": 4.603030303030304e-05,
      "loss": 0.1314,
      "step": 886
    },
    {
      "epoch": 0.0887,
      "grad_norm": 29.50817108154297,
      "learning_rate": 4.602525252525253e-05,
      "loss": 0.1771,
      "step": 887
    },
    {
      "epoch": 0.0888,
      "grad_norm": 4.626336097717285,
      "learning_rate": 4.602020202020202e-05,
      "loss": 0.2482,
      "step": 888
    },
    {
      "epoch": 0.0889,
      "grad_norm": 13.392169952392578,
      "learning_rate": 4.601515151515152e-05,
      "loss": 0.3322,
      "step": 889
    },
    {
      "epoch": 0.089,
      "grad_norm": 3.4384727478027344,
      "learning_rate": 4.6010101010101015e-05,
      "loss": 0.1133,
      "step": 890
    },
    {
      "epoch": 0.0891,
      "grad_norm": 1.8895155191421509,
      "learning_rate": 4.6005050505050504e-05,
      "loss": 0.0392,
      "step": 891
    },
    {
      "epoch": 0.0892,
      "grad_norm": 8.26074504852295,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1292,
      "step": 892
    },
    {
      "epoch": 0.0893,
      "grad_norm": 8.09462833404541,
      "learning_rate": 4.5994949494949496e-05,
      "loss": 0.1567,
      "step": 893
    },
    {
      "epoch": 0.0894,
      "grad_norm": 5.065523147583008,
      "learning_rate": 4.598989898989899e-05,
      "loss": 0.5747,
      "step": 894
    },
    {
      "epoch": 0.0895,
      "grad_norm": 8.600361824035645,
      "learning_rate": 4.598484848484849e-05,
      "loss": 0.7249,
      "step": 895
    },
    {
      "epoch": 0.0896,
      "grad_norm": 7.619458198547363,
      "learning_rate": 4.5979797979797984e-05,
      "loss": 0.3187,
      "step": 896
    },
    {
      "epoch": 0.0897,
      "grad_norm": 5.107491970062256,
      "learning_rate": 4.597474747474747e-05,
      "loss": 0.1842,
      "step": 897
    },
    {
      "epoch": 0.0898,
      "grad_norm": 8.751899719238281,
      "learning_rate": 4.5969696969696976e-05,
      "loss": 0.4486,
      "step": 898
    },
    {
      "epoch": 0.0899,
      "grad_norm": 3.2890682220458984,
      "learning_rate": 4.5964646464646465e-05,
      "loss": 0.5197,
      "step": 899
    },
    {
      "epoch": 0.09,
      "grad_norm": 13.421348571777344,
      "learning_rate": 4.595959595959596e-05,
      "loss": 0.9067,
      "step": 900
    },
    {
      "epoch": 0.0901,
      "grad_norm": 13.2793550491333,
      "learning_rate": 4.595454545454546e-05,
      "loss": 0.5909,
      "step": 901
    },
    {
      "epoch": 0.0902,
      "grad_norm": 12.666945457458496,
      "learning_rate": 4.594949494949495e-05,
      "loss": 0.3497,
      "step": 902
    },
    {
      "epoch": 0.0903,
      "grad_norm": 9.584513664245605,
      "learning_rate": 4.594444444444444e-05,
      "loss": 0.1004,
      "step": 903
    },
    {
      "epoch": 0.0904,
      "grad_norm": 6.409865379333496,
      "learning_rate": 4.5939393939393945e-05,
      "loss": 0.264,
      "step": 904
    },
    {
      "epoch": 0.0905,
      "grad_norm": 7.714510917663574,
      "learning_rate": 4.5934343434343434e-05,
      "loss": 0.7454,
      "step": 905
    },
    {
      "epoch": 0.0906,
      "grad_norm": 3.8850884437561035,
      "learning_rate": 4.592929292929293e-05,
      "loss": 0.1674,
      "step": 906
    },
    {
      "epoch": 0.0907,
      "grad_norm": 5.405892372131348,
      "learning_rate": 4.5924242424242426e-05,
      "loss": 0.1198,
      "step": 907
    },
    {
      "epoch": 0.0908,
      "grad_norm": 4.7098164558410645,
      "learning_rate": 4.591919191919192e-05,
      "loss": 0.1911,
      "step": 908
    },
    {
      "epoch": 0.0909,
      "grad_norm": 8.265875816345215,
      "learning_rate": 4.591414141414142e-05,
      "loss": 0.6159,
      "step": 909
    },
    {
      "epoch": 0.091,
      "grad_norm": 6.5548014640808105,
      "learning_rate": 4.5909090909090914e-05,
      "loss": 0.8189,
      "step": 910
    },
    {
      "epoch": 0.0911,
      "grad_norm": 35.42535400390625,
      "learning_rate": 4.5904040404040404e-05,
      "loss": 0.1774,
      "step": 911
    },
    {
      "epoch": 0.0912,
      "grad_norm": 5.9405107498168945,
      "learning_rate": 4.58989898989899e-05,
      "loss": 0.7115,
      "step": 912
    },
    {
      "epoch": 0.0913,
      "grad_norm": 5.822568416595459,
      "learning_rate": 4.5893939393939396e-05,
      "loss": 0.3534,
      "step": 913
    },
    {
      "epoch": 0.0914,
      "grad_norm": 5.529414653778076,
      "learning_rate": 4.588888888888889e-05,
      "loss": 0.3452,
      "step": 914
    },
    {
      "epoch": 0.0915,
      "grad_norm": 8.42989730834961,
      "learning_rate": 4.588383838383839e-05,
      "loss": 1.0339,
      "step": 915
    },
    {
      "epoch": 0.0916,
      "grad_norm": 11.330094337463379,
      "learning_rate": 4.5878787878787883e-05,
      "loss": 0.4238,
      "step": 916
    },
    {
      "epoch": 0.0917,
      "grad_norm": 6.355686187744141,
      "learning_rate": 4.587373737373737e-05,
      "loss": 0.3277,
      "step": 917
    },
    {
      "epoch": 0.0918,
      "grad_norm": 4.294715404510498,
      "learning_rate": 4.586868686868687e-05,
      "loss": 0.1103,
      "step": 918
    },
    {
      "epoch": 0.0919,
      "grad_norm": 10.160812377929688,
      "learning_rate": 4.5863636363636365e-05,
      "loss": 0.8742,
      "step": 919
    },
    {
      "epoch": 0.092,
      "grad_norm": 4.998605251312256,
      "learning_rate": 4.585858585858586e-05,
      "loss": 0.4403,
      "step": 920
    },
    {
      "epoch": 0.0921,
      "grad_norm": 8.2843656539917,
      "learning_rate": 4.585353535353536e-05,
      "loss": 0.1755,
      "step": 921
    },
    {
      "epoch": 0.0922,
      "grad_norm": 7.140223979949951,
      "learning_rate": 4.584848484848485e-05,
      "loss": 0.1372,
      "step": 922
    },
    {
      "epoch": 0.0923,
      "grad_norm": 1.8666291236877441,
      "learning_rate": 4.584343434343434e-05,
      "loss": 0.0569,
      "step": 923
    },
    {
      "epoch": 0.0924,
      "grad_norm": 25.738466262817383,
      "learning_rate": 4.5838383838383845e-05,
      "loss": 0.8808,
      "step": 924
    },
    {
      "epoch": 0.0925,
      "grad_norm": 4.348135471343994,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.1596,
      "step": 925
    },
    {
      "epoch": 0.0926,
      "grad_norm": 6.538527488708496,
      "learning_rate": 4.582828282828283e-05,
      "loss": 0.442,
      "step": 926
    },
    {
      "epoch": 0.0927,
      "grad_norm": 10.779641151428223,
      "learning_rate": 4.5823232323232326e-05,
      "loss": 0.7355,
      "step": 927
    },
    {
      "epoch": 0.0928,
      "grad_norm": 6.927086353302002,
      "learning_rate": 4.581818181818182e-05,
      "loss": 0.7278,
      "step": 928
    },
    {
      "epoch": 0.0929,
      "grad_norm": 6.171258926391602,
      "learning_rate": 4.581313131313131e-05,
      "loss": 0.4189,
      "step": 929
    },
    {
      "epoch": 0.093,
      "grad_norm": 7.647627353668213,
      "learning_rate": 4.5808080808080814e-05,
      "loss": 0.2922,
      "step": 930
    },
    {
      "epoch": 0.0931,
      "grad_norm": 4.289402484893799,
      "learning_rate": 4.58030303030303e-05,
      "loss": 0.1045,
      "step": 931
    },
    {
      "epoch": 0.0932,
      "grad_norm": 16.702116012573242,
      "learning_rate": 4.57979797979798e-05,
      "loss": 0.1824,
      "step": 932
    },
    {
      "epoch": 0.0933,
      "grad_norm": 2.9417214393615723,
      "learning_rate": 4.5792929292929295e-05,
      "loss": 0.5479,
      "step": 933
    },
    {
      "epoch": 0.0934,
      "grad_norm": 7.826621055603027,
      "learning_rate": 4.578787878787879e-05,
      "loss": 0.1396,
      "step": 934
    },
    {
      "epoch": 0.0935,
      "grad_norm": 3.2560505867004395,
      "learning_rate": 4.578282828282828e-05,
      "loss": 0.2731,
      "step": 935
    },
    {
      "epoch": 0.0936,
      "grad_norm": 33.456790924072266,
      "learning_rate": 4.577777777777778e-05,
      "loss": 0.4965,
      "step": 936
    },
    {
      "epoch": 0.0937,
      "grad_norm": 6.553860664367676,
      "learning_rate": 4.577272727272727e-05,
      "loss": 0.859,
      "step": 937
    },
    {
      "epoch": 0.0938,
      "grad_norm": 4.773195266723633,
      "learning_rate": 4.576767676767677e-05,
      "loss": 0.1426,
      "step": 938
    },
    {
      "epoch": 0.0939,
      "grad_norm": 11.289615631103516,
      "learning_rate": 4.5762626262626264e-05,
      "loss": 0.778,
      "step": 939
    },
    {
      "epoch": 0.094,
      "grad_norm": 7.001153945922852,
      "learning_rate": 4.575757575757576e-05,
      "loss": 0.2106,
      "step": 940
    },
    {
      "epoch": 0.0941,
      "grad_norm": 7.042220115661621,
      "learning_rate": 4.575252525252525e-05,
      "loss": 0.3976,
      "step": 941
    },
    {
      "epoch": 0.0942,
      "grad_norm": 7.495244979858398,
      "learning_rate": 4.574747474747475e-05,
      "loss": 0.6659,
      "step": 942
    },
    {
      "epoch": 0.0943,
      "grad_norm": 9.919684410095215,
      "learning_rate": 4.574242424242424e-05,
      "loss": 0.1637,
      "step": 943
    },
    {
      "epoch": 0.0944,
      "grad_norm": 21.829923629760742,
      "learning_rate": 4.573737373737374e-05,
      "loss": 0.2163,
      "step": 944
    },
    {
      "epoch": 0.0945,
      "grad_norm": 8.767940521240234,
      "learning_rate": 4.573232323232323e-05,
      "loss": 0.0732,
      "step": 945
    },
    {
      "epoch": 0.0946,
      "grad_norm": 9.799395561218262,
      "learning_rate": 4.572727272727273e-05,
      "loss": 0.3088,
      "step": 946
    },
    {
      "epoch": 0.0947,
      "grad_norm": 12.307817459106445,
      "learning_rate": 4.572222222222222e-05,
      "loss": 0.1376,
      "step": 947
    },
    {
      "epoch": 0.0948,
      "grad_norm": 5.993952751159668,
      "learning_rate": 4.571717171717172e-05,
      "loss": 0.6639,
      "step": 948
    },
    {
      "epoch": 0.0949,
      "grad_norm": 4.988731861114502,
      "learning_rate": 4.571212121212121e-05,
      "loss": 0.7027,
      "step": 949
    },
    {
      "epoch": 0.095,
      "grad_norm": 9.995146751403809,
      "learning_rate": 4.5707070707070706e-05,
      "loss": 0.6789,
      "step": 950
    },
    {
      "epoch": 0.0951,
      "grad_norm": 5.945037364959717,
      "learning_rate": 4.57020202020202e-05,
      "loss": 0.3538,
      "step": 951
    },
    {
      "epoch": 0.0952,
      "grad_norm": 12.200356483459473,
      "learning_rate": 4.56969696969697e-05,
      "loss": 0.501,
      "step": 952
    },
    {
      "epoch": 0.0953,
      "grad_norm": 5.317443370819092,
      "learning_rate": 4.5691919191919194e-05,
      "loss": 0.1576,
      "step": 953
    },
    {
      "epoch": 0.0954,
      "grad_norm": 3.020796775817871,
      "learning_rate": 4.568686868686869e-05,
      "loss": 0.1269,
      "step": 954
    },
    {
      "epoch": 0.0955,
      "grad_norm": 5.694932460784912,
      "learning_rate": 4.5681818181818186e-05,
      "loss": 0.24,
      "step": 955
    },
    {
      "epoch": 0.0956,
      "grad_norm": 8.580689430236816,
      "learning_rate": 4.5676767676767676e-05,
      "loss": 0.5069,
      "step": 956
    },
    {
      "epoch": 0.0957,
      "grad_norm": 5.759493827819824,
      "learning_rate": 4.567171717171718e-05,
      "loss": 0.445,
      "step": 957
    },
    {
      "epoch": 0.0958,
      "grad_norm": 4.957225322723389,
      "learning_rate": 4.566666666666667e-05,
      "loss": 0.5382,
      "step": 958
    },
    {
      "epoch": 0.0959,
      "grad_norm": 4.369378089904785,
      "learning_rate": 4.5661616161616164e-05,
      "loss": 0.1068,
      "step": 959
    },
    {
      "epoch": 0.096,
      "grad_norm": 5.830442428588867,
      "learning_rate": 4.565656565656566e-05,
      "loss": 0.3576,
      "step": 960
    },
    {
      "epoch": 0.0961,
      "grad_norm": 6.699178695678711,
      "learning_rate": 4.5651515151515156e-05,
      "loss": 0.4354,
      "step": 961
    },
    {
      "epoch": 0.0962,
      "grad_norm": 4.652170658111572,
      "learning_rate": 4.564646464646465e-05,
      "loss": 0.2335,
      "step": 962
    },
    {
      "epoch": 0.0963,
      "grad_norm": 3.1104087829589844,
      "learning_rate": 4.564141414141415e-05,
      "loss": 0.069,
      "step": 963
    },
    {
      "epoch": 0.0964,
      "grad_norm": 4.308261394500732,
      "learning_rate": 4.563636363636364e-05,
      "loss": 0.0581,
      "step": 964
    },
    {
      "epoch": 0.0965,
      "grad_norm": 6.639202117919922,
      "learning_rate": 4.563131313131313e-05,
      "loss": 0.2515,
      "step": 965
    },
    {
      "epoch": 0.0966,
      "grad_norm": 5.714690685272217,
      "learning_rate": 4.562626262626263e-05,
      "loss": 0.3071,
      "step": 966
    },
    {
      "epoch": 0.0967,
      "grad_norm": 4.454766273498535,
      "learning_rate": 4.5621212121212125e-05,
      "loss": 0.2059,
      "step": 967
    },
    {
      "epoch": 0.0968,
      "grad_norm": 4.767248630523682,
      "learning_rate": 4.561616161616162e-05,
      "loss": 0.1358,
      "step": 968
    },
    {
      "epoch": 0.0969,
      "grad_norm": 5.7643561363220215,
      "learning_rate": 4.561111111111112e-05,
      "loss": 0.3042,
      "step": 969
    },
    {
      "epoch": 0.097,
      "grad_norm": 16.4570255279541,
      "learning_rate": 4.5606060606060606e-05,
      "loss": 0.7744,
      "step": 970
    },
    {
      "epoch": 0.0971,
      "grad_norm": 8.21341609954834,
      "learning_rate": 4.560101010101011e-05,
      "loss": 0.1138,
      "step": 971
    },
    {
      "epoch": 0.0972,
      "grad_norm": 3.894282102584839,
      "learning_rate": 4.55959595959596e-05,
      "loss": 0.0995,
      "step": 972
    },
    {
      "epoch": 0.0973,
      "grad_norm": 16.011388778686523,
      "learning_rate": 4.5590909090909094e-05,
      "loss": 0.3244,
      "step": 973
    },
    {
      "epoch": 0.0974,
      "grad_norm": 3.7134029865264893,
      "learning_rate": 4.558585858585859e-05,
      "loss": 0.0801,
      "step": 974
    },
    {
      "epoch": 0.0975,
      "grad_norm": 4.8818888664245605,
      "learning_rate": 4.5580808080808086e-05,
      "loss": 1.0962,
      "step": 975
    },
    {
      "epoch": 0.0976,
      "grad_norm": 7.127344131469727,
      "learning_rate": 4.5575757575757575e-05,
      "loss": 0.4646,
      "step": 976
    },
    {
      "epoch": 0.0977,
      "grad_norm": 11.484536170959473,
      "learning_rate": 4.557070707070708e-05,
      "loss": 0.7825,
      "step": 977
    },
    {
      "epoch": 0.0978,
      "grad_norm": 24.641904830932617,
      "learning_rate": 4.556565656565657e-05,
      "loss": 0.7697,
      "step": 978
    },
    {
      "epoch": 0.0979,
      "grad_norm": 3.091977834701538,
      "learning_rate": 4.556060606060606e-05,
      "loss": 0.0811,
      "step": 979
    },
    {
      "epoch": 0.098,
      "grad_norm": 8.114965438842773,
      "learning_rate": 4.555555555555556e-05,
      "loss": 0.5016,
      "step": 980
    },
    {
      "epoch": 0.0981,
      "grad_norm": 10.134278297424316,
      "learning_rate": 4.5550505050505055e-05,
      "loss": 1.0426,
      "step": 981
    },
    {
      "epoch": 0.0982,
      "grad_norm": 9.804427146911621,
      "learning_rate": 4.5545454545454544e-05,
      "loss": 0.1939,
      "step": 982
    },
    {
      "epoch": 0.0983,
      "grad_norm": 5.553919315338135,
      "learning_rate": 4.554040404040405e-05,
      "loss": 0.1507,
      "step": 983
    },
    {
      "epoch": 0.0984,
      "grad_norm": 6.762911796569824,
      "learning_rate": 4.5535353535353536e-05,
      "loss": 0.3409,
      "step": 984
    },
    {
      "epoch": 0.0985,
      "grad_norm": 10.344657897949219,
      "learning_rate": 4.553030303030303e-05,
      "loss": 0.226,
      "step": 985
    },
    {
      "epoch": 0.0986,
      "grad_norm": 7.25610876083374,
      "learning_rate": 4.552525252525253e-05,
      "loss": 0.8518,
      "step": 986
    },
    {
      "epoch": 0.0987,
      "grad_norm": 7.446904182434082,
      "learning_rate": 4.5520202020202024e-05,
      "loss": 0.1748,
      "step": 987
    },
    {
      "epoch": 0.0988,
      "grad_norm": 6.528019905090332,
      "learning_rate": 4.5515151515151513e-05,
      "loss": 0.2939,
      "step": 988
    },
    {
      "epoch": 0.0989,
      "grad_norm": 7.952997207641602,
      "learning_rate": 4.5510101010101016e-05,
      "loss": 0.6744,
      "step": 989
    },
    {
      "epoch": 0.099,
      "grad_norm": 3.7128865718841553,
      "learning_rate": 4.5505050505050505e-05,
      "loss": 0.268,
      "step": 990
    },
    {
      "epoch": 0.0991,
      "grad_norm": 2.1966121196746826,
      "learning_rate": 4.55e-05,
      "loss": 0.0559,
      "step": 991
    },
    {
      "epoch": 0.0992,
      "grad_norm": 5.3280839920043945,
      "learning_rate": 4.54949494949495e-05,
      "loss": 0.1065,
      "step": 992
    },
    {
      "epoch": 0.0993,
      "grad_norm": 3.6213219165802,
      "learning_rate": 4.5489898989898993e-05,
      "loss": 0.2988,
      "step": 993
    },
    {
      "epoch": 0.0994,
      "grad_norm": 8.313054084777832,
      "learning_rate": 4.548484848484848e-05,
      "loss": 0.687,
      "step": 994
    },
    {
      "epoch": 0.0995,
      "grad_norm": 2.6076388359069824,
      "learning_rate": 4.5479797979797985e-05,
      "loss": 0.0369,
      "step": 995
    },
    {
      "epoch": 0.0996,
      "grad_norm": 5.1254191398620605,
      "learning_rate": 4.5474747474747475e-05,
      "loss": 0.1965,
      "step": 996
    },
    {
      "epoch": 0.0997,
      "grad_norm": 6.859347343444824,
      "learning_rate": 4.546969696969697e-05,
      "loss": 0.5297,
      "step": 997
    },
    {
      "epoch": 0.0998,
      "grad_norm": 5.358429908752441,
      "learning_rate": 4.5464646464646467e-05,
      "loss": 0.1465,
      "step": 998
    },
    {
      "epoch": 0.0999,
      "grad_norm": 6.72743034362793,
      "learning_rate": 4.545959595959596e-05,
      "loss": 0.4199,
      "step": 999
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.7050514221191406,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0678,
      "step": 1000
    },
    {
      "epoch": 0.1001,
      "grad_norm": 2.7966434955596924,
      "learning_rate": 4.5449494949494955e-05,
      "loss": 0.0697,
      "step": 1001
    },
    {
      "epoch": 0.1002,
      "grad_norm": 4.801110744476318,
      "learning_rate": 4.5444444444444444e-05,
      "loss": 0.1471,
      "step": 1002
    },
    {
      "epoch": 0.1003,
      "grad_norm": 8.521698951721191,
      "learning_rate": 4.543939393939394e-05,
      "loss": 0.5703,
      "step": 1003
    },
    {
      "epoch": 0.1004,
      "grad_norm": 8.462376594543457,
      "learning_rate": 4.5434343434343436e-05,
      "loss": 0.6075,
      "step": 1004
    },
    {
      "epoch": 0.1005,
      "grad_norm": 6.338038444519043,
      "learning_rate": 4.542929292929293e-05,
      "loss": 1.1665,
      "step": 1005
    },
    {
      "epoch": 0.1006,
      "grad_norm": 15.578302383422852,
      "learning_rate": 4.542424242424243e-05,
      "loss": 0.6203,
      "step": 1006
    },
    {
      "epoch": 0.1007,
      "grad_norm": 5.346531867980957,
      "learning_rate": 4.5419191919191924e-05,
      "loss": 0.1176,
      "step": 1007
    },
    {
      "epoch": 0.1008,
      "grad_norm": 3.769832134246826,
      "learning_rate": 4.541414141414141e-05,
      "loss": 0.1227,
      "step": 1008
    },
    {
      "epoch": 0.1009,
      "grad_norm": 20.163101196289062,
      "learning_rate": 4.540909090909091e-05,
      "loss": 0.2068,
      "step": 1009
    },
    {
      "epoch": 0.101,
      "grad_norm": 3.133239507675171,
      "learning_rate": 4.5404040404040405e-05,
      "loss": 0.0627,
      "step": 1010
    },
    {
      "epoch": 0.1011,
      "grad_norm": 8.049696922302246,
      "learning_rate": 4.53989898989899e-05,
      "loss": 0.4749,
      "step": 1011
    },
    {
      "epoch": 0.1012,
      "grad_norm": 3.262397527694702,
      "learning_rate": 4.53939393939394e-05,
      "loss": 0.0654,
      "step": 1012
    },
    {
      "epoch": 0.1013,
      "grad_norm": 9.071566581726074,
      "learning_rate": 4.538888888888889e-05,
      "loss": 0.2434,
      "step": 1013
    },
    {
      "epoch": 0.1014,
      "grad_norm": 14.909294128417969,
      "learning_rate": 4.538383838383838e-05,
      "loss": 0.4829,
      "step": 1014
    },
    {
      "epoch": 0.1015,
      "grad_norm": 6.807608604431152,
      "learning_rate": 4.5378787878787885e-05,
      "loss": 0.5337,
      "step": 1015
    },
    {
      "epoch": 0.1016,
      "grad_norm": 7.6401753425598145,
      "learning_rate": 4.5373737373737374e-05,
      "loss": 0.3835,
      "step": 1016
    },
    {
      "epoch": 0.1017,
      "grad_norm": 2.410707950592041,
      "learning_rate": 4.536868686868687e-05,
      "loss": 0.5289,
      "step": 1017
    },
    {
      "epoch": 0.1018,
      "grad_norm": 6.805196762084961,
      "learning_rate": 4.5363636363636366e-05,
      "loss": 0.2549,
      "step": 1018
    },
    {
      "epoch": 0.1019,
      "grad_norm": 8.602869987487793,
      "learning_rate": 4.535858585858586e-05,
      "loss": 0.2672,
      "step": 1019
    },
    {
      "epoch": 0.102,
      "grad_norm": 4.515635967254639,
      "learning_rate": 4.535353535353535e-05,
      "loss": 0.3745,
      "step": 1020
    },
    {
      "epoch": 0.1021,
      "grad_norm": 10.757957458496094,
      "learning_rate": 4.5348484848484854e-05,
      "loss": 0.2787,
      "step": 1021
    },
    {
      "epoch": 0.1022,
      "grad_norm": 7.765588283538818,
      "learning_rate": 4.534343434343434e-05,
      "loss": 0.2231,
      "step": 1022
    },
    {
      "epoch": 0.1023,
      "grad_norm": 3.89194917678833,
      "learning_rate": 4.533838383838384e-05,
      "loss": 0.7192,
      "step": 1023
    },
    {
      "epoch": 0.1024,
      "grad_norm": 12.474686622619629,
      "learning_rate": 4.5333333333333335e-05,
      "loss": 0.8825,
      "step": 1024
    },
    {
      "epoch": 0.1025,
      "grad_norm": 7.3239545822143555,
      "learning_rate": 4.532828282828283e-05,
      "loss": 0.3418,
      "step": 1025
    },
    {
      "epoch": 0.1026,
      "grad_norm": 2.7193379402160645,
      "learning_rate": 4.532323232323232e-05,
      "loss": 0.0403,
      "step": 1026
    },
    {
      "epoch": 0.1027,
      "grad_norm": 16.332971572875977,
      "learning_rate": 4.531818181818182e-05,
      "loss": 0.6075,
      "step": 1027
    },
    {
      "epoch": 0.1028,
      "grad_norm": 14.829005241394043,
      "learning_rate": 4.531313131313131e-05,
      "loss": 0.5648,
      "step": 1028
    },
    {
      "epoch": 0.1029,
      "grad_norm": 11.664813041687012,
      "learning_rate": 4.530808080808081e-05,
      "loss": 0.7446,
      "step": 1029
    },
    {
      "epoch": 0.103,
      "grad_norm": 3.9862895011901855,
      "learning_rate": 4.5303030303030304e-05,
      "loss": 0.1807,
      "step": 1030
    },
    {
      "epoch": 0.1031,
      "grad_norm": 11.110523223876953,
      "learning_rate": 4.52979797979798e-05,
      "loss": 0.1911,
      "step": 1031
    },
    {
      "epoch": 0.1032,
      "grad_norm": 7.116184234619141,
      "learning_rate": 4.529292929292929e-05,
      "loss": 0.2981,
      "step": 1032
    },
    {
      "epoch": 0.1033,
      "grad_norm": 6.691664218902588,
      "learning_rate": 4.528787878787879e-05,
      "loss": 0.734,
      "step": 1033
    },
    {
      "epoch": 0.1034,
      "grad_norm": 12.140087127685547,
      "learning_rate": 4.528282828282828e-05,
      "loss": 0.4822,
      "step": 1034
    },
    {
      "epoch": 0.1035,
      "grad_norm": 6.690937519073486,
      "learning_rate": 4.527777777777778e-05,
      "loss": 1.029,
      "step": 1035
    },
    {
      "epoch": 0.1036,
      "grad_norm": 4.175722122192383,
      "learning_rate": 4.5272727272727274e-05,
      "loss": 0.1482,
      "step": 1036
    },
    {
      "epoch": 0.1037,
      "grad_norm": 12.56698226928711,
      "learning_rate": 4.526767676767677e-05,
      "loss": 0.4227,
      "step": 1037
    },
    {
      "epoch": 0.1038,
      "grad_norm": 12.867484092712402,
      "learning_rate": 4.526262626262626e-05,
      "loss": 0.8536,
      "step": 1038
    },
    {
      "epoch": 0.1039,
      "grad_norm": 20.446142196655273,
      "learning_rate": 4.525757575757576e-05,
      "loss": 0.2251,
      "step": 1039
    },
    {
      "epoch": 0.104,
      "grad_norm": 15.488157272338867,
      "learning_rate": 4.525252525252526e-05,
      "loss": 0.6445,
      "step": 1040
    },
    {
      "epoch": 0.1041,
      "grad_norm": 5.051919460296631,
      "learning_rate": 4.524747474747475e-05,
      "loss": 0.5766,
      "step": 1041
    },
    {
      "epoch": 0.1042,
      "grad_norm": 6.564742088317871,
      "learning_rate": 4.524242424242425e-05,
      "loss": 0.3409,
      "step": 1042
    },
    {
      "epoch": 0.1043,
      "grad_norm": 3.0302977561950684,
      "learning_rate": 4.523737373737374e-05,
      "loss": 0.0372,
      "step": 1043
    },
    {
      "epoch": 0.1044,
      "grad_norm": 17.859586715698242,
      "learning_rate": 4.5232323232323235e-05,
      "loss": 0.1341,
      "step": 1044
    },
    {
      "epoch": 0.1045,
      "grad_norm": 8.162103652954102,
      "learning_rate": 4.522727272727273e-05,
      "loss": 0.1241,
      "step": 1045
    },
    {
      "epoch": 0.1046,
      "grad_norm": 5.893673896789551,
      "learning_rate": 4.522222222222223e-05,
      "loss": 0.2645,
      "step": 1046
    },
    {
      "epoch": 0.1047,
      "grad_norm": 2.167904853820801,
      "learning_rate": 4.5217171717171716e-05,
      "loss": 0.0519,
      "step": 1047
    },
    {
      "epoch": 0.1048,
      "grad_norm": 4.966544151306152,
      "learning_rate": 4.521212121212122e-05,
      "loss": 0.2769,
      "step": 1048
    },
    {
      "epoch": 0.1049,
      "grad_norm": 23.265810012817383,
      "learning_rate": 4.520707070707071e-05,
      "loss": 1.0905,
      "step": 1049
    },
    {
      "epoch": 0.105,
      "grad_norm": 2.749390125274658,
      "learning_rate": 4.5202020202020204e-05,
      "loss": 0.052,
      "step": 1050
    },
    {
      "epoch": 0.1051,
      "grad_norm": 17.251720428466797,
      "learning_rate": 4.51969696969697e-05,
      "loss": 0.1874,
      "step": 1051
    },
    {
      "epoch": 0.1052,
      "grad_norm": 8.531375885009766,
      "learning_rate": 4.5191919191919196e-05,
      "loss": 0.3082,
      "step": 1052
    },
    {
      "epoch": 0.1053,
      "grad_norm": 11.067678451538086,
      "learning_rate": 4.518686868686869e-05,
      "loss": 0.2634,
      "step": 1053
    },
    {
      "epoch": 0.1054,
      "grad_norm": 5.23277473449707,
      "learning_rate": 4.518181818181819e-05,
      "loss": 0.1117,
      "step": 1054
    },
    {
      "epoch": 0.1055,
      "grad_norm": 3.8088037967681885,
      "learning_rate": 4.517676767676768e-05,
      "loss": 0.1409,
      "step": 1055
    },
    {
      "epoch": 0.1056,
      "grad_norm": 4.35794734954834,
      "learning_rate": 4.517171717171717e-05,
      "loss": 0.1754,
      "step": 1056
    },
    {
      "epoch": 0.1057,
      "grad_norm": 2.471139430999756,
      "learning_rate": 4.516666666666667e-05,
      "loss": 0.5224,
      "step": 1057
    },
    {
      "epoch": 0.1058,
      "grad_norm": 6.178231239318848,
      "learning_rate": 4.5161616161616165e-05,
      "loss": 0.201,
      "step": 1058
    },
    {
      "epoch": 0.1059,
      "grad_norm": 9.900121688842773,
      "learning_rate": 4.515656565656566e-05,
      "loss": 0.99,
      "step": 1059
    },
    {
      "epoch": 0.106,
      "grad_norm": 6.302709102630615,
      "learning_rate": 4.515151515151516e-05,
      "loss": 0.5927,
      "step": 1060
    },
    {
      "epoch": 0.1061,
      "grad_norm": 2.072674512863159,
      "learning_rate": 4.5146464646464646e-05,
      "loss": 0.0358,
      "step": 1061
    },
    {
      "epoch": 0.1062,
      "grad_norm": 4.2309770584106445,
      "learning_rate": 4.514141414141415e-05,
      "loss": 0.1929,
      "step": 1062
    },
    {
      "epoch": 0.1063,
      "grad_norm": 6.7845778465271,
      "learning_rate": 4.513636363636364e-05,
      "loss": 0.345,
      "step": 1063
    },
    {
      "epoch": 0.1064,
      "grad_norm": 2.7231786251068115,
      "learning_rate": 4.5131313131313134e-05,
      "loss": 0.0586,
      "step": 1064
    },
    {
      "epoch": 0.1065,
      "grad_norm": 10.000534057617188,
      "learning_rate": 4.512626262626263e-05,
      "loss": 0.1093,
      "step": 1065
    },
    {
      "epoch": 0.1066,
      "grad_norm": 11.757752418518066,
      "learning_rate": 4.5121212121212126e-05,
      "loss": 0.7494,
      "step": 1066
    },
    {
      "epoch": 0.1067,
      "grad_norm": 6.9282732009887695,
      "learning_rate": 4.5116161616161615e-05,
      "loss": 0.613,
      "step": 1067
    },
    {
      "epoch": 0.1068,
      "grad_norm": 5.957398891448975,
      "learning_rate": 4.511111111111112e-05,
      "loss": 0.4477,
      "step": 1068
    },
    {
      "epoch": 0.1069,
      "grad_norm": 26.443105697631836,
      "learning_rate": 4.510606060606061e-05,
      "loss": 1.0087,
      "step": 1069
    },
    {
      "epoch": 0.107,
      "grad_norm": 9.564922332763672,
      "learning_rate": 4.51010101010101e-05,
      "loss": 0.1666,
      "step": 1070
    },
    {
      "epoch": 0.1071,
      "grad_norm": 13.430631637573242,
      "learning_rate": 4.50959595959596e-05,
      "loss": 0.3674,
      "step": 1071
    },
    {
      "epoch": 0.1072,
      "grad_norm": 8.489232063293457,
      "learning_rate": 4.5090909090909095e-05,
      "loss": 0.9446,
      "step": 1072
    },
    {
      "epoch": 0.1073,
      "grad_norm": 8.65447998046875,
      "learning_rate": 4.5085858585858585e-05,
      "loss": 0.5461,
      "step": 1073
    },
    {
      "epoch": 0.1074,
      "grad_norm": 10.628327369689941,
      "learning_rate": 4.508080808080809e-05,
      "loss": 0.5755,
      "step": 1074
    },
    {
      "epoch": 0.1075,
      "grad_norm": 4.940393447875977,
      "learning_rate": 4.5075757575757577e-05,
      "loss": 0.5701,
      "step": 1075
    },
    {
      "epoch": 0.1076,
      "grad_norm": 4.016294479370117,
      "learning_rate": 4.507070707070707e-05,
      "loss": 0.6005,
      "step": 1076
    },
    {
      "epoch": 0.1077,
      "grad_norm": 5.706302165985107,
      "learning_rate": 4.506565656565657e-05,
      "loss": 0.1487,
      "step": 1077
    },
    {
      "epoch": 0.1078,
      "grad_norm": 10.525063514709473,
      "learning_rate": 4.5060606060606065e-05,
      "loss": 0.2575,
      "step": 1078
    },
    {
      "epoch": 0.1079,
      "grad_norm": 9.318092346191406,
      "learning_rate": 4.5055555555555554e-05,
      "loss": 1.234,
      "step": 1079
    },
    {
      "epoch": 0.108,
      "grad_norm": 13.029769897460938,
      "learning_rate": 4.5050505050505056e-05,
      "loss": 1.113,
      "step": 1080
    },
    {
      "epoch": 0.1081,
      "grad_norm": 8.973505973815918,
      "learning_rate": 4.5045454545454546e-05,
      "loss": 0.5219,
      "step": 1081
    },
    {
      "epoch": 0.1082,
      "grad_norm": 8.861003875732422,
      "learning_rate": 4.504040404040404e-05,
      "loss": 0.5777,
      "step": 1082
    },
    {
      "epoch": 0.1083,
      "grad_norm": 17.022499084472656,
      "learning_rate": 4.503535353535354e-05,
      "loss": 0.6011,
      "step": 1083
    },
    {
      "epoch": 0.1084,
      "grad_norm": 4.243625164031982,
      "learning_rate": 4.5030303030303034e-05,
      "loss": 0.1368,
      "step": 1084
    },
    {
      "epoch": 0.1085,
      "grad_norm": 2.3952231407165527,
      "learning_rate": 4.502525252525252e-05,
      "loss": 0.0456,
      "step": 1085
    },
    {
      "epoch": 0.1086,
      "grad_norm": 7.48283576965332,
      "learning_rate": 4.5020202020202026e-05,
      "loss": 0.4203,
      "step": 1086
    },
    {
      "epoch": 0.1087,
      "grad_norm": 5.670314311981201,
      "learning_rate": 4.5015151515151515e-05,
      "loss": 0.5825,
      "step": 1087
    },
    {
      "epoch": 0.1088,
      "grad_norm": 12.387399673461914,
      "learning_rate": 4.501010101010101e-05,
      "loss": 0.9744,
      "step": 1088
    },
    {
      "epoch": 0.1089,
      "grad_norm": 18.44068717956543,
      "learning_rate": 4.500505050505051e-05,
      "loss": 0.2191,
      "step": 1089
    },
    {
      "epoch": 0.109,
      "grad_norm": 12.893267631530762,
      "learning_rate": 4.5e-05,
      "loss": 0.1734,
      "step": 1090
    },
    {
      "epoch": 0.1091,
      "grad_norm": 8.427271842956543,
      "learning_rate": 4.499494949494949e-05,
      "loss": 0.4266,
      "step": 1091
    },
    {
      "epoch": 0.1092,
      "grad_norm": 6.068711757659912,
      "learning_rate": 4.4989898989898995e-05,
      "loss": 0.1867,
      "step": 1092
    },
    {
      "epoch": 0.1093,
      "grad_norm": 3.7673492431640625,
      "learning_rate": 4.4984848484848484e-05,
      "loss": 0.0924,
      "step": 1093
    },
    {
      "epoch": 0.1094,
      "grad_norm": 3.2130677700042725,
      "learning_rate": 4.497979797979798e-05,
      "loss": 0.0718,
      "step": 1094
    },
    {
      "epoch": 0.1095,
      "grad_norm": 4.297408580780029,
      "learning_rate": 4.4974747474747476e-05,
      "loss": 0.2709,
      "step": 1095
    },
    {
      "epoch": 0.1096,
      "grad_norm": 3.953362464904785,
      "learning_rate": 4.496969696969697e-05,
      "loss": 0.5602,
      "step": 1096
    },
    {
      "epoch": 0.1097,
      "grad_norm": 20.66802406311035,
      "learning_rate": 4.496464646464647e-05,
      "loss": 0.2584,
      "step": 1097
    },
    {
      "epoch": 0.1098,
      "grad_norm": 7.889944076538086,
      "learning_rate": 4.4959595959595964e-05,
      "loss": 0.3948,
      "step": 1098
    },
    {
      "epoch": 0.1099,
      "grad_norm": 7.195418834686279,
      "learning_rate": 4.495454545454545e-05,
      "loss": 0.7356,
      "step": 1099
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.17496395111084,
      "learning_rate": 4.494949494949495e-05,
      "loss": 0.4161,
      "step": 1100
    },
    {
      "epoch": 0.1101,
      "grad_norm": 12.366473197937012,
      "learning_rate": 4.4944444444444445e-05,
      "loss": 0.5328,
      "step": 1101
    },
    {
      "epoch": 0.1102,
      "grad_norm": 7.36449670791626,
      "learning_rate": 4.493939393939394e-05,
      "loss": 1.0954,
      "step": 1102
    },
    {
      "epoch": 0.1103,
      "grad_norm": 8.068277359008789,
      "learning_rate": 4.493434343434344e-05,
      "loss": 0.4487,
      "step": 1103
    },
    {
      "epoch": 0.1104,
      "grad_norm": 27.101909637451172,
      "learning_rate": 4.492929292929293e-05,
      "loss": 0.2846,
      "step": 1104
    },
    {
      "epoch": 0.1105,
      "grad_norm": 6.1408185958862305,
      "learning_rate": 4.492424242424242e-05,
      "loss": 0.2001,
      "step": 1105
    },
    {
      "epoch": 0.1106,
      "grad_norm": 7.50735330581665,
      "learning_rate": 4.4919191919191925e-05,
      "loss": 0.1426,
      "step": 1106
    },
    {
      "epoch": 0.1107,
      "grad_norm": 14.337894439697266,
      "learning_rate": 4.4914141414141414e-05,
      "loss": 0.5612,
      "step": 1107
    },
    {
      "epoch": 0.1108,
      "grad_norm": 5.432666301727295,
      "learning_rate": 4.490909090909091e-05,
      "loss": 0.3085,
      "step": 1108
    },
    {
      "epoch": 0.1109,
      "grad_norm": 9.03901195526123,
      "learning_rate": 4.4904040404040406e-05,
      "loss": 0.8959,
      "step": 1109
    },
    {
      "epoch": 0.111,
      "grad_norm": 6.998770236968994,
      "learning_rate": 4.48989898989899e-05,
      "loss": 0.5281,
      "step": 1110
    },
    {
      "epoch": 0.1111,
      "grad_norm": 4.3077802658081055,
      "learning_rate": 4.489393939393939e-05,
      "loss": 0.1069,
      "step": 1111
    },
    {
      "epoch": 0.1112,
      "grad_norm": 8.19987964630127,
      "learning_rate": 4.4888888888888894e-05,
      "loss": 0.1434,
      "step": 1112
    },
    {
      "epoch": 0.1113,
      "grad_norm": 6.147011756896973,
      "learning_rate": 4.4883838383838384e-05,
      "loss": 0.7805,
      "step": 1113
    },
    {
      "epoch": 0.1114,
      "grad_norm": 4.361591339111328,
      "learning_rate": 4.487878787878788e-05,
      "loss": 0.038,
      "step": 1114
    },
    {
      "epoch": 0.1115,
      "grad_norm": 7.8595476150512695,
      "learning_rate": 4.4873737373737375e-05,
      "loss": 0.2053,
      "step": 1115
    },
    {
      "epoch": 0.1116,
      "grad_norm": 11.272153854370117,
      "learning_rate": 4.486868686868687e-05,
      "loss": 0.5411,
      "step": 1116
    },
    {
      "epoch": 0.1117,
      "grad_norm": 5.56926965713501,
      "learning_rate": 4.486363636363636e-05,
      "loss": 0.1655,
      "step": 1117
    },
    {
      "epoch": 0.1118,
      "grad_norm": 3.395500659942627,
      "learning_rate": 4.4858585858585863e-05,
      "loss": 0.1246,
      "step": 1118
    },
    {
      "epoch": 0.1119,
      "grad_norm": 7.012539863586426,
      "learning_rate": 4.485353535353535e-05,
      "loss": 0.4581,
      "step": 1119
    },
    {
      "epoch": 0.112,
      "grad_norm": 6.211499214172363,
      "learning_rate": 4.484848484848485e-05,
      "loss": 0.1607,
      "step": 1120
    },
    {
      "epoch": 0.1121,
      "grad_norm": 11.355920791625977,
      "learning_rate": 4.4843434343434345e-05,
      "loss": 1.3705,
      "step": 1121
    },
    {
      "epoch": 0.1122,
      "grad_norm": 9.894545555114746,
      "learning_rate": 4.483838383838384e-05,
      "loss": 0.0659,
      "step": 1122
    },
    {
      "epoch": 0.1123,
      "grad_norm": 29.38459014892578,
      "learning_rate": 4.483333333333333e-05,
      "loss": 0.4724,
      "step": 1123
    },
    {
      "epoch": 0.1124,
      "grad_norm": 8.854192733764648,
      "learning_rate": 4.482828282828283e-05,
      "loss": 0.3125,
      "step": 1124
    },
    {
      "epoch": 0.1125,
      "grad_norm": 16.061552047729492,
      "learning_rate": 4.482323232323233e-05,
      "loss": 0.1558,
      "step": 1125
    },
    {
      "epoch": 0.1126,
      "grad_norm": 17.867448806762695,
      "learning_rate": 4.481818181818182e-05,
      "loss": 0.665,
      "step": 1126
    },
    {
      "epoch": 0.1127,
      "grad_norm": 9.948736190795898,
      "learning_rate": 4.481313131313132e-05,
      "loss": 0.2474,
      "step": 1127
    },
    {
      "epoch": 0.1128,
      "grad_norm": 7.072950839996338,
      "learning_rate": 4.480808080808081e-05,
      "loss": 0.4601,
      "step": 1128
    },
    {
      "epoch": 0.1129,
      "grad_norm": 11.43730354309082,
      "learning_rate": 4.4803030303030306e-05,
      "loss": 1.043,
      "step": 1129
    },
    {
      "epoch": 0.113,
      "grad_norm": 2.3476924896240234,
      "learning_rate": 4.47979797979798e-05,
      "loss": 0.0541,
      "step": 1130
    },
    {
      "epoch": 0.1131,
      "grad_norm": 5.2770233154296875,
      "learning_rate": 4.47929292929293e-05,
      "loss": 0.1521,
      "step": 1131
    },
    {
      "epoch": 0.1132,
      "grad_norm": 10.142876625061035,
      "learning_rate": 4.478787878787879e-05,
      "loss": 0.6063,
      "step": 1132
    },
    {
      "epoch": 0.1133,
      "grad_norm": 7.009751796722412,
      "learning_rate": 4.478282828282829e-05,
      "loss": 0.585,
      "step": 1133
    },
    {
      "epoch": 0.1134,
      "grad_norm": 7.300774097442627,
      "learning_rate": 4.477777777777778e-05,
      "loss": 0.949,
      "step": 1134
    },
    {
      "epoch": 0.1135,
      "grad_norm": 6.127287864685059,
      "learning_rate": 4.4772727272727275e-05,
      "loss": 0.6037,
      "step": 1135
    },
    {
      "epoch": 0.1136,
      "grad_norm": 10.074549674987793,
      "learning_rate": 4.476767676767677e-05,
      "loss": 0.4733,
      "step": 1136
    },
    {
      "epoch": 0.1137,
      "grad_norm": 12.124725341796875,
      "learning_rate": 4.476262626262627e-05,
      "loss": 1.2318,
      "step": 1137
    },
    {
      "epoch": 0.1138,
      "grad_norm": 2.310382604598999,
      "learning_rate": 4.4757575757575756e-05,
      "loss": 0.0155,
      "step": 1138
    },
    {
      "epoch": 0.1139,
      "grad_norm": 9.640390396118164,
      "learning_rate": 4.475252525252526e-05,
      "loss": 0.6536,
      "step": 1139
    },
    {
      "epoch": 0.114,
      "grad_norm": 8.65160083770752,
      "learning_rate": 4.474747474747475e-05,
      "loss": 0.6011,
      "step": 1140
    },
    {
      "epoch": 0.1141,
      "grad_norm": 10.773650169372559,
      "learning_rate": 4.4742424242424244e-05,
      "loss": 0.6512,
      "step": 1141
    },
    {
      "epoch": 0.1142,
      "grad_norm": 5.146366596221924,
      "learning_rate": 4.473737373737374e-05,
      "loss": 0.1181,
      "step": 1142
    },
    {
      "epoch": 0.1143,
      "grad_norm": 4.857442378997803,
      "learning_rate": 4.4732323232323236e-05,
      "loss": 0.3416,
      "step": 1143
    },
    {
      "epoch": 0.1144,
      "grad_norm": 19.073820114135742,
      "learning_rate": 4.472727272727273e-05,
      "loss": 0.7915,
      "step": 1144
    },
    {
      "epoch": 0.1145,
      "grad_norm": 7.444849491119385,
      "learning_rate": 4.472222222222223e-05,
      "loss": 0.5404,
      "step": 1145
    },
    {
      "epoch": 0.1146,
      "grad_norm": 10.282227516174316,
      "learning_rate": 4.471717171717172e-05,
      "loss": 0.595,
      "step": 1146
    },
    {
      "epoch": 0.1147,
      "grad_norm": 6.329639434814453,
      "learning_rate": 4.471212121212121e-05,
      "loss": 0.5423,
      "step": 1147
    },
    {
      "epoch": 0.1148,
      "grad_norm": 16.063196182250977,
      "learning_rate": 4.470707070707071e-05,
      "loss": 0.5137,
      "step": 1148
    },
    {
      "epoch": 0.1149,
      "grad_norm": 8.994356155395508,
      "learning_rate": 4.4702020202020205e-05,
      "loss": 1.105,
      "step": 1149
    },
    {
      "epoch": 0.115,
      "grad_norm": 19.905059814453125,
      "learning_rate": 4.46969696969697e-05,
      "loss": 1.3545,
      "step": 1150
    },
    {
      "epoch": 0.1151,
      "grad_norm": 21.658096313476562,
      "learning_rate": 4.46919191919192e-05,
      "loss": 0.1572,
      "step": 1151
    },
    {
      "epoch": 0.1152,
      "grad_norm": 4.982419013977051,
      "learning_rate": 4.4686868686868686e-05,
      "loss": 0.5677,
      "step": 1152
    },
    {
      "epoch": 0.1153,
      "grad_norm": 2.9153759479522705,
      "learning_rate": 4.468181818181818e-05,
      "loss": 0.112,
      "step": 1153
    },
    {
      "epoch": 0.1154,
      "grad_norm": 4.274677753448486,
      "learning_rate": 4.467676767676768e-05,
      "loss": 0.0827,
      "step": 1154
    },
    {
      "epoch": 0.1155,
      "grad_norm": 6.110588073730469,
      "learning_rate": 4.4671717171717174e-05,
      "loss": 0.9852,
      "step": 1155
    },
    {
      "epoch": 0.1156,
      "grad_norm": 10.049993515014648,
      "learning_rate": 4.466666666666667e-05,
      "loss": 0.2466,
      "step": 1156
    },
    {
      "epoch": 0.1157,
      "grad_norm": 8.98173713684082,
      "learning_rate": 4.4661616161616166e-05,
      "loss": 0.3951,
      "step": 1157
    },
    {
      "epoch": 0.1158,
      "grad_norm": 4.786052227020264,
      "learning_rate": 4.4656565656565656e-05,
      "loss": 0.0951,
      "step": 1158
    },
    {
      "epoch": 0.1159,
      "grad_norm": 4.401653289794922,
      "learning_rate": 4.465151515151516e-05,
      "loss": 0.1209,
      "step": 1159
    },
    {
      "epoch": 0.116,
      "grad_norm": 2.0766563415527344,
      "learning_rate": 4.464646464646465e-05,
      "loss": 0.0337,
      "step": 1160
    },
    {
      "epoch": 0.1161,
      "grad_norm": 11.426548957824707,
      "learning_rate": 4.4641414141414144e-05,
      "loss": 0.2109,
      "step": 1161
    },
    {
      "epoch": 0.1162,
      "grad_norm": 9.803309440612793,
      "learning_rate": 4.463636363636364e-05,
      "loss": 0.811,
      "step": 1162
    },
    {
      "epoch": 0.1163,
      "grad_norm": 7.788112163543701,
      "learning_rate": 4.4631313131313136e-05,
      "loss": 0.1422,
      "step": 1163
    },
    {
      "epoch": 0.1164,
      "grad_norm": 3.6096062660217285,
      "learning_rate": 4.4626262626262625e-05,
      "loss": 0.0345,
      "step": 1164
    },
    {
      "epoch": 0.1165,
      "grad_norm": 8.135470390319824,
      "learning_rate": 4.462121212121213e-05,
      "loss": 0.55,
      "step": 1165
    },
    {
      "epoch": 0.1166,
      "grad_norm": 5.639625072479248,
      "learning_rate": 4.461616161616162e-05,
      "loss": 0.3051,
      "step": 1166
    },
    {
      "epoch": 0.1167,
      "grad_norm": 7.760005474090576,
      "learning_rate": 4.461111111111111e-05,
      "loss": 0.3625,
      "step": 1167
    },
    {
      "epoch": 0.1168,
      "grad_norm": 7.010201930999756,
      "learning_rate": 4.460606060606061e-05,
      "loss": 0.4425,
      "step": 1168
    },
    {
      "epoch": 0.1169,
      "grad_norm": 3.206435203552246,
      "learning_rate": 4.4601010101010105e-05,
      "loss": 0.0577,
      "step": 1169
    },
    {
      "epoch": 0.117,
      "grad_norm": 12.447701454162598,
      "learning_rate": 4.4595959595959594e-05,
      "loss": 0.1609,
      "step": 1170
    },
    {
      "epoch": 0.1171,
      "grad_norm": 5.868978023529053,
      "learning_rate": 4.45909090909091e-05,
      "loss": 0.1715,
      "step": 1171
    },
    {
      "epoch": 0.1172,
      "grad_norm": 6.553044319152832,
      "learning_rate": 4.4585858585858586e-05,
      "loss": 0.232,
      "step": 1172
    },
    {
      "epoch": 0.1173,
      "grad_norm": 8.230452537536621,
      "learning_rate": 4.458080808080808e-05,
      "loss": 0.2353,
      "step": 1173
    },
    {
      "epoch": 0.1174,
      "grad_norm": 3.585400342941284,
      "learning_rate": 4.457575757575758e-05,
      "loss": 0.0478,
      "step": 1174
    },
    {
      "epoch": 0.1175,
      "grad_norm": 1.887675404548645,
      "learning_rate": 4.4570707070707074e-05,
      "loss": 0.0481,
      "step": 1175
    },
    {
      "epoch": 0.1176,
      "grad_norm": 4.644859790802002,
      "learning_rate": 4.456565656565656e-05,
      "loss": 0.5824,
      "step": 1176
    },
    {
      "epoch": 0.1177,
      "grad_norm": 5.269247055053711,
      "learning_rate": 4.4560606060606066e-05,
      "loss": 0.0939,
      "step": 1177
    },
    {
      "epoch": 0.1178,
      "grad_norm": 9.606133460998535,
      "learning_rate": 4.4555555555555555e-05,
      "loss": 0.2268,
      "step": 1178
    },
    {
      "epoch": 0.1179,
      "grad_norm": 3.4100289344787598,
      "learning_rate": 4.455050505050505e-05,
      "loss": 0.0641,
      "step": 1179
    },
    {
      "epoch": 0.118,
      "grad_norm": 4.9919915199279785,
      "learning_rate": 4.454545454545455e-05,
      "loss": 0.4761,
      "step": 1180
    },
    {
      "epoch": 0.1181,
      "grad_norm": 1.7008004188537598,
      "learning_rate": 4.454040404040404e-05,
      "loss": 0.0142,
      "step": 1181
    },
    {
      "epoch": 0.1182,
      "grad_norm": 10.036349296569824,
      "learning_rate": 4.453535353535353e-05,
      "loss": 1.4267,
      "step": 1182
    },
    {
      "epoch": 0.1183,
      "grad_norm": 9.742138862609863,
      "learning_rate": 4.4530303030303035e-05,
      "loss": 0.3534,
      "step": 1183
    },
    {
      "epoch": 0.1184,
      "grad_norm": 4.172444820404053,
      "learning_rate": 4.4525252525252524e-05,
      "loss": 0.3566,
      "step": 1184
    },
    {
      "epoch": 0.1185,
      "grad_norm": 6.811132907867432,
      "learning_rate": 4.452020202020202e-05,
      "loss": 0.4557,
      "step": 1185
    },
    {
      "epoch": 0.1186,
      "grad_norm": 6.019387245178223,
      "learning_rate": 4.4515151515151516e-05,
      "loss": 0.2163,
      "step": 1186
    },
    {
      "epoch": 0.1187,
      "grad_norm": 18.54116439819336,
      "learning_rate": 4.451010101010101e-05,
      "loss": 0.9289,
      "step": 1187
    },
    {
      "epoch": 0.1188,
      "grad_norm": 5.9078898429870605,
      "learning_rate": 4.450505050505051e-05,
      "loss": 0.0883,
      "step": 1188
    },
    {
      "epoch": 0.1189,
      "grad_norm": 4.111548900604248,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 0.3771,
      "step": 1189
    },
    {
      "epoch": 0.119,
      "grad_norm": 4.032368183135986,
      "learning_rate": 4.4494949494949493e-05,
      "loss": 0.0933,
      "step": 1190
    },
    {
      "epoch": 0.1191,
      "grad_norm": 11.645378112792969,
      "learning_rate": 4.448989898989899e-05,
      "loss": 0.1669,
      "step": 1191
    },
    {
      "epoch": 0.1192,
      "grad_norm": 4.993757724761963,
      "learning_rate": 4.4484848484848485e-05,
      "loss": 0.1808,
      "step": 1192
    },
    {
      "epoch": 0.1193,
      "grad_norm": 24.81055450439453,
      "learning_rate": 4.447979797979798e-05,
      "loss": 0.848,
      "step": 1193
    },
    {
      "epoch": 0.1194,
      "grad_norm": 8.030146598815918,
      "learning_rate": 4.447474747474748e-05,
      "loss": 0.2885,
      "step": 1194
    },
    {
      "epoch": 0.1195,
      "grad_norm": 7.1218743324279785,
      "learning_rate": 4.4469696969696973e-05,
      "loss": 0.0641,
      "step": 1195
    },
    {
      "epoch": 0.1196,
      "grad_norm": 3.776982307434082,
      "learning_rate": 4.446464646464646e-05,
      "loss": 0.4669,
      "step": 1196
    },
    {
      "epoch": 0.1197,
      "grad_norm": 3.734731674194336,
      "learning_rate": 4.4459595959595965e-05,
      "loss": 0.1017,
      "step": 1197
    },
    {
      "epoch": 0.1198,
      "grad_norm": 5.913084506988525,
      "learning_rate": 4.4454545454545455e-05,
      "loss": 0.2401,
      "step": 1198
    },
    {
      "epoch": 0.1199,
      "grad_norm": 8.518794059753418,
      "learning_rate": 4.444949494949495e-05,
      "loss": 0.7842,
      "step": 1199
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.942734718322754,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.4672,
      "step": 1200
    },
    {
      "epoch": 0.1201,
      "grad_norm": 5.863522529602051,
      "learning_rate": 4.443939393939394e-05,
      "loss": 0.3512,
      "step": 1201
    },
    {
      "epoch": 0.1202,
      "grad_norm": 6.591408729553223,
      "learning_rate": 4.443434343434343e-05,
      "loss": 0.484,
      "step": 1202
    },
    {
      "epoch": 0.1203,
      "grad_norm": 6.335476398468018,
      "learning_rate": 4.4429292929292935e-05,
      "loss": 0.1366,
      "step": 1203
    },
    {
      "epoch": 0.1204,
      "grad_norm": 10.778593063354492,
      "learning_rate": 4.4424242424242424e-05,
      "loss": 0.0923,
      "step": 1204
    },
    {
      "epoch": 0.1205,
      "grad_norm": 3.834106206893921,
      "learning_rate": 4.441919191919192e-05,
      "loss": 0.0741,
      "step": 1205
    },
    {
      "epoch": 0.1206,
      "grad_norm": 4.841442108154297,
      "learning_rate": 4.4414141414141416e-05,
      "loss": 0.3747,
      "step": 1206
    },
    {
      "epoch": 0.1207,
      "grad_norm": 4.5432515144348145,
      "learning_rate": 4.440909090909091e-05,
      "loss": 0.0722,
      "step": 1207
    },
    {
      "epoch": 0.1208,
      "grad_norm": 3.1321566104888916,
      "learning_rate": 4.44040404040404e-05,
      "loss": 0.4839,
      "step": 1208
    },
    {
      "epoch": 0.1209,
      "grad_norm": 8.720353126525879,
      "learning_rate": 4.4398989898989904e-05,
      "loss": 0.2491,
      "step": 1209
    },
    {
      "epoch": 0.121,
      "grad_norm": 11.165698051452637,
      "learning_rate": 4.43939393939394e-05,
      "loss": 0.9673,
      "step": 1210
    },
    {
      "epoch": 0.1211,
      "grad_norm": 6.231223106384277,
      "learning_rate": 4.438888888888889e-05,
      "loss": 0.8276,
      "step": 1211
    },
    {
      "epoch": 0.1212,
      "grad_norm": 5.251862049102783,
      "learning_rate": 4.438383838383839e-05,
      "loss": 0.333,
      "step": 1212
    },
    {
      "epoch": 0.1213,
      "grad_norm": 9.02682113647461,
      "learning_rate": 4.437878787878788e-05,
      "loss": 0.3585,
      "step": 1213
    },
    {
      "epoch": 0.1214,
      "grad_norm": 5.070196151733398,
      "learning_rate": 4.437373737373738e-05,
      "loss": 0.5721,
      "step": 1214
    },
    {
      "epoch": 0.1215,
      "grad_norm": 2.8672547340393066,
      "learning_rate": 4.436868686868687e-05,
      "loss": 0.1289,
      "step": 1215
    },
    {
      "epoch": 0.1216,
      "grad_norm": 12.949196815490723,
      "learning_rate": 4.436363636363637e-05,
      "loss": 0.681,
      "step": 1216
    },
    {
      "epoch": 0.1217,
      "grad_norm": 10.080541610717773,
      "learning_rate": 4.435858585858586e-05,
      "loss": 0.409,
      "step": 1217
    },
    {
      "epoch": 0.1218,
      "grad_norm": 6.476293087005615,
      "learning_rate": 4.435353535353536e-05,
      "loss": 0.6175,
      "step": 1218
    },
    {
      "epoch": 0.1219,
      "grad_norm": 7.235831260681152,
      "learning_rate": 4.434848484848485e-05,
      "loss": 0.7315,
      "step": 1219
    },
    {
      "epoch": 0.122,
      "grad_norm": 19.88304328918457,
      "learning_rate": 4.4343434343434346e-05,
      "loss": 0.4451,
      "step": 1220
    },
    {
      "epoch": 0.1221,
      "grad_norm": 10.404736518859863,
      "learning_rate": 4.433838383838384e-05,
      "loss": 0.4506,
      "step": 1221
    },
    {
      "epoch": 0.1222,
      "grad_norm": 10.841041564941406,
      "learning_rate": 4.433333333333334e-05,
      "loss": 0.2469,
      "step": 1222
    },
    {
      "epoch": 0.1223,
      "grad_norm": 3.6923727989196777,
      "learning_rate": 4.432828282828283e-05,
      "loss": 0.1418,
      "step": 1223
    },
    {
      "epoch": 0.1224,
      "grad_norm": 6.295706748962402,
      "learning_rate": 4.432323232323233e-05,
      "loss": 0.512,
      "step": 1224
    },
    {
      "epoch": 0.1225,
      "grad_norm": 20.84015655517578,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.7291,
      "step": 1225
    },
    {
      "epoch": 0.1226,
      "grad_norm": 45.681434631347656,
      "learning_rate": 4.4313131313131315e-05,
      "loss": 0.6792,
      "step": 1226
    },
    {
      "epoch": 0.1227,
      "grad_norm": 7.992038249969482,
      "learning_rate": 4.430808080808081e-05,
      "loss": 0.279,
      "step": 1227
    },
    {
      "epoch": 0.1228,
      "grad_norm": 6.093226909637451,
      "learning_rate": 4.430303030303031e-05,
      "loss": 0.1906,
      "step": 1228
    },
    {
      "epoch": 0.1229,
      "grad_norm": 5.226652145385742,
      "learning_rate": 4.4297979797979796e-05,
      "loss": 0.3908,
      "step": 1229
    },
    {
      "epoch": 0.123,
      "grad_norm": 10.860602378845215,
      "learning_rate": 4.42929292929293e-05,
      "loss": 0.7431,
      "step": 1230
    },
    {
      "epoch": 0.1231,
      "grad_norm": 6.15314245223999,
      "learning_rate": 4.428787878787879e-05,
      "loss": 0.268,
      "step": 1231
    },
    {
      "epoch": 0.1232,
      "grad_norm": 4.672374725341797,
      "learning_rate": 4.4282828282828284e-05,
      "loss": 0.5261,
      "step": 1232
    },
    {
      "epoch": 0.1233,
      "grad_norm": 2.7219204902648926,
      "learning_rate": 4.427777777777778e-05,
      "loss": 0.4843,
      "step": 1233
    },
    {
      "epoch": 0.1234,
      "grad_norm": 12.937013626098633,
      "learning_rate": 4.4272727272727276e-05,
      "loss": 0.5703,
      "step": 1234
    },
    {
      "epoch": 0.1235,
      "grad_norm": 10.619874954223633,
      "learning_rate": 4.426767676767677e-05,
      "loss": 0.5402,
      "step": 1235
    },
    {
      "epoch": 0.1236,
      "grad_norm": 11.357385635375977,
      "learning_rate": 4.426262626262627e-05,
      "loss": 0.5419,
      "step": 1236
    },
    {
      "epoch": 0.1237,
      "grad_norm": 10.362465858459473,
      "learning_rate": 4.425757575757576e-05,
      "loss": 0.7472,
      "step": 1237
    },
    {
      "epoch": 0.1238,
      "grad_norm": 7.339501857757568,
      "learning_rate": 4.4252525252525254e-05,
      "loss": 0.2635,
      "step": 1238
    },
    {
      "epoch": 0.1239,
      "grad_norm": 2.238816976547241,
      "learning_rate": 4.424747474747475e-05,
      "loss": 0.4808,
      "step": 1239
    },
    {
      "epoch": 0.124,
      "grad_norm": 24.70423126220703,
      "learning_rate": 4.4242424242424246e-05,
      "loss": 0.335,
      "step": 1240
    },
    {
      "epoch": 0.1241,
      "grad_norm": 8.073127746582031,
      "learning_rate": 4.423737373737374e-05,
      "loss": 0.3877,
      "step": 1241
    },
    {
      "epoch": 0.1242,
      "grad_norm": 3.4055113792419434,
      "learning_rate": 4.423232323232324e-05,
      "loss": 0.0957,
      "step": 1242
    },
    {
      "epoch": 0.1243,
      "grad_norm": 14.234118461608887,
      "learning_rate": 4.422727272727273e-05,
      "loss": 0.3325,
      "step": 1243
    },
    {
      "epoch": 0.1244,
      "grad_norm": 3.8062021732330322,
      "learning_rate": 4.422222222222222e-05,
      "loss": 0.5492,
      "step": 1244
    },
    {
      "epoch": 0.1245,
      "grad_norm": 3.345273017883301,
      "learning_rate": 4.421717171717172e-05,
      "loss": 0.1114,
      "step": 1245
    },
    {
      "epoch": 0.1246,
      "grad_norm": 9.491399765014648,
      "learning_rate": 4.4212121212121215e-05,
      "loss": 0.8431,
      "step": 1246
    },
    {
      "epoch": 0.1247,
      "grad_norm": 4.54273796081543,
      "learning_rate": 4.420707070707071e-05,
      "loss": 0.3704,
      "step": 1247
    },
    {
      "epoch": 0.1248,
      "grad_norm": 5.306372165679932,
      "learning_rate": 4.420202020202021e-05,
      "loss": 0.1219,
      "step": 1248
    },
    {
      "epoch": 0.1249,
      "grad_norm": 4.535835266113281,
      "learning_rate": 4.4196969696969696e-05,
      "loss": 0.1611,
      "step": 1249
    },
    {
      "epoch": 0.125,
      "grad_norm": 6.9904561042785645,
      "learning_rate": 4.41919191919192e-05,
      "loss": 0.3994,
      "step": 1250
    },
    {
      "epoch": 0.1251,
      "grad_norm": 5.363622665405273,
      "learning_rate": 4.418686868686869e-05,
      "loss": 0.7066,
      "step": 1251
    },
    {
      "epoch": 0.1252,
      "grad_norm": 6.170074462890625,
      "learning_rate": 4.4181818181818184e-05,
      "loss": 0.4342,
      "step": 1252
    },
    {
      "epoch": 0.1253,
      "grad_norm": 17.242006301879883,
      "learning_rate": 4.417676767676768e-05,
      "loss": 1.1948,
      "step": 1253
    },
    {
      "epoch": 0.1254,
      "grad_norm": 13.71372127532959,
      "learning_rate": 4.4171717171717176e-05,
      "loss": 0.6004,
      "step": 1254
    },
    {
      "epoch": 0.1255,
      "grad_norm": 5.361608982086182,
      "learning_rate": 4.4166666666666665e-05,
      "loss": 0.5912,
      "step": 1255
    },
    {
      "epoch": 0.1256,
      "grad_norm": 14.937453269958496,
      "learning_rate": 4.416161616161617e-05,
      "loss": 0.2749,
      "step": 1256
    },
    {
      "epoch": 0.1257,
      "grad_norm": 21.312585830688477,
      "learning_rate": 4.415656565656566e-05,
      "loss": 0.8121,
      "step": 1257
    },
    {
      "epoch": 0.1258,
      "grad_norm": 10.799200057983398,
      "learning_rate": 4.415151515151515e-05,
      "loss": 0.7342,
      "step": 1258
    },
    {
      "epoch": 0.1259,
      "grad_norm": 6.927521705627441,
      "learning_rate": 4.414646464646465e-05,
      "loss": 0.446,
      "step": 1259
    },
    {
      "epoch": 0.126,
      "grad_norm": 8.539000511169434,
      "learning_rate": 4.4141414141414145e-05,
      "loss": 0.2418,
      "step": 1260
    },
    {
      "epoch": 0.1261,
      "grad_norm": 4.654475212097168,
      "learning_rate": 4.4136363636363634e-05,
      "loss": 0.2605,
      "step": 1261
    },
    {
      "epoch": 0.1262,
      "grad_norm": 10.76779842376709,
      "learning_rate": 4.413131313131314e-05,
      "loss": 0.3138,
      "step": 1262
    },
    {
      "epoch": 0.1263,
      "grad_norm": 12.550968170166016,
      "learning_rate": 4.4126262626262626e-05,
      "loss": 0.291,
      "step": 1263
    },
    {
      "epoch": 0.1264,
      "grad_norm": 9.075549125671387,
      "learning_rate": 4.412121212121212e-05,
      "loss": 0.6299,
      "step": 1264
    },
    {
      "epoch": 0.1265,
      "grad_norm": 8.371729850769043,
      "learning_rate": 4.411616161616162e-05,
      "loss": 0.3716,
      "step": 1265
    },
    {
      "epoch": 0.1266,
      "grad_norm": 5.233151435852051,
      "learning_rate": 4.4111111111111114e-05,
      "loss": 0.3912,
      "step": 1266
    },
    {
      "epoch": 0.1267,
      "grad_norm": 7.246476650238037,
      "learning_rate": 4.41060606060606e-05,
      "loss": 0.7154,
      "step": 1267
    },
    {
      "epoch": 0.1268,
      "grad_norm": 20.818300247192383,
      "learning_rate": 4.4101010101010106e-05,
      "loss": 0.8008,
      "step": 1268
    },
    {
      "epoch": 0.1269,
      "grad_norm": 10.91275691986084,
      "learning_rate": 4.4095959595959595e-05,
      "loss": 0.2746,
      "step": 1269
    },
    {
      "epoch": 0.127,
      "grad_norm": 7.837447166442871,
      "learning_rate": 4.409090909090909e-05,
      "loss": 0.7036,
      "step": 1270
    },
    {
      "epoch": 0.1271,
      "grad_norm": 4.871488094329834,
      "learning_rate": 4.408585858585859e-05,
      "loss": 0.1309,
      "step": 1271
    },
    {
      "epoch": 0.1272,
      "grad_norm": 5.38530158996582,
      "learning_rate": 4.408080808080808e-05,
      "loss": 0.3501,
      "step": 1272
    },
    {
      "epoch": 0.1273,
      "grad_norm": 2.9805421829223633,
      "learning_rate": 4.407575757575757e-05,
      "loss": 0.0566,
      "step": 1273
    },
    {
      "epoch": 0.1274,
      "grad_norm": 10.267814636230469,
      "learning_rate": 4.4070707070707075e-05,
      "loss": 0.5796,
      "step": 1274
    },
    {
      "epoch": 0.1275,
      "grad_norm": 7.002389430999756,
      "learning_rate": 4.4065656565656565e-05,
      "loss": 0.1273,
      "step": 1275
    },
    {
      "epoch": 0.1276,
      "grad_norm": 4.04514741897583,
      "learning_rate": 4.406060606060606e-05,
      "loss": 0.1021,
      "step": 1276
    },
    {
      "epoch": 0.1277,
      "grad_norm": 10.235160827636719,
      "learning_rate": 4.4055555555555557e-05,
      "loss": 0.3806,
      "step": 1277
    },
    {
      "epoch": 0.1278,
      "grad_norm": 14.341047286987305,
      "learning_rate": 4.405050505050505e-05,
      "loss": 0.7543,
      "step": 1278
    },
    {
      "epoch": 0.1279,
      "grad_norm": 14.393119812011719,
      "learning_rate": 4.404545454545455e-05,
      "loss": 0.6644,
      "step": 1279
    },
    {
      "epoch": 0.128,
      "grad_norm": 6.11290979385376,
      "learning_rate": 4.4040404040404044e-05,
      "loss": 0.6223,
      "step": 1280
    },
    {
      "epoch": 0.1281,
      "grad_norm": 3.613393545150757,
      "learning_rate": 4.4035353535353534e-05,
      "loss": 0.1176,
      "step": 1281
    },
    {
      "epoch": 0.1282,
      "grad_norm": 6.4732441902160645,
      "learning_rate": 4.403030303030303e-05,
      "loss": 0.202,
      "step": 1282
    },
    {
      "epoch": 0.1283,
      "grad_norm": 4.847947597503662,
      "learning_rate": 4.4025252525252526e-05,
      "loss": 0.2818,
      "step": 1283
    },
    {
      "epoch": 0.1284,
      "grad_norm": 3.0768141746520996,
      "learning_rate": 4.402020202020202e-05,
      "loss": 0.1477,
      "step": 1284
    },
    {
      "epoch": 0.1285,
      "grad_norm": 11.616692543029785,
      "learning_rate": 4.401515151515152e-05,
      "loss": 0.4114,
      "step": 1285
    },
    {
      "epoch": 0.1286,
      "grad_norm": 19.34842872619629,
      "learning_rate": 4.4010101010101014e-05,
      "loss": 0.8344,
      "step": 1286
    },
    {
      "epoch": 0.1287,
      "grad_norm": 5.717264175415039,
      "learning_rate": 4.40050505050505e-05,
      "loss": 0.3433,
      "step": 1287
    },
    {
      "epoch": 0.1288,
      "grad_norm": 5.733270645141602,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1165,
      "step": 1288
    },
    {
      "epoch": 0.1289,
      "grad_norm": 7.348781108856201,
      "learning_rate": 4.3994949494949495e-05,
      "loss": 0.8719,
      "step": 1289
    },
    {
      "epoch": 0.129,
      "grad_norm": 2.6085989475250244,
      "learning_rate": 4.398989898989899e-05,
      "loss": 0.0654,
      "step": 1290
    },
    {
      "epoch": 0.1291,
      "grad_norm": 9.059029579162598,
      "learning_rate": 4.398484848484849e-05,
      "loss": 0.2979,
      "step": 1291
    },
    {
      "epoch": 0.1292,
      "grad_norm": 10.162759780883789,
      "learning_rate": 4.397979797979798e-05,
      "loss": 0.3522,
      "step": 1292
    },
    {
      "epoch": 0.1293,
      "grad_norm": 4.220986366271973,
      "learning_rate": 4.397474747474747e-05,
      "loss": 0.2519,
      "step": 1293
    },
    {
      "epoch": 0.1294,
      "grad_norm": 6.055702209472656,
      "learning_rate": 4.3969696969696975e-05,
      "loss": 0.1066,
      "step": 1294
    },
    {
      "epoch": 0.1295,
      "grad_norm": 9.936673164367676,
      "learning_rate": 4.396464646464647e-05,
      "loss": 0.2802,
      "step": 1295
    },
    {
      "epoch": 0.1296,
      "grad_norm": 2.1967952251434326,
      "learning_rate": 4.395959595959596e-05,
      "loss": 0.0769,
      "step": 1296
    },
    {
      "epoch": 0.1297,
      "grad_norm": 5.135561943054199,
      "learning_rate": 4.3954545454545456e-05,
      "loss": 0.2413,
      "step": 1297
    },
    {
      "epoch": 0.1298,
      "grad_norm": 4.98421049118042,
      "learning_rate": 4.394949494949495e-05,
      "loss": 0.6846,
      "step": 1298
    },
    {
      "epoch": 0.1299,
      "grad_norm": 14.037317276000977,
      "learning_rate": 4.394444444444445e-05,
      "loss": 0.4088,
      "step": 1299
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.805152416229248,
      "learning_rate": 4.3939393939393944e-05,
      "loss": 0.3408,
      "step": 1300
    },
    {
      "epoch": 0.1301,
      "grad_norm": 13.31623649597168,
      "learning_rate": 4.393434343434344e-05,
      "loss": 0.0899,
      "step": 1301
    },
    {
      "epoch": 0.1302,
      "grad_norm": 6.111833572387695,
      "learning_rate": 4.392929292929293e-05,
      "loss": 0.0493,
      "step": 1302
    },
    {
      "epoch": 0.1303,
      "grad_norm": 18.687177658081055,
      "learning_rate": 4.392424242424243e-05,
      "loss": 0.3506,
      "step": 1303
    },
    {
      "epoch": 0.1304,
      "grad_norm": 3.926973342895508,
      "learning_rate": 4.391919191919192e-05,
      "loss": 0.1703,
      "step": 1304
    },
    {
      "epoch": 0.1305,
      "grad_norm": 5.7456231117248535,
      "learning_rate": 4.391414141414142e-05,
      "loss": 0.2712,
      "step": 1305
    },
    {
      "epoch": 0.1306,
      "grad_norm": 6.249138355255127,
      "learning_rate": 4.390909090909091e-05,
      "loss": 0.3754,
      "step": 1306
    },
    {
      "epoch": 0.1307,
      "grad_norm": 6.590536117553711,
      "learning_rate": 4.390404040404041e-05,
      "loss": 0.263,
      "step": 1307
    },
    {
      "epoch": 0.1308,
      "grad_norm": 6.49874210357666,
      "learning_rate": 4.38989898989899e-05,
      "loss": 0.2706,
      "step": 1308
    },
    {
      "epoch": 0.1309,
      "grad_norm": 4.8616814613342285,
      "learning_rate": 4.38939393939394e-05,
      "loss": 0.2714,
      "step": 1309
    },
    {
      "epoch": 0.131,
      "grad_norm": 25.424386978149414,
      "learning_rate": 4.388888888888889e-05,
      "loss": 0.5086,
      "step": 1310
    },
    {
      "epoch": 0.1311,
      "grad_norm": 2.312626600265503,
      "learning_rate": 4.3883838383838386e-05,
      "loss": 0.0567,
      "step": 1311
    },
    {
      "epoch": 0.1312,
      "grad_norm": 13.389031410217285,
      "learning_rate": 4.387878787878788e-05,
      "loss": 0.3663,
      "step": 1312
    },
    {
      "epoch": 0.1313,
      "grad_norm": 9.517316818237305,
      "learning_rate": 4.387373737373738e-05,
      "loss": 0.2592,
      "step": 1313
    },
    {
      "epoch": 0.1314,
      "grad_norm": 15.286215782165527,
      "learning_rate": 4.386868686868687e-05,
      "loss": 0.3153,
      "step": 1314
    },
    {
      "epoch": 0.1315,
      "grad_norm": 14.566010475158691,
      "learning_rate": 4.386363636363637e-05,
      "loss": 0.7198,
      "step": 1315
    },
    {
      "epoch": 0.1316,
      "grad_norm": 12.479262351989746,
      "learning_rate": 4.385858585858586e-05,
      "loss": 0.7364,
      "step": 1316
    },
    {
      "epoch": 0.1317,
      "grad_norm": 5.911183834075928,
      "learning_rate": 4.3853535353535355e-05,
      "loss": 0.2676,
      "step": 1317
    },
    {
      "epoch": 0.1318,
      "grad_norm": 6.119168281555176,
      "learning_rate": 4.384848484848485e-05,
      "loss": 0.8388,
      "step": 1318
    },
    {
      "epoch": 0.1319,
      "grad_norm": 5.779472351074219,
      "learning_rate": 4.384343434343435e-05,
      "loss": 0.1835,
      "step": 1319
    },
    {
      "epoch": 0.132,
      "grad_norm": 10.971961975097656,
      "learning_rate": 4.383838383838384e-05,
      "loss": 0.3496,
      "step": 1320
    },
    {
      "epoch": 0.1321,
      "grad_norm": 4.795050144195557,
      "learning_rate": 4.383333333333334e-05,
      "loss": 0.1515,
      "step": 1321
    },
    {
      "epoch": 0.1322,
      "grad_norm": 2.9982450008392334,
      "learning_rate": 4.382828282828283e-05,
      "loss": 0.2574,
      "step": 1322
    },
    {
      "epoch": 0.1323,
      "grad_norm": 4.611610412597656,
      "learning_rate": 4.3823232323232325e-05,
      "loss": 0.2823,
      "step": 1323
    },
    {
      "epoch": 0.1324,
      "grad_norm": 5.2248148918151855,
      "learning_rate": 4.381818181818182e-05,
      "loss": 0.0589,
      "step": 1324
    },
    {
      "epoch": 0.1325,
      "grad_norm": 8.940719604492188,
      "learning_rate": 4.381313131313132e-05,
      "loss": 0.3906,
      "step": 1325
    },
    {
      "epoch": 0.1326,
      "grad_norm": 14.954671859741211,
      "learning_rate": 4.3808080808080806e-05,
      "loss": 0.4126,
      "step": 1326
    },
    {
      "epoch": 0.1327,
      "grad_norm": 4.939717769622803,
      "learning_rate": 4.380303030303031e-05,
      "loss": 0.2116,
      "step": 1327
    },
    {
      "epoch": 0.1328,
      "grad_norm": 11.728196144104004,
      "learning_rate": 4.37979797979798e-05,
      "loss": 0.7109,
      "step": 1328
    },
    {
      "epoch": 0.1329,
      "grad_norm": 7.671423435211182,
      "learning_rate": 4.3792929292929294e-05,
      "loss": 0.5424,
      "step": 1329
    },
    {
      "epoch": 0.133,
      "grad_norm": 4.968948841094971,
      "learning_rate": 4.378787878787879e-05,
      "loss": 0.1774,
      "step": 1330
    },
    {
      "epoch": 0.1331,
      "grad_norm": 8.962488174438477,
      "learning_rate": 4.3782828282828286e-05,
      "loss": 0.7355,
      "step": 1331
    },
    {
      "epoch": 0.1332,
      "grad_norm": 5.736128330230713,
      "learning_rate": 4.377777777777778e-05,
      "loss": 0.4189,
      "step": 1332
    },
    {
      "epoch": 0.1333,
      "grad_norm": 4.680136680603027,
      "learning_rate": 4.377272727272728e-05,
      "loss": 0.5398,
      "step": 1333
    },
    {
      "epoch": 0.1334,
      "grad_norm": 9.091719627380371,
      "learning_rate": 4.376767676767677e-05,
      "loss": 0.6167,
      "step": 1334
    },
    {
      "epoch": 0.1335,
      "grad_norm": 35.11933135986328,
      "learning_rate": 4.376262626262626e-05,
      "loss": 0.3548,
      "step": 1335
    },
    {
      "epoch": 0.1336,
      "grad_norm": 7.413740634918213,
      "learning_rate": 4.375757575757576e-05,
      "loss": 0.2677,
      "step": 1336
    },
    {
      "epoch": 0.1337,
      "grad_norm": 5.657708644866943,
      "learning_rate": 4.3752525252525255e-05,
      "loss": 0.2797,
      "step": 1337
    },
    {
      "epoch": 0.1338,
      "grad_norm": 13.687578201293945,
      "learning_rate": 4.374747474747475e-05,
      "loss": 0.3945,
      "step": 1338
    },
    {
      "epoch": 0.1339,
      "grad_norm": 9.862324714660645,
      "learning_rate": 4.374242424242425e-05,
      "loss": 0.3407,
      "step": 1339
    },
    {
      "epoch": 0.134,
      "grad_norm": 4.7790021896362305,
      "learning_rate": 4.3737373737373736e-05,
      "loss": 0.2202,
      "step": 1340
    },
    {
      "epoch": 0.1341,
      "grad_norm": 8.168787956237793,
      "learning_rate": 4.373232323232324e-05,
      "loss": 0.8579,
      "step": 1341
    },
    {
      "epoch": 0.1342,
      "grad_norm": 7.860549449920654,
      "learning_rate": 4.372727272727273e-05,
      "loss": 0.4639,
      "step": 1342
    },
    {
      "epoch": 0.1343,
      "grad_norm": 8.82792854309082,
      "learning_rate": 4.3722222222222224e-05,
      "loss": 0.5619,
      "step": 1343
    },
    {
      "epoch": 0.1344,
      "grad_norm": 16.86528968811035,
      "learning_rate": 4.371717171717172e-05,
      "loss": 0.6086,
      "step": 1344
    },
    {
      "epoch": 0.1345,
      "grad_norm": 1.7972744703292847,
      "learning_rate": 4.3712121212121216e-05,
      "loss": 0.0279,
      "step": 1345
    },
    {
      "epoch": 0.1346,
      "grad_norm": 3.8329432010650635,
      "learning_rate": 4.3707070707070705e-05,
      "loss": 0.0914,
      "step": 1346
    },
    {
      "epoch": 0.1347,
      "grad_norm": 7.668546676635742,
      "learning_rate": 4.370202020202021e-05,
      "loss": 0.0728,
      "step": 1347
    },
    {
      "epoch": 0.1348,
      "grad_norm": 9.323616981506348,
      "learning_rate": 4.36969696969697e-05,
      "loss": 0.1018,
      "step": 1348
    },
    {
      "epoch": 0.1349,
      "grad_norm": 19.29063606262207,
      "learning_rate": 4.369191919191919e-05,
      "loss": 0.3579,
      "step": 1349
    },
    {
      "epoch": 0.135,
      "grad_norm": 4.020677089691162,
      "learning_rate": 4.368686868686869e-05,
      "loss": 0.0779,
      "step": 1350
    },
    {
      "epoch": 0.1351,
      "grad_norm": 11.977575302124023,
      "learning_rate": 4.3681818181818185e-05,
      "loss": 0.458,
      "step": 1351
    },
    {
      "epoch": 0.1352,
      "grad_norm": 3.1319525241851807,
      "learning_rate": 4.3676767676767674e-05,
      "loss": 0.1634,
      "step": 1352
    },
    {
      "epoch": 0.1353,
      "grad_norm": 31.928936004638672,
      "learning_rate": 4.367171717171718e-05,
      "loss": 0.1969,
      "step": 1353
    },
    {
      "epoch": 0.1354,
      "grad_norm": 3.9399070739746094,
      "learning_rate": 4.3666666666666666e-05,
      "loss": 0.1682,
      "step": 1354
    },
    {
      "epoch": 0.1355,
      "grad_norm": 5.212310791015625,
      "learning_rate": 4.366161616161616e-05,
      "loss": 0.0996,
      "step": 1355
    },
    {
      "epoch": 0.1356,
      "grad_norm": 5.016822338104248,
      "learning_rate": 4.365656565656566e-05,
      "loss": 0.3085,
      "step": 1356
    },
    {
      "epoch": 0.1357,
      "grad_norm": 2.816803216934204,
      "learning_rate": 4.3651515151515154e-05,
      "loss": 0.0449,
      "step": 1357
    },
    {
      "epoch": 0.1358,
      "grad_norm": 6.835114002227783,
      "learning_rate": 4.3646464646464644e-05,
      "loss": 0.348,
      "step": 1358
    },
    {
      "epoch": 0.1359,
      "grad_norm": 8.686152458190918,
      "learning_rate": 4.3641414141414146e-05,
      "loss": 0.2435,
      "step": 1359
    },
    {
      "epoch": 0.136,
      "grad_norm": 2.205014944076538,
      "learning_rate": 4.3636363636363636e-05,
      "loss": 0.0396,
      "step": 1360
    },
    {
      "epoch": 0.1361,
      "grad_norm": 3.6454286575317383,
      "learning_rate": 4.363131313131313e-05,
      "loss": 0.1565,
      "step": 1361
    },
    {
      "epoch": 0.1362,
      "grad_norm": 8.434383392333984,
      "learning_rate": 4.362626262626263e-05,
      "loss": 0.3028,
      "step": 1362
    },
    {
      "epoch": 0.1363,
      "grad_norm": 9.068832397460938,
      "learning_rate": 4.3621212121212124e-05,
      "loss": 0.4142,
      "step": 1363
    },
    {
      "epoch": 0.1364,
      "grad_norm": 6.26732873916626,
      "learning_rate": 4.361616161616161e-05,
      "loss": 0.1887,
      "step": 1364
    },
    {
      "epoch": 0.1365,
      "grad_norm": 5.351733207702637,
      "learning_rate": 4.3611111111111116e-05,
      "loss": 0.1601,
      "step": 1365
    },
    {
      "epoch": 0.1366,
      "grad_norm": 17.0987491607666,
      "learning_rate": 4.3606060606060605e-05,
      "loss": 0.1123,
      "step": 1366
    },
    {
      "epoch": 0.1367,
      "grad_norm": 5.360751628875732,
      "learning_rate": 4.36010101010101e-05,
      "loss": 0.7033,
      "step": 1367
    },
    {
      "epoch": 0.1368,
      "grad_norm": 1.787258505821228,
      "learning_rate": 4.35959595959596e-05,
      "loss": 0.0754,
      "step": 1368
    },
    {
      "epoch": 0.1369,
      "grad_norm": 20.680818557739258,
      "learning_rate": 4.359090909090909e-05,
      "loss": 0.884,
      "step": 1369
    },
    {
      "epoch": 0.137,
      "grad_norm": 6.5299482345581055,
      "learning_rate": 4.358585858585859e-05,
      "loss": 0.1109,
      "step": 1370
    },
    {
      "epoch": 0.1371,
      "grad_norm": 4.60192346572876,
      "learning_rate": 4.3580808080808085e-05,
      "loss": 0.3814,
      "step": 1371
    },
    {
      "epoch": 0.1372,
      "grad_norm": 21.077770233154297,
      "learning_rate": 4.3575757575757574e-05,
      "loss": 0.3892,
      "step": 1372
    },
    {
      "epoch": 0.1373,
      "grad_norm": 10.666337013244629,
      "learning_rate": 4.357070707070707e-05,
      "loss": 1.1837,
      "step": 1373
    },
    {
      "epoch": 0.1374,
      "grad_norm": 8.153608322143555,
      "learning_rate": 4.3565656565656566e-05,
      "loss": 0.515,
      "step": 1374
    },
    {
      "epoch": 0.1375,
      "grad_norm": 10.197620391845703,
      "learning_rate": 4.356060606060606e-05,
      "loss": 0.2309,
      "step": 1375
    },
    {
      "epoch": 0.1376,
      "grad_norm": 2.0087814331054688,
      "learning_rate": 4.355555555555556e-05,
      "loss": 0.0444,
      "step": 1376
    },
    {
      "epoch": 0.1377,
      "grad_norm": 5.640796184539795,
      "learning_rate": 4.3550505050505054e-05,
      "loss": 0.3226,
      "step": 1377
    },
    {
      "epoch": 0.1378,
      "grad_norm": 5.304194927215576,
      "learning_rate": 4.354545454545454e-05,
      "loss": 0.2234,
      "step": 1378
    },
    {
      "epoch": 0.1379,
      "grad_norm": 40.4586296081543,
      "learning_rate": 4.3540404040404046e-05,
      "loss": 0.2534,
      "step": 1379
    },
    {
      "epoch": 0.138,
      "grad_norm": 4.144929885864258,
      "learning_rate": 4.3535353535353535e-05,
      "loss": 0.1533,
      "step": 1380
    },
    {
      "epoch": 0.1381,
      "grad_norm": 9.435275077819824,
      "learning_rate": 4.353030303030303e-05,
      "loss": 0.513,
      "step": 1381
    },
    {
      "epoch": 0.1382,
      "grad_norm": 6.468306064605713,
      "learning_rate": 4.352525252525253e-05,
      "loss": 0.6215,
      "step": 1382
    },
    {
      "epoch": 0.1383,
      "grad_norm": 8.39622974395752,
      "learning_rate": 4.352020202020202e-05,
      "loss": 0.0925,
      "step": 1383
    },
    {
      "epoch": 0.1384,
      "grad_norm": 7.853567123413086,
      "learning_rate": 4.351515151515152e-05,
      "loss": 0.4621,
      "step": 1384
    },
    {
      "epoch": 0.1385,
      "grad_norm": 23.85093116760254,
      "learning_rate": 4.3510101010101015e-05,
      "loss": 1.4664,
      "step": 1385
    },
    {
      "epoch": 0.1386,
      "grad_norm": 7.760303497314453,
      "learning_rate": 4.350505050505051e-05,
      "loss": 0.2987,
      "step": 1386
    },
    {
      "epoch": 0.1387,
      "grad_norm": 12.9166898727417,
      "learning_rate": 4.35e-05,
      "loss": 1.0176,
      "step": 1387
    },
    {
      "epoch": 0.1388,
      "grad_norm": 8.086881637573242,
      "learning_rate": 4.3494949494949496e-05,
      "loss": 0.236,
      "step": 1388
    },
    {
      "epoch": 0.1389,
      "grad_norm": 4.423867225646973,
      "learning_rate": 4.348989898989899e-05,
      "loss": 0.1278,
      "step": 1389
    },
    {
      "epoch": 0.139,
      "grad_norm": 5.831171035766602,
      "learning_rate": 4.348484848484849e-05,
      "loss": 0.3659,
      "step": 1390
    },
    {
      "epoch": 0.1391,
      "grad_norm": 8.338103294372559,
      "learning_rate": 4.3479797979797984e-05,
      "loss": 0.1454,
      "step": 1391
    },
    {
      "epoch": 0.1392,
      "grad_norm": 5.738827228546143,
      "learning_rate": 4.347474747474748e-05,
      "loss": 0.6124,
      "step": 1392
    },
    {
      "epoch": 0.1393,
      "grad_norm": 8.433737754821777,
      "learning_rate": 4.346969696969697e-05,
      "loss": 0.5094,
      "step": 1393
    },
    {
      "epoch": 0.1394,
      "grad_norm": 6.175227642059326,
      "learning_rate": 4.346464646464647e-05,
      "loss": 0.4065,
      "step": 1394
    },
    {
      "epoch": 0.1395,
      "grad_norm": 6.4999189376831055,
      "learning_rate": 4.345959595959596e-05,
      "loss": 1.329,
      "step": 1395
    },
    {
      "epoch": 0.1396,
      "grad_norm": 8.801280975341797,
      "learning_rate": 4.345454545454546e-05,
      "loss": 0.411,
      "step": 1396
    },
    {
      "epoch": 0.1397,
      "grad_norm": 24.540620803833008,
      "learning_rate": 4.344949494949495e-05,
      "loss": 0.75,
      "step": 1397
    },
    {
      "epoch": 0.1398,
      "grad_norm": 5.729910850524902,
      "learning_rate": 4.344444444444445e-05,
      "loss": 0.6272,
      "step": 1398
    },
    {
      "epoch": 0.1399,
      "grad_norm": 4.245127201080322,
      "learning_rate": 4.343939393939394e-05,
      "loss": 0.3477,
      "step": 1399
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.506329536437988,
      "learning_rate": 4.343434343434344e-05,
      "loss": 0.4903,
      "step": 1400
    },
    {
      "epoch": 0.1401,
      "grad_norm": 5.857060432434082,
      "learning_rate": 4.342929292929293e-05,
      "loss": 0.2884,
      "step": 1401
    },
    {
      "epoch": 0.1402,
      "grad_norm": 3.5903615951538086,
      "learning_rate": 4.3424242424242427e-05,
      "loss": 0.107,
      "step": 1402
    },
    {
      "epoch": 0.1403,
      "grad_norm": 7.293788909912109,
      "learning_rate": 4.341919191919192e-05,
      "loss": 0.1451,
      "step": 1403
    },
    {
      "epoch": 0.1404,
      "grad_norm": 7.704268455505371,
      "learning_rate": 4.341414141414142e-05,
      "loss": 0.2531,
      "step": 1404
    },
    {
      "epoch": 0.1405,
      "grad_norm": 1.743229627609253,
      "learning_rate": 4.340909090909091e-05,
      "loss": 0.4463,
      "step": 1405
    },
    {
      "epoch": 0.1406,
      "grad_norm": 3.9533469676971436,
      "learning_rate": 4.340404040404041e-05,
      "loss": 0.2764,
      "step": 1406
    },
    {
      "epoch": 0.1407,
      "grad_norm": 7.215447425842285,
      "learning_rate": 4.33989898989899e-05,
      "loss": 0.5127,
      "step": 1407
    },
    {
      "epoch": 0.1408,
      "grad_norm": 4.007427215576172,
      "learning_rate": 4.3393939393939396e-05,
      "loss": 0.0985,
      "step": 1408
    },
    {
      "epoch": 0.1409,
      "grad_norm": 11.303470611572266,
      "learning_rate": 4.338888888888889e-05,
      "loss": 0.8221,
      "step": 1409
    },
    {
      "epoch": 0.141,
      "grad_norm": 10.285551071166992,
      "learning_rate": 4.338383838383839e-05,
      "loss": 0.1572,
      "step": 1410
    },
    {
      "epoch": 0.1411,
      "grad_norm": 8.734354019165039,
      "learning_rate": 4.337878787878788e-05,
      "loss": 0.7642,
      "step": 1411
    },
    {
      "epoch": 0.1412,
      "grad_norm": 3.602816104888916,
      "learning_rate": 4.337373737373738e-05,
      "loss": 0.0369,
      "step": 1412
    },
    {
      "epoch": 0.1413,
      "grad_norm": 4.219107151031494,
      "learning_rate": 4.336868686868687e-05,
      "loss": 0.7007,
      "step": 1413
    },
    {
      "epoch": 0.1414,
      "grad_norm": 5.221271514892578,
      "learning_rate": 4.3363636363636365e-05,
      "loss": 0.4064,
      "step": 1414
    },
    {
      "epoch": 0.1415,
      "grad_norm": 10.191288948059082,
      "learning_rate": 4.335858585858586e-05,
      "loss": 0.542,
      "step": 1415
    },
    {
      "epoch": 0.1416,
      "grad_norm": 4.913425922393799,
      "learning_rate": 4.335353535353536e-05,
      "loss": 0.2088,
      "step": 1416
    },
    {
      "epoch": 0.1417,
      "grad_norm": 3.630795955657959,
      "learning_rate": 4.3348484848484846e-05,
      "loss": 0.2707,
      "step": 1417
    },
    {
      "epoch": 0.1418,
      "grad_norm": 7.9361090660095215,
      "learning_rate": 4.334343434343435e-05,
      "loss": 1.4392,
      "step": 1418
    },
    {
      "epoch": 0.1419,
      "grad_norm": 7.5339741706848145,
      "learning_rate": 4.333838383838384e-05,
      "loss": 0.3626,
      "step": 1419
    },
    {
      "epoch": 0.142,
      "grad_norm": 3.126086950302124,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.1309,
      "step": 1420
    },
    {
      "epoch": 0.1421,
      "grad_norm": 6.512105941772461,
      "learning_rate": 4.332828282828283e-05,
      "loss": 0.4885,
      "step": 1421
    },
    {
      "epoch": 0.1422,
      "grad_norm": 7.473788261413574,
      "learning_rate": 4.3323232323232326e-05,
      "loss": 0.2646,
      "step": 1422
    },
    {
      "epoch": 0.1423,
      "grad_norm": 7.343514919281006,
      "learning_rate": 4.331818181818182e-05,
      "loss": 0.26,
      "step": 1423
    },
    {
      "epoch": 0.1424,
      "grad_norm": 6.532415390014648,
      "learning_rate": 4.331313131313132e-05,
      "loss": 0.4382,
      "step": 1424
    },
    {
      "epoch": 0.1425,
      "grad_norm": 6.421276569366455,
      "learning_rate": 4.330808080808081e-05,
      "loss": 0.8816,
      "step": 1425
    },
    {
      "epoch": 0.1426,
      "grad_norm": 9.461780548095703,
      "learning_rate": 4.33030303030303e-05,
      "loss": 0.4592,
      "step": 1426
    },
    {
      "epoch": 0.1427,
      "grad_norm": 19.428728103637695,
      "learning_rate": 4.32979797979798e-05,
      "loss": 0.74,
      "step": 1427
    },
    {
      "epoch": 0.1428,
      "grad_norm": 9.862836837768555,
      "learning_rate": 4.3292929292929295e-05,
      "loss": 0.4721,
      "step": 1428
    },
    {
      "epoch": 0.1429,
      "grad_norm": 10.136397361755371,
      "learning_rate": 4.328787878787879e-05,
      "loss": 0.6084,
      "step": 1429
    },
    {
      "epoch": 0.143,
      "grad_norm": 9.468661308288574,
      "learning_rate": 4.328282828282829e-05,
      "loss": 0.5581,
      "step": 1430
    },
    {
      "epoch": 0.1431,
      "grad_norm": 10.0414457321167,
      "learning_rate": 4.3277777777777776e-05,
      "loss": 0.0752,
      "step": 1431
    },
    {
      "epoch": 0.1432,
      "grad_norm": 7.389156818389893,
      "learning_rate": 4.327272727272728e-05,
      "loss": 0.3064,
      "step": 1432
    },
    {
      "epoch": 0.1433,
      "grad_norm": 4.942269325256348,
      "learning_rate": 4.326767676767677e-05,
      "loss": 0.2423,
      "step": 1433
    },
    {
      "epoch": 0.1434,
      "grad_norm": 2.6441099643707275,
      "learning_rate": 4.3262626262626264e-05,
      "loss": 0.0481,
      "step": 1434
    },
    {
      "epoch": 0.1435,
      "grad_norm": 8.25572681427002,
      "learning_rate": 4.325757575757576e-05,
      "loss": 0.5143,
      "step": 1435
    },
    {
      "epoch": 0.1436,
      "grad_norm": 8.341739654541016,
      "learning_rate": 4.3252525252525256e-05,
      "loss": 0.273,
      "step": 1436
    },
    {
      "epoch": 0.1437,
      "grad_norm": 8.19671630859375,
      "learning_rate": 4.3247474747474746e-05,
      "loss": 0.234,
      "step": 1437
    },
    {
      "epoch": 0.1438,
      "grad_norm": 3.4410104751586914,
      "learning_rate": 4.324242424242425e-05,
      "loss": 0.0629,
      "step": 1438
    },
    {
      "epoch": 0.1439,
      "grad_norm": 33.81544876098633,
      "learning_rate": 4.323737373737374e-05,
      "loss": 0.3117,
      "step": 1439
    },
    {
      "epoch": 0.144,
      "grad_norm": 4.328913688659668,
      "learning_rate": 4.3232323232323234e-05,
      "loss": 0.0598,
      "step": 1440
    },
    {
      "epoch": 0.1441,
      "grad_norm": 8.803826332092285,
      "learning_rate": 4.322727272727273e-05,
      "loss": 0.8094,
      "step": 1441
    },
    {
      "epoch": 0.1442,
      "grad_norm": 4.734263896942139,
      "learning_rate": 4.3222222222222226e-05,
      "loss": 0.0607,
      "step": 1442
    },
    {
      "epoch": 0.1443,
      "grad_norm": 5.3934760093688965,
      "learning_rate": 4.3217171717171715e-05,
      "loss": 0.1786,
      "step": 1443
    },
    {
      "epoch": 0.1444,
      "grad_norm": 3.9927473068237305,
      "learning_rate": 4.321212121212122e-05,
      "loss": 0.5326,
      "step": 1444
    },
    {
      "epoch": 0.1445,
      "grad_norm": 6.196213245391846,
      "learning_rate": 4.320707070707071e-05,
      "loss": 0.5993,
      "step": 1445
    },
    {
      "epoch": 0.1446,
      "grad_norm": 4.581137180328369,
      "learning_rate": 4.32020202020202e-05,
      "loss": 0.1542,
      "step": 1446
    },
    {
      "epoch": 0.1447,
      "grad_norm": 7.2821855545043945,
      "learning_rate": 4.31969696969697e-05,
      "loss": 0.5233,
      "step": 1447
    },
    {
      "epoch": 0.1448,
      "grad_norm": 11.308349609375,
      "learning_rate": 4.3191919191919195e-05,
      "loss": 0.2803,
      "step": 1448
    },
    {
      "epoch": 0.1449,
      "grad_norm": 50.67920684814453,
      "learning_rate": 4.3186868686868684e-05,
      "loss": 0.5341,
      "step": 1449
    },
    {
      "epoch": 0.145,
      "grad_norm": 3.88204026222229,
      "learning_rate": 4.318181818181819e-05,
      "loss": 0.1299,
      "step": 1450
    },
    {
      "epoch": 0.1451,
      "grad_norm": 16.210590362548828,
      "learning_rate": 4.3176767676767676e-05,
      "loss": 0.2434,
      "step": 1451
    },
    {
      "epoch": 0.1452,
      "grad_norm": 9.559014320373535,
      "learning_rate": 4.317171717171717e-05,
      "loss": 0.6671,
      "step": 1452
    },
    {
      "epoch": 0.1453,
      "grad_norm": 23.914873123168945,
      "learning_rate": 4.316666666666667e-05,
      "loss": 0.5168,
      "step": 1453
    },
    {
      "epoch": 0.1454,
      "grad_norm": 4.069071292877197,
      "learning_rate": 4.3161616161616164e-05,
      "loss": 0.0547,
      "step": 1454
    },
    {
      "epoch": 0.1455,
      "grad_norm": 6.175732612609863,
      "learning_rate": 4.315656565656565e-05,
      "loss": 0.3716,
      "step": 1455
    },
    {
      "epoch": 0.1456,
      "grad_norm": 26.09285545349121,
      "learning_rate": 4.3151515151515156e-05,
      "loss": 0.6356,
      "step": 1456
    },
    {
      "epoch": 0.1457,
      "grad_norm": 3.2828948497772217,
      "learning_rate": 4.3146464646464645e-05,
      "loss": 0.0512,
      "step": 1457
    },
    {
      "epoch": 0.1458,
      "grad_norm": 23.764963150024414,
      "learning_rate": 4.314141414141414e-05,
      "loss": 0.4594,
      "step": 1458
    },
    {
      "epoch": 0.1459,
      "grad_norm": 6.513148784637451,
      "learning_rate": 4.313636363636364e-05,
      "loss": 0.6715,
      "step": 1459
    },
    {
      "epoch": 0.146,
      "grad_norm": 4.478639602661133,
      "learning_rate": 4.313131313131313e-05,
      "loss": 0.5378,
      "step": 1460
    },
    {
      "epoch": 0.1461,
      "grad_norm": 17.077129364013672,
      "learning_rate": 4.312626262626263e-05,
      "loss": 0.5307,
      "step": 1461
    },
    {
      "epoch": 0.1462,
      "grad_norm": 36.59617614746094,
      "learning_rate": 4.3121212121212125e-05,
      "loss": 0.6138,
      "step": 1462
    },
    {
      "epoch": 0.1463,
      "grad_norm": 5.438115119934082,
      "learning_rate": 4.3116161616161614e-05,
      "loss": 0.0302,
      "step": 1463
    },
    {
      "epoch": 0.1464,
      "grad_norm": 3.7317874431610107,
      "learning_rate": 4.311111111111111e-05,
      "loss": 0.037,
      "step": 1464
    },
    {
      "epoch": 0.1465,
      "grad_norm": 7.099480628967285,
      "learning_rate": 4.3106060606060606e-05,
      "loss": 0.2071,
      "step": 1465
    },
    {
      "epoch": 0.1466,
      "grad_norm": 2.586479663848877,
      "learning_rate": 4.31010101010101e-05,
      "loss": 0.0807,
      "step": 1466
    },
    {
      "epoch": 0.1467,
      "grad_norm": 17.309226989746094,
      "learning_rate": 4.30959595959596e-05,
      "loss": 0.4634,
      "step": 1467
    },
    {
      "epoch": 0.1468,
      "grad_norm": 4.248676300048828,
      "learning_rate": 4.3090909090909094e-05,
      "loss": 0.1327,
      "step": 1468
    },
    {
      "epoch": 0.1469,
      "grad_norm": 7.472452163696289,
      "learning_rate": 4.308585858585859e-05,
      "loss": 0.4468,
      "step": 1469
    },
    {
      "epoch": 0.147,
      "grad_norm": 6.667840003967285,
      "learning_rate": 4.308080808080808e-05,
      "loss": 0.6258,
      "step": 1470
    },
    {
      "epoch": 0.1471,
      "grad_norm": 9.167384147644043,
      "learning_rate": 4.307575757575758e-05,
      "loss": 1.0976,
      "step": 1471
    },
    {
      "epoch": 0.1472,
      "grad_norm": 15.021002769470215,
      "learning_rate": 4.307070707070707e-05,
      "loss": 0.8687,
      "step": 1472
    },
    {
      "epoch": 0.1473,
      "grad_norm": 5.59685754776001,
      "learning_rate": 4.306565656565657e-05,
      "loss": 0.1807,
      "step": 1473
    },
    {
      "epoch": 0.1474,
      "grad_norm": 5.1817240715026855,
      "learning_rate": 4.306060606060606e-05,
      "loss": 0.3897,
      "step": 1474
    },
    {
      "epoch": 0.1475,
      "grad_norm": 3.6456823348999023,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.3132,
      "step": 1475
    },
    {
      "epoch": 0.1476,
      "grad_norm": 8.992445945739746,
      "learning_rate": 4.3050505050505055e-05,
      "loss": 0.7563,
      "step": 1476
    },
    {
      "epoch": 0.1477,
      "grad_norm": 9.952363967895508,
      "learning_rate": 4.304545454545455e-05,
      "loss": 0.5192,
      "step": 1477
    },
    {
      "epoch": 0.1478,
      "grad_norm": 7.6289777755737305,
      "learning_rate": 4.304040404040404e-05,
      "loss": 0.1586,
      "step": 1478
    },
    {
      "epoch": 0.1479,
      "grad_norm": 3.2134172916412354,
      "learning_rate": 4.3035353535353536e-05,
      "loss": 0.2095,
      "step": 1479
    },
    {
      "epoch": 0.148,
      "grad_norm": 10.25228500366211,
      "learning_rate": 4.303030303030303e-05,
      "loss": 0.1172,
      "step": 1480
    },
    {
      "epoch": 0.1481,
      "grad_norm": 3.082158327102661,
      "learning_rate": 4.302525252525253e-05,
      "loss": 0.5405,
      "step": 1481
    },
    {
      "epoch": 0.1482,
      "grad_norm": 5.241281032562256,
      "learning_rate": 4.3020202020202024e-05,
      "loss": 0.2105,
      "step": 1482
    },
    {
      "epoch": 0.1483,
      "grad_norm": 4.3411030769348145,
      "learning_rate": 4.301515151515152e-05,
      "loss": 0.2538,
      "step": 1483
    },
    {
      "epoch": 0.1484,
      "grad_norm": 3.2876789569854736,
      "learning_rate": 4.301010101010101e-05,
      "loss": 0.0705,
      "step": 1484
    },
    {
      "epoch": 0.1485,
      "grad_norm": 8.532051086425781,
      "learning_rate": 4.300505050505051e-05,
      "loss": 0.942,
      "step": 1485
    },
    {
      "epoch": 0.1486,
      "grad_norm": 4.407407283782959,
      "learning_rate": 4.3e-05,
      "loss": 0.1745,
      "step": 1486
    },
    {
      "epoch": 0.1487,
      "grad_norm": 3.286289691925049,
      "learning_rate": 4.29949494949495e-05,
      "loss": 0.5387,
      "step": 1487
    },
    {
      "epoch": 0.1488,
      "grad_norm": 15.809852600097656,
      "learning_rate": 4.2989898989898994e-05,
      "loss": 0.3083,
      "step": 1488
    },
    {
      "epoch": 0.1489,
      "grad_norm": 8.969032287597656,
      "learning_rate": 4.298484848484849e-05,
      "loss": 0.1544,
      "step": 1489
    },
    {
      "epoch": 0.149,
      "grad_norm": 9.35049819946289,
      "learning_rate": 4.297979797979798e-05,
      "loss": 0.2667,
      "step": 1490
    },
    {
      "epoch": 0.1491,
      "grad_norm": 6.6220245361328125,
      "learning_rate": 4.297474747474748e-05,
      "loss": 0.3914,
      "step": 1491
    },
    {
      "epoch": 0.1492,
      "grad_norm": 5.5303239822387695,
      "learning_rate": 4.296969696969697e-05,
      "loss": 0.724,
      "step": 1492
    },
    {
      "epoch": 0.1493,
      "grad_norm": 2.2613861560821533,
      "learning_rate": 4.296464646464647e-05,
      "loss": 0.0267,
      "step": 1493
    },
    {
      "epoch": 0.1494,
      "grad_norm": 3.7318828105926514,
      "learning_rate": 4.295959595959596e-05,
      "loss": 0.1308,
      "step": 1494
    },
    {
      "epoch": 0.1495,
      "grad_norm": 6.114651203155518,
      "learning_rate": 4.295454545454546e-05,
      "loss": 0.2511,
      "step": 1495
    },
    {
      "epoch": 0.1496,
      "grad_norm": 7.221536159515381,
      "learning_rate": 4.294949494949495e-05,
      "loss": 0.7645,
      "step": 1496
    },
    {
      "epoch": 0.1497,
      "grad_norm": 9.114614486694336,
      "learning_rate": 4.294444444444445e-05,
      "loss": 0.4095,
      "step": 1497
    },
    {
      "epoch": 0.1498,
      "grad_norm": 7.012844085693359,
      "learning_rate": 4.293939393939394e-05,
      "loss": 0.2202,
      "step": 1498
    },
    {
      "epoch": 0.1499,
      "grad_norm": 25.470645904541016,
      "learning_rate": 4.2934343434343436e-05,
      "loss": 0.6141,
      "step": 1499
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.279896259307861,
      "learning_rate": 4.292929292929293e-05,
      "loss": 0.3519,
      "step": 1500
    },
    {
      "epoch": 0.1501,
      "grad_norm": 7.947887897491455,
      "learning_rate": 4.292424242424243e-05,
      "loss": 0.069,
      "step": 1501
    },
    {
      "epoch": 0.1502,
      "grad_norm": 3.2142751216888428,
      "learning_rate": 4.291919191919192e-05,
      "loss": 0.2829,
      "step": 1502
    },
    {
      "epoch": 0.1503,
      "grad_norm": 6.186308860778809,
      "learning_rate": 4.291414141414142e-05,
      "loss": 0.057,
      "step": 1503
    },
    {
      "epoch": 0.1504,
      "grad_norm": 6.81362247467041,
      "learning_rate": 4.290909090909091e-05,
      "loss": 0.4249,
      "step": 1504
    },
    {
      "epoch": 0.1505,
      "grad_norm": 7.538792610168457,
      "learning_rate": 4.2904040404040405e-05,
      "loss": 1.1305,
      "step": 1505
    },
    {
      "epoch": 0.1506,
      "grad_norm": 6.21569299697876,
      "learning_rate": 4.28989898989899e-05,
      "loss": 0.3593,
      "step": 1506
    },
    {
      "epoch": 0.1507,
      "grad_norm": 12.673759460449219,
      "learning_rate": 4.28939393939394e-05,
      "loss": 0.2578,
      "step": 1507
    },
    {
      "epoch": 0.1508,
      "grad_norm": 5.957897663116455,
      "learning_rate": 4.2888888888888886e-05,
      "loss": 0.1777,
      "step": 1508
    },
    {
      "epoch": 0.1509,
      "grad_norm": 6.4707770347595215,
      "learning_rate": 4.288383838383839e-05,
      "loss": 0.4338,
      "step": 1509
    },
    {
      "epoch": 0.151,
      "grad_norm": 5.291756629943848,
      "learning_rate": 4.287878787878788e-05,
      "loss": 0.1828,
      "step": 1510
    },
    {
      "epoch": 0.1511,
      "grad_norm": 26.556415557861328,
      "learning_rate": 4.2873737373737374e-05,
      "loss": 0.7224,
      "step": 1511
    },
    {
      "epoch": 0.1512,
      "grad_norm": 4.817185878753662,
      "learning_rate": 4.286868686868687e-05,
      "loss": 0.6342,
      "step": 1512
    },
    {
      "epoch": 0.1513,
      "grad_norm": 5.514959335327148,
      "learning_rate": 4.2863636363636366e-05,
      "loss": 0.2873,
      "step": 1513
    },
    {
      "epoch": 0.1514,
      "grad_norm": 24.58442497253418,
      "learning_rate": 4.285858585858586e-05,
      "loss": 0.7183,
      "step": 1514
    },
    {
      "epoch": 0.1515,
      "grad_norm": 3.67441725730896,
      "learning_rate": 4.285353535353536e-05,
      "loss": 0.2256,
      "step": 1515
    },
    {
      "epoch": 0.1516,
      "grad_norm": 3.547889232635498,
      "learning_rate": 4.284848484848485e-05,
      "loss": 0.1621,
      "step": 1516
    },
    {
      "epoch": 0.1517,
      "grad_norm": 10.409196853637695,
      "learning_rate": 4.2843434343434343e-05,
      "loss": 0.2953,
      "step": 1517
    },
    {
      "epoch": 0.1518,
      "grad_norm": 19.14298439025879,
      "learning_rate": 4.283838383838384e-05,
      "loss": 0.2956,
      "step": 1518
    },
    {
      "epoch": 0.1519,
      "grad_norm": 9.037729263305664,
      "learning_rate": 4.2833333333333335e-05,
      "loss": 0.2347,
      "step": 1519
    },
    {
      "epoch": 0.152,
      "grad_norm": 9.270950317382812,
      "learning_rate": 4.282828282828283e-05,
      "loss": 0.1962,
      "step": 1520
    },
    {
      "epoch": 0.1521,
      "grad_norm": 7.499726295471191,
      "learning_rate": 4.282323232323233e-05,
      "loss": 0.3642,
      "step": 1521
    },
    {
      "epoch": 0.1522,
      "grad_norm": 4.1794939041137695,
      "learning_rate": 4.281818181818182e-05,
      "loss": 0.1287,
      "step": 1522
    },
    {
      "epoch": 0.1523,
      "grad_norm": 2.302596092224121,
      "learning_rate": 4.281313131313132e-05,
      "loss": 0.0604,
      "step": 1523
    },
    {
      "epoch": 0.1524,
      "grad_norm": 11.337059020996094,
      "learning_rate": 4.280808080808081e-05,
      "loss": 0.5762,
      "step": 1524
    },
    {
      "epoch": 0.1525,
      "grad_norm": 4.800475597381592,
      "learning_rate": 4.2803030303030305e-05,
      "loss": 0.488,
      "step": 1525
    },
    {
      "epoch": 0.1526,
      "grad_norm": 4.152645111083984,
      "learning_rate": 4.27979797979798e-05,
      "loss": 0.131,
      "step": 1526
    },
    {
      "epoch": 0.1527,
      "grad_norm": 6.27304744720459,
      "learning_rate": 4.2792929292929297e-05,
      "loss": 0.2998,
      "step": 1527
    },
    {
      "epoch": 0.1528,
      "grad_norm": 2.894152879714966,
      "learning_rate": 4.2787878787878786e-05,
      "loss": 0.1165,
      "step": 1528
    },
    {
      "epoch": 0.1529,
      "grad_norm": 6.460586071014404,
      "learning_rate": 4.278282828282829e-05,
      "loss": 0.7047,
      "step": 1529
    },
    {
      "epoch": 0.153,
      "grad_norm": 23.18444061279297,
      "learning_rate": 4.277777777777778e-05,
      "loss": 0.1843,
      "step": 1530
    },
    {
      "epoch": 0.1531,
      "grad_norm": 8.706353187561035,
      "learning_rate": 4.2772727272727274e-05,
      "loss": 0.2451,
      "step": 1531
    },
    {
      "epoch": 0.1532,
      "grad_norm": 7.80864953994751,
      "learning_rate": 4.276767676767677e-05,
      "loss": 0.3449,
      "step": 1532
    },
    {
      "epoch": 0.1533,
      "grad_norm": 4.059741020202637,
      "learning_rate": 4.2762626262626266e-05,
      "loss": 0.4859,
      "step": 1533
    },
    {
      "epoch": 0.1534,
      "grad_norm": 5.22976541519165,
      "learning_rate": 4.2757575757575755e-05,
      "loss": 0.3361,
      "step": 1534
    },
    {
      "epoch": 0.1535,
      "grad_norm": 4.700899124145508,
      "learning_rate": 4.275252525252526e-05,
      "loss": 0.5629,
      "step": 1535
    },
    {
      "epoch": 0.1536,
      "grad_norm": 7.955204010009766,
      "learning_rate": 4.274747474747475e-05,
      "loss": 0.1567,
      "step": 1536
    },
    {
      "epoch": 0.1537,
      "grad_norm": 7.358720779418945,
      "learning_rate": 4.274242424242424e-05,
      "loss": 0.3165,
      "step": 1537
    },
    {
      "epoch": 0.1538,
      "grad_norm": 6.936886787414551,
      "learning_rate": 4.273737373737374e-05,
      "loss": 0.1467,
      "step": 1538
    },
    {
      "epoch": 0.1539,
      "grad_norm": 7.461055755615234,
      "learning_rate": 4.2732323232323235e-05,
      "loss": 0.2203,
      "step": 1539
    },
    {
      "epoch": 0.154,
      "grad_norm": 15.448769569396973,
      "learning_rate": 4.2727272727272724e-05,
      "loss": 0.1415,
      "step": 1540
    },
    {
      "epoch": 0.1541,
      "grad_norm": 8.93331527709961,
      "learning_rate": 4.272222222222223e-05,
      "loss": 0.2787,
      "step": 1541
    },
    {
      "epoch": 0.1542,
      "grad_norm": 12.618269920349121,
      "learning_rate": 4.2717171717171716e-05,
      "loss": 0.2316,
      "step": 1542
    },
    {
      "epoch": 0.1543,
      "grad_norm": 7.656423568725586,
      "learning_rate": 4.271212121212121e-05,
      "loss": 0.5987,
      "step": 1543
    },
    {
      "epoch": 0.1544,
      "grad_norm": 7.900589466094971,
      "learning_rate": 4.270707070707071e-05,
      "loss": 0.4012,
      "step": 1544
    },
    {
      "epoch": 0.1545,
      "grad_norm": 2.1343820095062256,
      "learning_rate": 4.2702020202020204e-05,
      "loss": 0.0528,
      "step": 1545
    },
    {
      "epoch": 0.1546,
      "grad_norm": 5.457420825958252,
      "learning_rate": 4.269696969696969e-05,
      "loss": 0.1179,
      "step": 1546
    },
    {
      "epoch": 0.1547,
      "grad_norm": 8.106205940246582,
      "learning_rate": 4.2691919191919196e-05,
      "loss": 0.4445,
      "step": 1547
    },
    {
      "epoch": 0.1548,
      "grad_norm": 2.8535947799682617,
      "learning_rate": 4.2686868686868685e-05,
      "loss": 0.0303,
      "step": 1548
    },
    {
      "epoch": 0.1549,
      "grad_norm": 15.331854820251465,
      "learning_rate": 4.268181818181818e-05,
      "loss": 0.0749,
      "step": 1549
    },
    {
      "epoch": 0.155,
      "grad_norm": 8.649993896484375,
      "learning_rate": 4.267676767676768e-05,
      "loss": 0.3532,
      "step": 1550
    },
    {
      "epoch": 0.1551,
      "grad_norm": 9.131808280944824,
      "learning_rate": 4.267171717171717e-05,
      "loss": 0.92,
      "step": 1551
    },
    {
      "epoch": 0.1552,
      "grad_norm": 9.457326889038086,
      "learning_rate": 4.266666666666667e-05,
      "loss": 0.5677,
      "step": 1552
    },
    {
      "epoch": 0.1553,
      "grad_norm": 7.8602705001831055,
      "learning_rate": 4.2661616161616165e-05,
      "loss": 0.1712,
      "step": 1553
    },
    {
      "epoch": 0.1554,
      "grad_norm": 6.807799816131592,
      "learning_rate": 4.265656565656566e-05,
      "loss": 0.3186,
      "step": 1554
    },
    {
      "epoch": 0.1555,
      "grad_norm": 7.793826103210449,
      "learning_rate": 4.265151515151515e-05,
      "loss": 0.4938,
      "step": 1555
    },
    {
      "epoch": 0.1556,
      "grad_norm": 8.182577133178711,
      "learning_rate": 4.264646464646465e-05,
      "loss": 0.1896,
      "step": 1556
    },
    {
      "epoch": 0.1557,
      "grad_norm": 1.2277477979660034,
      "learning_rate": 4.264141414141414e-05,
      "loss": 0.0165,
      "step": 1557
    },
    {
      "epoch": 0.1558,
      "grad_norm": 14.473286628723145,
      "learning_rate": 4.263636363636364e-05,
      "loss": 0.3957,
      "step": 1558
    },
    {
      "epoch": 0.1559,
      "grad_norm": 4.292314052581787,
      "learning_rate": 4.2631313131313134e-05,
      "loss": 0.1719,
      "step": 1559
    },
    {
      "epoch": 0.156,
      "grad_norm": 9.440757751464844,
      "learning_rate": 4.262626262626263e-05,
      "loss": 0.0888,
      "step": 1560
    },
    {
      "epoch": 0.1561,
      "grad_norm": 4.398626327514648,
      "learning_rate": 4.262121212121212e-05,
      "loss": 0.24,
      "step": 1561
    },
    {
      "epoch": 0.1562,
      "grad_norm": 7.486146926879883,
      "learning_rate": 4.261616161616162e-05,
      "loss": 0.1928,
      "step": 1562
    },
    {
      "epoch": 0.1563,
      "grad_norm": 20.507328033447266,
      "learning_rate": 4.261111111111111e-05,
      "loss": 0.2861,
      "step": 1563
    },
    {
      "epoch": 0.1564,
      "grad_norm": 4.200345993041992,
      "learning_rate": 4.260606060606061e-05,
      "loss": 0.5555,
      "step": 1564
    },
    {
      "epoch": 0.1565,
      "grad_norm": 7.981325149536133,
      "learning_rate": 4.2601010101010104e-05,
      "loss": 0.4871,
      "step": 1565
    },
    {
      "epoch": 0.1566,
      "grad_norm": 7.725915908813477,
      "learning_rate": 4.25959595959596e-05,
      "loss": 0.9604,
      "step": 1566
    },
    {
      "epoch": 0.1567,
      "grad_norm": 12.717632293701172,
      "learning_rate": 4.2590909090909096e-05,
      "loss": 0.0879,
      "step": 1567
    },
    {
      "epoch": 0.1568,
      "grad_norm": 22.88072967529297,
      "learning_rate": 4.258585858585859e-05,
      "loss": 0.6635,
      "step": 1568
    },
    {
      "epoch": 0.1569,
      "grad_norm": 11.492015838623047,
      "learning_rate": 4.258080808080808e-05,
      "loss": 0.7442,
      "step": 1569
    },
    {
      "epoch": 0.157,
      "grad_norm": 5.1918182373046875,
      "learning_rate": 4.257575757575758e-05,
      "loss": 0.2955,
      "step": 1570
    },
    {
      "epoch": 0.1571,
      "grad_norm": 37.195186614990234,
      "learning_rate": 4.257070707070707e-05,
      "loss": 0.3811,
      "step": 1571
    },
    {
      "epoch": 0.1572,
      "grad_norm": 3.079880952835083,
      "learning_rate": 4.256565656565657e-05,
      "loss": 0.1837,
      "step": 1572
    },
    {
      "epoch": 0.1573,
      "grad_norm": 2.850325107574463,
      "learning_rate": 4.2560606060606065e-05,
      "loss": 0.2443,
      "step": 1573
    },
    {
      "epoch": 0.1574,
      "grad_norm": 4.355357646942139,
      "learning_rate": 4.255555555555556e-05,
      "loss": 0.1609,
      "step": 1574
    },
    {
      "epoch": 0.1575,
      "grad_norm": 11.286717414855957,
      "learning_rate": 4.255050505050505e-05,
      "loss": 0.4733,
      "step": 1575
    },
    {
      "epoch": 0.1576,
      "grad_norm": 9.159163475036621,
      "learning_rate": 4.254545454545455e-05,
      "loss": 0.3705,
      "step": 1576
    },
    {
      "epoch": 0.1577,
      "grad_norm": 3.7005715370178223,
      "learning_rate": 4.254040404040404e-05,
      "loss": 0.1055,
      "step": 1577
    },
    {
      "epoch": 0.1578,
      "grad_norm": 23.223508834838867,
      "learning_rate": 4.253535353535354e-05,
      "loss": 0.6384,
      "step": 1578
    },
    {
      "epoch": 0.1579,
      "grad_norm": 7.6237287521362305,
      "learning_rate": 4.2530303030303034e-05,
      "loss": 0.2285,
      "step": 1579
    },
    {
      "epoch": 0.158,
      "grad_norm": 18.105928421020508,
      "learning_rate": 4.252525252525253e-05,
      "loss": 0.0741,
      "step": 1580
    },
    {
      "epoch": 0.1581,
      "grad_norm": 6.118851184844971,
      "learning_rate": 4.252020202020202e-05,
      "loss": 0.3034,
      "step": 1581
    },
    {
      "epoch": 0.1582,
      "grad_norm": 16.454137802124023,
      "learning_rate": 4.251515151515152e-05,
      "loss": 0.232,
      "step": 1582
    },
    {
      "epoch": 0.1583,
      "grad_norm": 25.458532333374023,
      "learning_rate": 4.251010101010101e-05,
      "loss": 0.0619,
      "step": 1583
    },
    {
      "epoch": 0.1584,
      "grad_norm": 6.507909297943115,
      "learning_rate": 4.250505050505051e-05,
      "loss": 0.4698,
      "step": 1584
    },
    {
      "epoch": 0.1585,
      "grad_norm": 11.719447135925293,
      "learning_rate": 4.25e-05,
      "loss": 0.2426,
      "step": 1585
    },
    {
      "epoch": 0.1586,
      "grad_norm": 8.033214569091797,
      "learning_rate": 4.24949494949495e-05,
      "loss": 0.3261,
      "step": 1586
    },
    {
      "epoch": 0.1587,
      "grad_norm": 5.689024448394775,
      "learning_rate": 4.248989898989899e-05,
      "loss": 0.2572,
      "step": 1587
    },
    {
      "epoch": 0.1588,
      "grad_norm": 5.594699382781982,
      "learning_rate": 4.248484848484849e-05,
      "loss": 0.349,
      "step": 1588
    },
    {
      "epoch": 0.1589,
      "grad_norm": 8.080381393432617,
      "learning_rate": 4.247979797979798e-05,
      "loss": 0.7611,
      "step": 1589
    },
    {
      "epoch": 0.159,
      "grad_norm": 10.03982162475586,
      "learning_rate": 4.2474747474747476e-05,
      "loss": 0.5967,
      "step": 1590
    },
    {
      "epoch": 0.1591,
      "grad_norm": 9.265560150146484,
      "learning_rate": 4.246969696969697e-05,
      "loss": 0.5755,
      "step": 1591
    },
    {
      "epoch": 0.1592,
      "grad_norm": 18.917011260986328,
      "learning_rate": 4.246464646464647e-05,
      "loss": 0.1859,
      "step": 1592
    },
    {
      "epoch": 0.1593,
      "grad_norm": 16.131250381469727,
      "learning_rate": 4.245959595959596e-05,
      "loss": 0.4478,
      "step": 1593
    },
    {
      "epoch": 0.1594,
      "grad_norm": 11.745388984680176,
      "learning_rate": 4.245454545454546e-05,
      "loss": 0.1582,
      "step": 1594
    },
    {
      "epoch": 0.1595,
      "grad_norm": 32.300628662109375,
      "learning_rate": 4.244949494949495e-05,
      "loss": 0.5692,
      "step": 1595
    },
    {
      "epoch": 0.1596,
      "grad_norm": 7.64410924911499,
      "learning_rate": 4.2444444444444445e-05,
      "loss": 0.9972,
      "step": 1596
    },
    {
      "epoch": 0.1597,
      "grad_norm": 6.137459754943848,
      "learning_rate": 4.243939393939394e-05,
      "loss": 0.734,
      "step": 1597
    },
    {
      "epoch": 0.1598,
      "grad_norm": 13.619617462158203,
      "learning_rate": 4.243434343434344e-05,
      "loss": 0.3025,
      "step": 1598
    },
    {
      "epoch": 0.1599,
      "grad_norm": 32.69888687133789,
      "learning_rate": 4.2429292929292927e-05,
      "loss": 0.2483,
      "step": 1599
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.546449661254883,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.3682,
      "step": 1600
    },
    {
      "epoch": 0.1601,
      "grad_norm": 13.956707000732422,
      "learning_rate": 4.241919191919192e-05,
      "loss": 0.5379,
      "step": 1601
    },
    {
      "epoch": 0.1602,
      "grad_norm": 8.188984870910645,
      "learning_rate": 4.2414141414141415e-05,
      "loss": 0.3634,
      "step": 1602
    },
    {
      "epoch": 0.1603,
      "grad_norm": 10.614768028259277,
      "learning_rate": 4.240909090909091e-05,
      "loss": 0.3287,
      "step": 1603
    },
    {
      "epoch": 0.1604,
      "grad_norm": 8.154572486877441,
      "learning_rate": 4.2404040404040407e-05,
      "loss": 0.7782,
      "step": 1604
    },
    {
      "epoch": 0.1605,
      "grad_norm": 6.429945945739746,
      "learning_rate": 4.23989898989899e-05,
      "loss": 0.2744,
      "step": 1605
    },
    {
      "epoch": 0.1606,
      "grad_norm": 7.445661544799805,
      "learning_rate": 4.23939393939394e-05,
      "loss": 1.11,
      "step": 1606
    },
    {
      "epoch": 0.1607,
      "grad_norm": 8.306814193725586,
      "learning_rate": 4.238888888888889e-05,
      "loss": 0.265,
      "step": 1607
    },
    {
      "epoch": 0.1608,
      "grad_norm": 5.470815658569336,
      "learning_rate": 4.2383838383838384e-05,
      "loss": 1.0066,
      "step": 1608
    },
    {
      "epoch": 0.1609,
      "grad_norm": 11.20921516418457,
      "learning_rate": 4.237878787878788e-05,
      "loss": 0.3212,
      "step": 1609
    },
    {
      "epoch": 0.161,
      "grad_norm": 12.51895523071289,
      "learning_rate": 4.2373737373737376e-05,
      "loss": 0.224,
      "step": 1610
    },
    {
      "epoch": 0.1611,
      "grad_norm": 7.640239715576172,
      "learning_rate": 4.236868686868687e-05,
      "loss": 0.2881,
      "step": 1611
    },
    {
      "epoch": 0.1612,
      "grad_norm": 18.26462745666504,
      "learning_rate": 4.236363636363637e-05,
      "loss": 0.5673,
      "step": 1612
    },
    {
      "epoch": 0.1613,
      "grad_norm": 12.443288803100586,
      "learning_rate": 4.235858585858586e-05,
      "loss": 0.7819,
      "step": 1613
    },
    {
      "epoch": 0.1614,
      "grad_norm": 17.0102596282959,
      "learning_rate": 4.235353535353536e-05,
      "loss": 0.2034,
      "step": 1614
    },
    {
      "epoch": 0.1615,
      "grad_norm": 5.508864402770996,
      "learning_rate": 4.234848484848485e-05,
      "loss": 0.0938,
      "step": 1615
    },
    {
      "epoch": 0.1616,
      "grad_norm": 4.716686248779297,
      "learning_rate": 4.2343434343434345e-05,
      "loss": 0.1805,
      "step": 1616
    },
    {
      "epoch": 0.1617,
      "grad_norm": 11.7072172164917,
      "learning_rate": 4.233838383838384e-05,
      "loss": 0.4299,
      "step": 1617
    },
    {
      "epoch": 0.1618,
      "grad_norm": 6.941978454589844,
      "learning_rate": 4.233333333333334e-05,
      "loss": 0.1285,
      "step": 1618
    },
    {
      "epoch": 0.1619,
      "grad_norm": 5.537332057952881,
      "learning_rate": 4.2328282828282826e-05,
      "loss": 0.1157,
      "step": 1619
    },
    {
      "epoch": 0.162,
      "grad_norm": 3.896941900253296,
      "learning_rate": 4.232323232323233e-05,
      "loss": 0.5157,
      "step": 1620
    },
    {
      "epoch": 0.1621,
      "grad_norm": 13.695942878723145,
      "learning_rate": 4.231818181818182e-05,
      "loss": 0.6969,
      "step": 1621
    },
    {
      "epoch": 0.1622,
      "grad_norm": 6.521020412445068,
      "learning_rate": 4.2313131313131314e-05,
      "loss": 0.4262,
      "step": 1622
    },
    {
      "epoch": 0.1623,
      "grad_norm": 7.907081604003906,
      "learning_rate": 4.230808080808081e-05,
      "loss": 0.3952,
      "step": 1623
    },
    {
      "epoch": 0.1624,
      "grad_norm": 16.84214210510254,
      "learning_rate": 4.2303030303030306e-05,
      "loss": 0.2046,
      "step": 1624
    },
    {
      "epoch": 0.1625,
      "grad_norm": 8.589207649230957,
      "learning_rate": 4.2297979797979795e-05,
      "loss": 0.1901,
      "step": 1625
    },
    {
      "epoch": 0.1626,
      "grad_norm": 5.49573278427124,
      "learning_rate": 4.22929292929293e-05,
      "loss": 0.2971,
      "step": 1626
    },
    {
      "epoch": 0.1627,
      "grad_norm": 10.363228797912598,
      "learning_rate": 4.228787878787879e-05,
      "loss": 0.0983,
      "step": 1627
    },
    {
      "epoch": 0.1628,
      "grad_norm": 6.6771559715271,
      "learning_rate": 4.228282828282828e-05,
      "loss": 0.146,
      "step": 1628
    },
    {
      "epoch": 0.1629,
      "grad_norm": 15.553668022155762,
      "learning_rate": 4.227777777777778e-05,
      "loss": 0.1937,
      "step": 1629
    },
    {
      "epoch": 0.163,
      "grad_norm": 5.771736145019531,
      "learning_rate": 4.2272727272727275e-05,
      "loss": 0.2263,
      "step": 1630
    },
    {
      "epoch": 0.1631,
      "grad_norm": 7.139773845672607,
      "learning_rate": 4.2267676767676764e-05,
      "loss": 0.3427,
      "step": 1631
    },
    {
      "epoch": 0.1632,
      "grad_norm": 11.685315132141113,
      "learning_rate": 4.226262626262627e-05,
      "loss": 0.395,
      "step": 1632
    },
    {
      "epoch": 0.1633,
      "grad_norm": 7.999608039855957,
      "learning_rate": 4.2257575757575756e-05,
      "loss": 1.9622,
      "step": 1633
    },
    {
      "epoch": 0.1634,
      "grad_norm": 6.761656284332275,
      "learning_rate": 4.225252525252525e-05,
      "loss": 0.2321,
      "step": 1634
    },
    {
      "epoch": 0.1635,
      "grad_norm": 4.602180480957031,
      "learning_rate": 4.224747474747475e-05,
      "loss": 0.1077,
      "step": 1635
    },
    {
      "epoch": 0.1636,
      "grad_norm": 8.275554656982422,
      "learning_rate": 4.2242424242424244e-05,
      "loss": 0.2929,
      "step": 1636
    },
    {
      "epoch": 0.1637,
      "grad_norm": 3.7078351974487305,
      "learning_rate": 4.223737373737374e-05,
      "loss": 0.3042,
      "step": 1637
    },
    {
      "epoch": 0.1638,
      "grad_norm": 5.386572360992432,
      "learning_rate": 4.2232323232323236e-05,
      "loss": 0.1246,
      "step": 1638
    },
    {
      "epoch": 0.1639,
      "grad_norm": 16.139705657958984,
      "learning_rate": 4.222727272727273e-05,
      "loss": 0.4228,
      "step": 1639
    },
    {
      "epoch": 0.164,
      "grad_norm": 5.443061828613281,
      "learning_rate": 4.222222222222222e-05,
      "loss": 0.8996,
      "step": 1640
    },
    {
      "epoch": 0.1641,
      "grad_norm": 12.850011825561523,
      "learning_rate": 4.2217171717171724e-05,
      "loss": 1.0305,
      "step": 1641
    },
    {
      "epoch": 0.1642,
      "grad_norm": 7.197300434112549,
      "learning_rate": 4.2212121212121214e-05,
      "loss": 0.2768,
      "step": 1642
    },
    {
      "epoch": 0.1643,
      "grad_norm": 5.38131046295166,
      "learning_rate": 4.220707070707071e-05,
      "loss": 0.1444,
      "step": 1643
    },
    {
      "epoch": 0.1644,
      "grad_norm": 11.471158027648926,
      "learning_rate": 4.2202020202020205e-05,
      "loss": 0.5416,
      "step": 1644
    },
    {
      "epoch": 0.1645,
      "grad_norm": 21.15543556213379,
      "learning_rate": 4.21969696969697e-05,
      "loss": 1.1858,
      "step": 1645
    },
    {
      "epoch": 0.1646,
      "grad_norm": 8.268730163574219,
      "learning_rate": 4.219191919191919e-05,
      "loss": 0.374,
      "step": 1646
    },
    {
      "epoch": 0.1647,
      "grad_norm": 8.389582633972168,
      "learning_rate": 4.2186868686868693e-05,
      "loss": 0.5293,
      "step": 1647
    },
    {
      "epoch": 0.1648,
      "grad_norm": 8.273667335510254,
      "learning_rate": 4.218181818181818e-05,
      "loss": 0.6938,
      "step": 1648
    },
    {
      "epoch": 0.1649,
      "grad_norm": 5.098051071166992,
      "learning_rate": 4.217676767676768e-05,
      "loss": 0.3358,
      "step": 1649
    },
    {
      "epoch": 0.165,
      "grad_norm": 11.158885955810547,
      "learning_rate": 4.2171717171717175e-05,
      "loss": 0.2376,
      "step": 1650
    },
    {
      "epoch": 0.1651,
      "grad_norm": 23.413677215576172,
      "learning_rate": 4.216666666666667e-05,
      "loss": 0.3709,
      "step": 1651
    },
    {
      "epoch": 0.1652,
      "grad_norm": 6.67278528213501,
      "learning_rate": 4.216161616161616e-05,
      "loss": 0.1687,
      "step": 1652
    },
    {
      "epoch": 0.1653,
      "grad_norm": 5.1817498207092285,
      "learning_rate": 4.215656565656566e-05,
      "loss": 0.2239,
      "step": 1653
    },
    {
      "epoch": 0.1654,
      "grad_norm": 6.34110689163208,
      "learning_rate": 4.215151515151515e-05,
      "loss": 0.577,
      "step": 1654
    },
    {
      "epoch": 0.1655,
      "grad_norm": 2.808788776397705,
      "learning_rate": 4.214646464646465e-05,
      "loss": 1.031,
      "step": 1655
    },
    {
      "epoch": 0.1656,
      "grad_norm": 2.06892466545105,
      "learning_rate": 4.2141414141414144e-05,
      "loss": 0.0517,
      "step": 1656
    },
    {
      "epoch": 0.1657,
      "grad_norm": 2.446943521499634,
      "learning_rate": 4.213636363636364e-05,
      "loss": 0.5079,
      "step": 1657
    },
    {
      "epoch": 0.1658,
      "grad_norm": 7.860481262207031,
      "learning_rate": 4.2131313131313136e-05,
      "loss": 0.3588,
      "step": 1658
    },
    {
      "epoch": 0.1659,
      "grad_norm": 2.4861528873443604,
      "learning_rate": 4.212626262626263e-05,
      "loss": 0.0525,
      "step": 1659
    },
    {
      "epoch": 0.166,
      "grad_norm": 6.412899017333984,
      "learning_rate": 4.212121212121212e-05,
      "loss": 0.4689,
      "step": 1660
    },
    {
      "epoch": 0.1661,
      "grad_norm": 3.991183280944824,
      "learning_rate": 4.211616161616162e-05,
      "loss": 0.0296,
      "step": 1661
    },
    {
      "epoch": 0.1662,
      "grad_norm": 9.403154373168945,
      "learning_rate": 4.211111111111111e-05,
      "loss": 0.3389,
      "step": 1662
    },
    {
      "epoch": 0.1663,
      "grad_norm": 11.052852630615234,
      "learning_rate": 4.210606060606061e-05,
      "loss": 0.5362,
      "step": 1663
    },
    {
      "epoch": 0.1664,
      "grad_norm": 11.590909957885742,
      "learning_rate": 4.2101010101010105e-05,
      "loss": 0.4356,
      "step": 1664
    },
    {
      "epoch": 0.1665,
      "grad_norm": 7.038582801818848,
      "learning_rate": 4.20959595959596e-05,
      "loss": 0.2731,
      "step": 1665
    },
    {
      "epoch": 0.1666,
      "grad_norm": 4.837848663330078,
      "learning_rate": 4.209090909090909e-05,
      "loss": 0.0455,
      "step": 1666
    },
    {
      "epoch": 0.1667,
      "grad_norm": 10.672060012817383,
      "learning_rate": 4.208585858585859e-05,
      "loss": 0.2552,
      "step": 1667
    },
    {
      "epoch": 0.1668,
      "grad_norm": 4.7543158531188965,
      "learning_rate": 4.208080808080808e-05,
      "loss": 0.2649,
      "step": 1668
    },
    {
      "epoch": 0.1669,
      "grad_norm": 4.519567489624023,
      "learning_rate": 4.207575757575758e-05,
      "loss": 0.2306,
      "step": 1669
    },
    {
      "epoch": 0.167,
      "grad_norm": 37.14533233642578,
      "learning_rate": 4.2070707070707074e-05,
      "loss": 0.3719,
      "step": 1670
    },
    {
      "epoch": 0.1671,
      "grad_norm": 3.6395277976989746,
      "learning_rate": 4.206565656565657e-05,
      "loss": 0.093,
      "step": 1671
    },
    {
      "epoch": 0.1672,
      "grad_norm": 7.084214210510254,
      "learning_rate": 4.206060606060606e-05,
      "loss": 0.23,
      "step": 1672
    },
    {
      "epoch": 0.1673,
      "grad_norm": 9.593965530395508,
      "learning_rate": 4.205555555555556e-05,
      "loss": 0.6847,
      "step": 1673
    },
    {
      "epoch": 0.1674,
      "grad_norm": 29.439861297607422,
      "learning_rate": 4.205050505050505e-05,
      "loss": 0.4061,
      "step": 1674
    },
    {
      "epoch": 0.1675,
      "grad_norm": 10.886629104614258,
      "learning_rate": 4.204545454545455e-05,
      "loss": 0.8852,
      "step": 1675
    },
    {
      "epoch": 0.1676,
      "grad_norm": 4.932504653930664,
      "learning_rate": 4.204040404040404e-05,
      "loss": 0.6745,
      "step": 1676
    },
    {
      "epoch": 0.1677,
      "grad_norm": 20.507287979125977,
      "learning_rate": 4.203535353535354e-05,
      "loss": 0.1539,
      "step": 1677
    },
    {
      "epoch": 0.1678,
      "grad_norm": 6.524774074554443,
      "learning_rate": 4.203030303030303e-05,
      "loss": 0.089,
      "step": 1678
    },
    {
      "epoch": 0.1679,
      "grad_norm": 5.514453411102295,
      "learning_rate": 4.202525252525253e-05,
      "loss": 0.0661,
      "step": 1679
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.9257525205612183,
      "learning_rate": 4.202020202020202e-05,
      "loss": 0.0226,
      "step": 1680
    },
    {
      "epoch": 0.1681,
      "grad_norm": 7.666175365447998,
      "learning_rate": 4.2015151515151516e-05,
      "loss": 0.0314,
      "step": 1681
    },
    {
      "epoch": 0.1682,
      "grad_norm": 7.748335361480713,
      "learning_rate": 4.201010101010101e-05,
      "loss": 0.1368,
      "step": 1682
    },
    {
      "epoch": 0.1683,
      "grad_norm": 6.280849933624268,
      "learning_rate": 4.200505050505051e-05,
      "loss": 0.1606,
      "step": 1683
    },
    {
      "epoch": 0.1684,
      "grad_norm": 9.120244979858398,
      "learning_rate": 4.2e-05,
      "loss": 0.2297,
      "step": 1684
    },
    {
      "epoch": 0.1685,
      "grad_norm": 41.20259094238281,
      "learning_rate": 4.19949494949495e-05,
      "loss": 0.0898,
      "step": 1685
    },
    {
      "epoch": 0.1686,
      "grad_norm": 4.769396781921387,
      "learning_rate": 4.198989898989899e-05,
      "loss": 0.2456,
      "step": 1686
    },
    {
      "epoch": 0.1687,
      "grad_norm": 11.770502090454102,
      "learning_rate": 4.1984848484848486e-05,
      "loss": 0.4452,
      "step": 1687
    },
    {
      "epoch": 0.1688,
      "grad_norm": 5.461766719818115,
      "learning_rate": 4.197979797979798e-05,
      "loss": 0.799,
      "step": 1688
    },
    {
      "epoch": 0.1689,
      "grad_norm": 1.4221233129501343,
      "learning_rate": 4.197474747474748e-05,
      "loss": 0.0239,
      "step": 1689
    },
    {
      "epoch": 0.169,
      "grad_norm": 4.5103020668029785,
      "learning_rate": 4.196969696969697e-05,
      "loss": 0.1044,
      "step": 1690
    },
    {
      "epoch": 0.1691,
      "grad_norm": 6.616169452667236,
      "learning_rate": 4.196464646464647e-05,
      "loss": 0.3607,
      "step": 1691
    },
    {
      "epoch": 0.1692,
      "grad_norm": 2.652167320251465,
      "learning_rate": 4.195959595959596e-05,
      "loss": 0.4799,
      "step": 1692
    },
    {
      "epoch": 0.1693,
      "grad_norm": 8.230294227600098,
      "learning_rate": 4.1954545454545455e-05,
      "loss": 0.7987,
      "step": 1693
    },
    {
      "epoch": 0.1694,
      "grad_norm": 8.455179214477539,
      "learning_rate": 4.194949494949495e-05,
      "loss": 0.207,
      "step": 1694
    },
    {
      "epoch": 0.1695,
      "grad_norm": 5.97882604598999,
      "learning_rate": 4.194444444444445e-05,
      "loss": 0.2514,
      "step": 1695
    },
    {
      "epoch": 0.1696,
      "grad_norm": 9.282645225524902,
      "learning_rate": 4.193939393939394e-05,
      "loss": 0.0768,
      "step": 1696
    },
    {
      "epoch": 0.1697,
      "grad_norm": 8.571797370910645,
      "learning_rate": 4.193434343434344e-05,
      "loss": 0.2625,
      "step": 1697
    },
    {
      "epoch": 0.1698,
      "grad_norm": 6.393579483032227,
      "learning_rate": 4.192929292929293e-05,
      "loss": 0.1806,
      "step": 1698
    },
    {
      "epoch": 0.1699,
      "grad_norm": 6.502730369567871,
      "learning_rate": 4.1924242424242424e-05,
      "loss": 0.2994,
      "step": 1699
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.53913688659668,
      "learning_rate": 4.191919191919192e-05,
      "loss": 0.3283,
      "step": 1700
    },
    {
      "epoch": 0.1701,
      "grad_norm": 5.016302108764648,
      "learning_rate": 4.1914141414141416e-05,
      "loss": 0.7426,
      "step": 1701
    },
    {
      "epoch": 0.1702,
      "grad_norm": 5.280808448791504,
      "learning_rate": 4.190909090909091e-05,
      "loss": 0.1355,
      "step": 1702
    },
    {
      "epoch": 0.1703,
      "grad_norm": 6.21004056930542,
      "learning_rate": 4.190404040404041e-05,
      "loss": 0.1678,
      "step": 1703
    },
    {
      "epoch": 0.1704,
      "grad_norm": 5.246786117553711,
      "learning_rate": 4.18989898989899e-05,
      "loss": 0.1429,
      "step": 1704
    },
    {
      "epoch": 0.1705,
      "grad_norm": 2.9734857082366943,
      "learning_rate": 4.189393939393939e-05,
      "loss": 0.2371,
      "step": 1705
    },
    {
      "epoch": 0.1706,
      "grad_norm": 4.808830261230469,
      "learning_rate": 4.188888888888889e-05,
      "loss": 0.1544,
      "step": 1706
    },
    {
      "epoch": 0.1707,
      "grad_norm": 7.059626579284668,
      "learning_rate": 4.1883838383838385e-05,
      "loss": 0.2545,
      "step": 1707
    },
    {
      "epoch": 0.1708,
      "grad_norm": 4.329983711242676,
      "learning_rate": 4.187878787878788e-05,
      "loss": 0.1494,
      "step": 1708
    },
    {
      "epoch": 0.1709,
      "grad_norm": 6.699649333953857,
      "learning_rate": 4.187373737373738e-05,
      "loss": 0.0248,
      "step": 1709
    },
    {
      "epoch": 0.171,
      "grad_norm": 5.032986164093018,
      "learning_rate": 4.1868686868686866e-05,
      "loss": 0.3566,
      "step": 1710
    },
    {
      "epoch": 0.1711,
      "grad_norm": 6.073639869689941,
      "learning_rate": 4.186363636363637e-05,
      "loss": 0.2558,
      "step": 1711
    },
    {
      "epoch": 0.1712,
      "grad_norm": 4.392465591430664,
      "learning_rate": 4.185858585858586e-05,
      "loss": 0.3233,
      "step": 1712
    },
    {
      "epoch": 0.1713,
      "grad_norm": 4.342381954193115,
      "learning_rate": 4.1853535353535354e-05,
      "loss": 0.0998,
      "step": 1713
    },
    {
      "epoch": 0.1714,
      "grad_norm": 3.809910774230957,
      "learning_rate": 4.184848484848485e-05,
      "loss": 0.1909,
      "step": 1714
    },
    {
      "epoch": 0.1715,
      "grad_norm": 9.85987377166748,
      "learning_rate": 4.1843434343434346e-05,
      "loss": 0.2843,
      "step": 1715
    },
    {
      "epoch": 0.1716,
      "grad_norm": 5.1785712242126465,
      "learning_rate": 4.1838383838383835e-05,
      "loss": 0.1561,
      "step": 1716
    },
    {
      "epoch": 0.1717,
      "grad_norm": 4.354910373687744,
      "learning_rate": 4.183333333333334e-05,
      "loss": 0.3311,
      "step": 1717
    },
    {
      "epoch": 0.1718,
      "grad_norm": 6.063541889190674,
      "learning_rate": 4.182828282828283e-05,
      "loss": 0.4148,
      "step": 1718
    },
    {
      "epoch": 0.1719,
      "grad_norm": 2.8170158863067627,
      "learning_rate": 4.1823232323232323e-05,
      "loss": 0.5362,
      "step": 1719
    },
    {
      "epoch": 0.172,
      "grad_norm": 6.377957820892334,
      "learning_rate": 4.181818181818182e-05,
      "loss": 0.2905,
      "step": 1720
    },
    {
      "epoch": 0.1721,
      "grad_norm": 10.472000122070312,
      "learning_rate": 4.1813131313131315e-05,
      "loss": 0.9179,
      "step": 1721
    },
    {
      "epoch": 0.1722,
      "grad_norm": 18.96717643737793,
      "learning_rate": 4.180808080808081e-05,
      "loss": 0.629,
      "step": 1722
    },
    {
      "epoch": 0.1723,
      "grad_norm": 3.3534114360809326,
      "learning_rate": 4.180303030303031e-05,
      "loss": 0.1383,
      "step": 1723
    },
    {
      "epoch": 0.1724,
      "grad_norm": 13.097563743591309,
      "learning_rate": 4.1797979797979803e-05,
      "loss": 0.1583,
      "step": 1724
    },
    {
      "epoch": 0.1725,
      "grad_norm": 7.929518699645996,
      "learning_rate": 4.179292929292929e-05,
      "loss": 0.5971,
      "step": 1725
    },
    {
      "epoch": 0.1726,
      "grad_norm": 4.927096366882324,
      "learning_rate": 4.1787878787878795e-05,
      "loss": 0.2294,
      "step": 1726
    },
    {
      "epoch": 0.1727,
      "grad_norm": 1.1531977653503418,
      "learning_rate": 4.1782828282828285e-05,
      "loss": 0.0197,
      "step": 1727
    },
    {
      "epoch": 0.1728,
      "grad_norm": 7.675385475158691,
      "learning_rate": 4.177777777777778e-05,
      "loss": 0.2239,
      "step": 1728
    },
    {
      "epoch": 0.1729,
      "grad_norm": 2.2766029834747314,
      "learning_rate": 4.1772727272727277e-05,
      "loss": 0.0699,
      "step": 1729
    },
    {
      "epoch": 0.173,
      "grad_norm": 3.610635757446289,
      "learning_rate": 4.176767676767677e-05,
      "loss": 0.2659,
      "step": 1730
    },
    {
      "epoch": 0.1731,
      "grad_norm": 6.59266996383667,
      "learning_rate": 4.176262626262626e-05,
      "loss": 0.7511,
      "step": 1731
    },
    {
      "epoch": 0.1732,
      "grad_norm": 7.2987165451049805,
      "learning_rate": 4.1757575757575765e-05,
      "loss": 0.4084,
      "step": 1732
    },
    {
      "epoch": 0.1733,
      "grad_norm": 6.670225143432617,
      "learning_rate": 4.1752525252525254e-05,
      "loss": 0.2536,
      "step": 1733
    },
    {
      "epoch": 0.1734,
      "grad_norm": 11.662955284118652,
      "learning_rate": 4.174747474747475e-05,
      "loss": 0.3367,
      "step": 1734
    },
    {
      "epoch": 0.1735,
      "grad_norm": 8.095751762390137,
      "learning_rate": 4.1742424242424246e-05,
      "loss": 0.2639,
      "step": 1735
    },
    {
      "epoch": 0.1736,
      "grad_norm": 3.40061354637146,
      "learning_rate": 4.173737373737374e-05,
      "loss": 0.7387,
      "step": 1736
    },
    {
      "epoch": 0.1737,
      "grad_norm": 7.575395107269287,
      "learning_rate": 4.173232323232323e-05,
      "loss": 0.2931,
      "step": 1737
    },
    {
      "epoch": 0.1738,
      "grad_norm": 5.528494358062744,
      "learning_rate": 4.1727272727272734e-05,
      "loss": 0.2485,
      "step": 1738
    },
    {
      "epoch": 0.1739,
      "grad_norm": 6.770495891571045,
      "learning_rate": 4.172222222222222e-05,
      "loss": 0.3828,
      "step": 1739
    },
    {
      "epoch": 0.174,
      "grad_norm": 4.44437837600708,
      "learning_rate": 4.171717171717172e-05,
      "loss": 0.384,
      "step": 1740
    },
    {
      "epoch": 0.1741,
      "grad_norm": 4.110107898712158,
      "learning_rate": 4.1712121212121215e-05,
      "loss": 0.1755,
      "step": 1741
    },
    {
      "epoch": 0.1742,
      "grad_norm": 2.655519723892212,
      "learning_rate": 4.170707070707071e-05,
      "loss": 0.0749,
      "step": 1742
    },
    {
      "epoch": 0.1743,
      "grad_norm": 2.0378520488739014,
      "learning_rate": 4.17020202020202e-05,
      "loss": 0.0324,
      "step": 1743
    },
    {
      "epoch": 0.1744,
      "grad_norm": 3.7926454544067383,
      "learning_rate": 4.16969696969697e-05,
      "loss": 0.0519,
      "step": 1744
    },
    {
      "epoch": 0.1745,
      "grad_norm": 4.882641315460205,
      "learning_rate": 4.169191919191919e-05,
      "loss": 0.0588,
      "step": 1745
    },
    {
      "epoch": 0.1746,
      "grad_norm": 7.19382905960083,
      "learning_rate": 4.168686868686869e-05,
      "loss": 0.6753,
      "step": 1746
    },
    {
      "epoch": 0.1747,
      "grad_norm": 7.738876819610596,
      "learning_rate": 4.1681818181818184e-05,
      "loss": 0.1649,
      "step": 1747
    },
    {
      "epoch": 0.1748,
      "grad_norm": 13.436308860778809,
      "learning_rate": 4.167676767676768e-05,
      "loss": 0.546,
      "step": 1748
    },
    {
      "epoch": 0.1749,
      "grad_norm": 14.871387481689453,
      "learning_rate": 4.1671717171717176e-05,
      "loss": 0.1366,
      "step": 1749
    },
    {
      "epoch": 0.175,
      "grad_norm": 3.7852416038513184,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.0974,
      "step": 1750
    },
    {
      "epoch": 0.1751,
      "grad_norm": 93.21372985839844,
      "learning_rate": 4.166161616161616e-05,
      "loss": 0.9168,
      "step": 1751
    },
    {
      "epoch": 0.1752,
      "grad_norm": 7.894371509552002,
      "learning_rate": 4.165656565656566e-05,
      "loss": 0.3108,
      "step": 1752
    },
    {
      "epoch": 0.1753,
      "grad_norm": 16.559513092041016,
      "learning_rate": 4.165151515151515e-05,
      "loss": 0.0545,
      "step": 1753
    },
    {
      "epoch": 0.1754,
      "grad_norm": 4.5353875160217285,
      "learning_rate": 4.164646464646465e-05,
      "loss": 0.1864,
      "step": 1754
    },
    {
      "epoch": 0.1755,
      "grad_norm": 3.7663447856903076,
      "learning_rate": 4.1641414141414145e-05,
      "loss": 0.1126,
      "step": 1755
    },
    {
      "epoch": 0.1756,
      "grad_norm": 58.135948181152344,
      "learning_rate": 4.163636363636364e-05,
      "loss": 0.1405,
      "step": 1756
    },
    {
      "epoch": 0.1757,
      "grad_norm": 10.129554748535156,
      "learning_rate": 4.163131313131313e-05,
      "loss": 1.5296,
      "step": 1757
    },
    {
      "epoch": 0.1758,
      "grad_norm": 55.81517028808594,
      "learning_rate": 4.162626262626263e-05,
      "loss": 0.234,
      "step": 1758
    },
    {
      "epoch": 0.1759,
      "grad_norm": 6.1236252784729,
      "learning_rate": 4.162121212121212e-05,
      "loss": 0.0796,
      "step": 1759
    },
    {
      "epoch": 0.176,
      "grad_norm": 3.625826358795166,
      "learning_rate": 4.161616161616162e-05,
      "loss": 0.1341,
      "step": 1760
    },
    {
      "epoch": 0.1761,
      "grad_norm": 8.232704162597656,
      "learning_rate": 4.1611111111111114e-05,
      "loss": 0.5892,
      "step": 1761
    },
    {
      "epoch": 0.1762,
      "grad_norm": 3.967101812362671,
      "learning_rate": 4.160606060606061e-05,
      "loss": 0.3208,
      "step": 1762
    },
    {
      "epoch": 0.1763,
      "grad_norm": 3.282834053039551,
      "learning_rate": 4.16010101010101e-05,
      "loss": 0.1183,
      "step": 1763
    },
    {
      "epoch": 0.1764,
      "grad_norm": 4.184635162353516,
      "learning_rate": 4.15959595959596e-05,
      "loss": 0.0822,
      "step": 1764
    },
    {
      "epoch": 0.1765,
      "grad_norm": 6.060056209564209,
      "learning_rate": 4.159090909090909e-05,
      "loss": 0.1332,
      "step": 1765
    },
    {
      "epoch": 0.1766,
      "grad_norm": 14.957667350769043,
      "learning_rate": 4.158585858585859e-05,
      "loss": 0.3305,
      "step": 1766
    },
    {
      "epoch": 0.1767,
      "grad_norm": 6.240828037261963,
      "learning_rate": 4.1580808080808084e-05,
      "loss": 0.4551,
      "step": 1767
    },
    {
      "epoch": 0.1768,
      "grad_norm": 4.225847244262695,
      "learning_rate": 4.157575757575758e-05,
      "loss": 0.0815,
      "step": 1768
    },
    {
      "epoch": 0.1769,
      "grad_norm": 6.872257709503174,
      "learning_rate": 4.157070707070707e-05,
      "loss": 0.2715,
      "step": 1769
    },
    {
      "epoch": 0.177,
      "grad_norm": 8.102657318115234,
      "learning_rate": 4.156565656565657e-05,
      "loss": 0.6155,
      "step": 1770
    },
    {
      "epoch": 0.1771,
      "grad_norm": 1.4285147190093994,
      "learning_rate": 4.156060606060606e-05,
      "loss": 0.0238,
      "step": 1771
    },
    {
      "epoch": 0.1772,
      "grad_norm": 12.795692443847656,
      "learning_rate": 4.155555555555556e-05,
      "loss": 0.4694,
      "step": 1772
    },
    {
      "epoch": 0.1773,
      "grad_norm": 16.63220977783203,
      "learning_rate": 4.155050505050505e-05,
      "loss": 0.1726,
      "step": 1773
    },
    {
      "epoch": 0.1774,
      "grad_norm": 3.708301305770874,
      "learning_rate": 4.154545454545455e-05,
      "loss": 0.0381,
      "step": 1774
    },
    {
      "epoch": 0.1775,
      "grad_norm": 4.885568141937256,
      "learning_rate": 4.154040404040404e-05,
      "loss": 1.0944,
      "step": 1775
    },
    {
      "epoch": 0.1776,
      "grad_norm": 2.7780284881591797,
      "learning_rate": 4.153535353535354e-05,
      "loss": 0.0943,
      "step": 1776
    },
    {
      "epoch": 0.1777,
      "grad_norm": 2.0329504013061523,
      "learning_rate": 4.153030303030303e-05,
      "loss": 0.0319,
      "step": 1777
    },
    {
      "epoch": 0.1778,
      "grad_norm": 10.213082313537598,
      "learning_rate": 4.1525252525252526e-05,
      "loss": 1.0207,
      "step": 1778
    },
    {
      "epoch": 0.1779,
      "grad_norm": 8.304875373840332,
      "learning_rate": 4.152020202020202e-05,
      "loss": 0.4232,
      "step": 1779
    },
    {
      "epoch": 0.178,
      "grad_norm": 2.8589560985565186,
      "learning_rate": 4.151515151515152e-05,
      "loss": 0.0433,
      "step": 1780
    },
    {
      "epoch": 0.1781,
      "grad_norm": 7.165769577026367,
      "learning_rate": 4.151010101010101e-05,
      "loss": 0.1409,
      "step": 1781
    },
    {
      "epoch": 0.1782,
      "grad_norm": 6.693011283874512,
      "learning_rate": 4.150505050505051e-05,
      "loss": 0.1825,
      "step": 1782
    },
    {
      "epoch": 0.1783,
      "grad_norm": 10.43548583984375,
      "learning_rate": 4.15e-05,
      "loss": 0.3977,
      "step": 1783
    },
    {
      "epoch": 0.1784,
      "grad_norm": 21.274873733520508,
      "learning_rate": 4.1494949494949495e-05,
      "loss": 0.1002,
      "step": 1784
    },
    {
      "epoch": 0.1785,
      "grad_norm": 11.042582511901855,
      "learning_rate": 4.148989898989899e-05,
      "loss": 0.6598,
      "step": 1785
    },
    {
      "epoch": 0.1786,
      "grad_norm": 3.1672465801239014,
      "learning_rate": 4.148484848484849e-05,
      "loss": 0.0536,
      "step": 1786
    },
    {
      "epoch": 0.1787,
      "grad_norm": 2.1595728397369385,
      "learning_rate": 4.147979797979798e-05,
      "loss": 0.9665,
      "step": 1787
    },
    {
      "epoch": 0.1788,
      "grad_norm": 6.408117771148682,
      "learning_rate": 4.147474747474748e-05,
      "loss": 0.3642,
      "step": 1788
    },
    {
      "epoch": 0.1789,
      "grad_norm": 9.086941719055176,
      "learning_rate": 4.146969696969697e-05,
      "loss": 0.9037,
      "step": 1789
    },
    {
      "epoch": 0.179,
      "grad_norm": 1.2455466985702515,
      "learning_rate": 4.1464646464646464e-05,
      "loss": 0.0109,
      "step": 1790
    },
    {
      "epoch": 0.1791,
      "grad_norm": 4.654355049133301,
      "learning_rate": 4.145959595959596e-05,
      "loss": 0.1326,
      "step": 1791
    },
    {
      "epoch": 0.1792,
      "grad_norm": 2.53128719329834,
      "learning_rate": 4.1454545454545456e-05,
      "loss": 0.0619,
      "step": 1792
    },
    {
      "epoch": 0.1793,
      "grad_norm": 6.211311340332031,
      "learning_rate": 4.144949494949495e-05,
      "loss": 0.2638,
      "step": 1793
    },
    {
      "epoch": 0.1794,
      "grad_norm": 7.729066371917725,
      "learning_rate": 4.144444444444445e-05,
      "loss": 0.1199,
      "step": 1794
    },
    {
      "epoch": 0.1795,
      "grad_norm": 4.379194736480713,
      "learning_rate": 4.143939393939394e-05,
      "loss": 0.677,
      "step": 1795
    },
    {
      "epoch": 0.1796,
      "grad_norm": 6.850740432739258,
      "learning_rate": 4.1434343434343433e-05,
      "loss": 0.6705,
      "step": 1796
    },
    {
      "epoch": 0.1797,
      "grad_norm": 4.593969345092773,
      "learning_rate": 4.142929292929293e-05,
      "loss": 0.1429,
      "step": 1797
    },
    {
      "epoch": 0.1798,
      "grad_norm": 2.739217758178711,
      "learning_rate": 4.1424242424242425e-05,
      "loss": 0.0518,
      "step": 1798
    },
    {
      "epoch": 0.1799,
      "grad_norm": 6.405355930328369,
      "learning_rate": 4.141919191919192e-05,
      "loss": 0.2233,
      "step": 1799
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.267935276031494,
      "learning_rate": 4.141414141414142e-05,
      "loss": 0.2202,
      "step": 1800
    },
    {
      "epoch": 0.1801,
      "grad_norm": 12.594554901123047,
      "learning_rate": 4.1409090909090907e-05,
      "loss": 0.5336,
      "step": 1801
    },
    {
      "epoch": 0.1802,
      "grad_norm": 2.219510555267334,
      "learning_rate": 4.140404040404041e-05,
      "loss": 0.1172,
      "step": 1802
    },
    {
      "epoch": 0.1803,
      "grad_norm": 10.376921653747559,
      "learning_rate": 4.13989898989899e-05,
      "loss": 0.7534,
      "step": 1803
    },
    {
      "epoch": 0.1804,
      "grad_norm": 49.42991256713867,
      "learning_rate": 4.1393939393939395e-05,
      "loss": 0.2717,
      "step": 1804
    },
    {
      "epoch": 0.1805,
      "grad_norm": 4.186535835266113,
      "learning_rate": 4.138888888888889e-05,
      "loss": 0.3455,
      "step": 1805
    },
    {
      "epoch": 0.1806,
      "grad_norm": 3.249621868133545,
      "learning_rate": 4.1383838383838387e-05,
      "loss": 0.1096,
      "step": 1806
    },
    {
      "epoch": 0.1807,
      "grad_norm": 8.392698287963867,
      "learning_rate": 4.137878787878788e-05,
      "loss": 0.4882,
      "step": 1807
    },
    {
      "epoch": 0.1808,
      "grad_norm": 2.1361546516418457,
      "learning_rate": 4.137373737373738e-05,
      "loss": 0.1051,
      "step": 1808
    },
    {
      "epoch": 0.1809,
      "grad_norm": 15.968500137329102,
      "learning_rate": 4.1368686868686874e-05,
      "loss": 0.7769,
      "step": 1809
    },
    {
      "epoch": 0.181,
      "grad_norm": 4.573246002197266,
      "learning_rate": 4.1363636363636364e-05,
      "loss": 0.3807,
      "step": 1810
    },
    {
      "epoch": 0.1811,
      "grad_norm": 5.895662784576416,
      "learning_rate": 4.1358585858585866e-05,
      "loss": 0.4697,
      "step": 1811
    },
    {
      "epoch": 0.1812,
      "grad_norm": 8.654304504394531,
      "learning_rate": 4.1353535353535356e-05,
      "loss": 0.168,
      "step": 1812
    },
    {
      "epoch": 0.1813,
      "grad_norm": 6.495271682739258,
      "learning_rate": 4.134848484848485e-05,
      "loss": 0.2686,
      "step": 1813
    },
    {
      "epoch": 0.1814,
      "grad_norm": 3.004690408706665,
      "learning_rate": 4.134343434343435e-05,
      "loss": 0.4928,
      "step": 1814
    },
    {
      "epoch": 0.1815,
      "grad_norm": 5.507744789123535,
      "learning_rate": 4.1338383838383844e-05,
      "loss": 0.4381,
      "step": 1815
    },
    {
      "epoch": 0.1816,
      "grad_norm": 7.575071811676025,
      "learning_rate": 4.133333333333333e-05,
      "loss": 0.7354,
      "step": 1816
    },
    {
      "epoch": 0.1817,
      "grad_norm": 4.5434699058532715,
      "learning_rate": 4.1328282828282836e-05,
      "loss": 0.6036,
      "step": 1817
    },
    {
      "epoch": 0.1818,
      "grad_norm": 9.841665267944336,
      "learning_rate": 4.1323232323232325e-05,
      "loss": 0.7461,
      "step": 1818
    },
    {
      "epoch": 0.1819,
      "grad_norm": 6.511895656585693,
      "learning_rate": 4.131818181818182e-05,
      "loss": 0.7122,
      "step": 1819
    },
    {
      "epoch": 0.182,
      "grad_norm": 9.843351364135742,
      "learning_rate": 4.131313131313132e-05,
      "loss": 0.4322,
      "step": 1820
    },
    {
      "epoch": 0.1821,
      "grad_norm": 7.342960834503174,
      "learning_rate": 4.130808080808081e-05,
      "loss": 0.3401,
      "step": 1821
    },
    {
      "epoch": 0.1822,
      "grad_norm": 4.160350799560547,
      "learning_rate": 4.13030303030303e-05,
      "loss": 0.1077,
      "step": 1822
    },
    {
      "epoch": 0.1823,
      "grad_norm": 56.98871994018555,
      "learning_rate": 4.1297979797979805e-05,
      "loss": 0.3708,
      "step": 1823
    },
    {
      "epoch": 0.1824,
      "grad_norm": 20.562061309814453,
      "learning_rate": 4.1292929292929294e-05,
      "loss": 0.1153,
      "step": 1824
    },
    {
      "epoch": 0.1825,
      "grad_norm": 3.6951284408569336,
      "learning_rate": 4.128787878787879e-05,
      "loss": 0.0958,
      "step": 1825
    },
    {
      "epoch": 0.1826,
      "grad_norm": 4.904908180236816,
      "learning_rate": 4.1282828282828286e-05,
      "loss": 0.2165,
      "step": 1826
    },
    {
      "epoch": 0.1827,
      "grad_norm": 3.0236170291900635,
      "learning_rate": 4.127777777777778e-05,
      "loss": 0.558,
      "step": 1827
    },
    {
      "epoch": 0.1828,
      "grad_norm": 10.99720287322998,
      "learning_rate": 4.127272727272727e-05,
      "loss": 0.7746,
      "step": 1828
    },
    {
      "epoch": 0.1829,
      "grad_norm": 3.722938060760498,
      "learning_rate": 4.1267676767676774e-05,
      "loss": 0.1124,
      "step": 1829
    },
    {
      "epoch": 0.183,
      "grad_norm": 2.111355781555176,
      "learning_rate": 4.126262626262626e-05,
      "loss": 0.0643,
      "step": 1830
    },
    {
      "epoch": 0.1831,
      "grad_norm": 8.77035903930664,
      "learning_rate": 4.125757575757576e-05,
      "loss": 0.2094,
      "step": 1831
    },
    {
      "epoch": 0.1832,
      "grad_norm": 13.943181991577148,
      "learning_rate": 4.1252525252525255e-05,
      "loss": 0.4252,
      "step": 1832
    },
    {
      "epoch": 0.1833,
      "grad_norm": 5.615204334259033,
      "learning_rate": 4.124747474747475e-05,
      "loss": 0.5538,
      "step": 1833
    },
    {
      "epoch": 0.1834,
      "grad_norm": 1.135623574256897,
      "learning_rate": 4.124242424242424e-05,
      "loss": 0.0313,
      "step": 1834
    },
    {
      "epoch": 0.1835,
      "grad_norm": 1.4566131830215454,
      "learning_rate": 4.123737373737374e-05,
      "loss": 0.4553,
      "step": 1835
    },
    {
      "epoch": 0.1836,
      "grad_norm": 4.315639495849609,
      "learning_rate": 4.123232323232323e-05,
      "loss": 0.2763,
      "step": 1836
    },
    {
      "epoch": 0.1837,
      "grad_norm": 5.090106964111328,
      "learning_rate": 4.122727272727273e-05,
      "loss": 0.117,
      "step": 1837
    },
    {
      "epoch": 0.1838,
      "grad_norm": 20.52265739440918,
      "learning_rate": 4.1222222222222224e-05,
      "loss": 0.2508,
      "step": 1838
    },
    {
      "epoch": 0.1839,
      "grad_norm": 3.01655912399292,
      "learning_rate": 4.121717171717172e-05,
      "loss": 0.1179,
      "step": 1839
    },
    {
      "epoch": 0.184,
      "grad_norm": 5.578568458557129,
      "learning_rate": 4.1212121212121216e-05,
      "loss": 0.2649,
      "step": 1840
    },
    {
      "epoch": 0.1841,
      "grad_norm": 3.3576748371124268,
      "learning_rate": 4.120707070707071e-05,
      "loss": 0.3261,
      "step": 1841
    },
    {
      "epoch": 0.1842,
      "grad_norm": 6.222786903381348,
      "learning_rate": 4.12020202020202e-05,
      "loss": 0.4091,
      "step": 1842
    },
    {
      "epoch": 0.1843,
      "grad_norm": 16.29450035095215,
      "learning_rate": 4.11969696969697e-05,
      "loss": 0.3857,
      "step": 1843
    },
    {
      "epoch": 0.1844,
      "grad_norm": 9.329377174377441,
      "learning_rate": 4.1191919191919193e-05,
      "loss": 0.3851,
      "step": 1844
    },
    {
      "epoch": 0.1845,
      "grad_norm": 6.360150337219238,
      "learning_rate": 4.118686868686869e-05,
      "loss": 1.0853,
      "step": 1845
    },
    {
      "epoch": 0.1846,
      "grad_norm": 5.429049968719482,
      "learning_rate": 4.1181818181818185e-05,
      "loss": 0.1193,
      "step": 1846
    },
    {
      "epoch": 0.1847,
      "grad_norm": 21.44135093688965,
      "learning_rate": 4.117676767676768e-05,
      "loss": 0.909,
      "step": 1847
    },
    {
      "epoch": 0.1848,
      "grad_norm": 7.075431823730469,
      "learning_rate": 4.117171717171717e-05,
      "loss": 0.3319,
      "step": 1848
    },
    {
      "epoch": 0.1849,
      "grad_norm": 8.160235404968262,
      "learning_rate": 4.116666666666667e-05,
      "loss": 0.3023,
      "step": 1849
    },
    {
      "epoch": 0.185,
      "grad_norm": 8.279582023620605,
      "learning_rate": 4.116161616161616e-05,
      "loss": 0.6073,
      "step": 1850
    },
    {
      "epoch": 0.1851,
      "grad_norm": 2.4211316108703613,
      "learning_rate": 4.115656565656566e-05,
      "loss": 0.087,
      "step": 1851
    },
    {
      "epoch": 0.1852,
      "grad_norm": 4.383516788482666,
      "learning_rate": 4.1151515151515155e-05,
      "loss": 0.3186,
      "step": 1852
    },
    {
      "epoch": 0.1853,
      "grad_norm": 7.412717819213867,
      "learning_rate": 4.114646464646465e-05,
      "loss": 0.1652,
      "step": 1853
    },
    {
      "epoch": 0.1854,
      "grad_norm": 5.364938259124756,
      "learning_rate": 4.114141414141414e-05,
      "loss": 0.1789,
      "step": 1854
    },
    {
      "epoch": 0.1855,
      "grad_norm": 3.9459195137023926,
      "learning_rate": 4.113636363636364e-05,
      "loss": 0.2522,
      "step": 1855
    },
    {
      "epoch": 0.1856,
      "grad_norm": 12.8801851272583,
      "learning_rate": 4.113131313131313e-05,
      "loss": 0.0674,
      "step": 1856
    },
    {
      "epoch": 0.1857,
      "grad_norm": 3.76902437210083,
      "learning_rate": 4.112626262626263e-05,
      "loss": 0.1445,
      "step": 1857
    },
    {
      "epoch": 0.1858,
      "grad_norm": 26.59897804260254,
      "learning_rate": 4.1121212121212124e-05,
      "loss": 0.534,
      "step": 1858
    },
    {
      "epoch": 0.1859,
      "grad_norm": 3.567960739135742,
      "learning_rate": 4.111616161616162e-05,
      "loss": 1.0684,
      "step": 1859
    },
    {
      "epoch": 0.186,
      "grad_norm": 6.4464545249938965,
      "learning_rate": 4.111111111111111e-05,
      "loss": 0.1541,
      "step": 1860
    },
    {
      "epoch": 0.1861,
      "grad_norm": 3.255634069442749,
      "learning_rate": 4.110606060606061e-05,
      "loss": 0.1739,
      "step": 1861
    },
    {
      "epoch": 0.1862,
      "grad_norm": 10.793841361999512,
      "learning_rate": 4.11010101010101e-05,
      "loss": 0.7362,
      "step": 1862
    },
    {
      "epoch": 0.1863,
      "grad_norm": 5.358344554901123,
      "learning_rate": 4.10959595959596e-05,
      "loss": 0.3095,
      "step": 1863
    },
    {
      "epoch": 0.1864,
      "grad_norm": 3.432511568069458,
      "learning_rate": 4.109090909090909e-05,
      "loss": 0.303,
      "step": 1864
    },
    {
      "epoch": 0.1865,
      "grad_norm": 6.50601053237915,
      "learning_rate": 4.108585858585859e-05,
      "loss": 0.6408,
      "step": 1865
    },
    {
      "epoch": 0.1866,
      "grad_norm": 6.7414469718933105,
      "learning_rate": 4.108080808080808e-05,
      "loss": 0.2203,
      "step": 1866
    },
    {
      "epoch": 0.1867,
      "grad_norm": 3.8642783164978027,
      "learning_rate": 4.107575757575758e-05,
      "loss": 0.1963,
      "step": 1867
    },
    {
      "epoch": 0.1868,
      "grad_norm": 4.127852439880371,
      "learning_rate": 4.107070707070707e-05,
      "loss": 0.137,
      "step": 1868
    },
    {
      "epoch": 0.1869,
      "grad_norm": 7.227301120758057,
      "learning_rate": 4.1065656565656566e-05,
      "loss": 0.1662,
      "step": 1869
    },
    {
      "epoch": 0.187,
      "grad_norm": 5.147050857543945,
      "learning_rate": 4.106060606060606e-05,
      "loss": 0.17,
      "step": 1870
    },
    {
      "epoch": 0.1871,
      "grad_norm": 4.87009334564209,
      "learning_rate": 4.105555555555556e-05,
      "loss": 0.4795,
      "step": 1871
    },
    {
      "epoch": 0.1872,
      "grad_norm": 5.3162126541137695,
      "learning_rate": 4.105050505050505e-05,
      "loss": 0.7452,
      "step": 1872
    },
    {
      "epoch": 0.1873,
      "grad_norm": 7.848865509033203,
      "learning_rate": 4.104545454545455e-05,
      "loss": 0.3476,
      "step": 1873
    },
    {
      "epoch": 0.1874,
      "grad_norm": 6.02916955947876,
      "learning_rate": 4.104040404040404e-05,
      "loss": 0.143,
      "step": 1874
    },
    {
      "epoch": 0.1875,
      "grad_norm": 5.8158040046691895,
      "learning_rate": 4.1035353535353535e-05,
      "loss": 0.189,
      "step": 1875
    },
    {
      "epoch": 0.1876,
      "grad_norm": 7.455306053161621,
      "learning_rate": 4.103030303030303e-05,
      "loss": 0.2038,
      "step": 1876
    },
    {
      "epoch": 0.1877,
      "grad_norm": 5.360888957977295,
      "learning_rate": 4.102525252525253e-05,
      "loss": 0.1571,
      "step": 1877
    },
    {
      "epoch": 0.1878,
      "grad_norm": 4.837212085723877,
      "learning_rate": 4.1020202020202017e-05,
      "loss": 0.3333,
      "step": 1878
    },
    {
      "epoch": 0.1879,
      "grad_norm": 27.492605209350586,
      "learning_rate": 4.101515151515152e-05,
      "loss": 0.1034,
      "step": 1879
    },
    {
      "epoch": 0.188,
      "grad_norm": 5.451103687286377,
      "learning_rate": 4.101010101010101e-05,
      "loss": 0.2781,
      "step": 1880
    },
    {
      "epoch": 0.1881,
      "grad_norm": 8.305437088012695,
      "learning_rate": 4.1005050505050504e-05,
      "loss": 0.8407,
      "step": 1881
    },
    {
      "epoch": 0.1882,
      "grad_norm": 5.863147258758545,
      "learning_rate": 4.1e-05,
      "loss": 1.0952,
      "step": 1882
    },
    {
      "epoch": 0.1883,
      "grad_norm": 4.861309051513672,
      "learning_rate": 4.0994949494949496e-05,
      "loss": 0.1798,
      "step": 1883
    },
    {
      "epoch": 0.1884,
      "grad_norm": 4.3185224533081055,
      "learning_rate": 4.098989898989899e-05,
      "loss": 1.089,
      "step": 1884
    },
    {
      "epoch": 0.1885,
      "grad_norm": 7.1203789710998535,
      "learning_rate": 4.098484848484849e-05,
      "loss": 0.3442,
      "step": 1885
    },
    {
      "epoch": 0.1886,
      "grad_norm": 20.753040313720703,
      "learning_rate": 4.097979797979798e-05,
      "loss": 0.2188,
      "step": 1886
    },
    {
      "epoch": 0.1887,
      "grad_norm": 5.2101922035217285,
      "learning_rate": 4.0974747474747474e-05,
      "loss": 0.548,
      "step": 1887
    },
    {
      "epoch": 0.1888,
      "grad_norm": 4.128884792327881,
      "learning_rate": 4.096969696969697e-05,
      "loss": 0.1879,
      "step": 1888
    },
    {
      "epoch": 0.1889,
      "grad_norm": 4.7189226150512695,
      "learning_rate": 4.0964646464646466e-05,
      "loss": 0.063,
      "step": 1889
    },
    {
      "epoch": 0.189,
      "grad_norm": 6.4641337394714355,
      "learning_rate": 4.095959595959596e-05,
      "loss": 0.7229,
      "step": 1890
    },
    {
      "epoch": 0.1891,
      "grad_norm": 6.1957106590271,
      "learning_rate": 4.095454545454546e-05,
      "loss": 0.1946,
      "step": 1891
    },
    {
      "epoch": 0.1892,
      "grad_norm": 1.9290704727172852,
      "learning_rate": 4.094949494949495e-05,
      "loss": 0.0261,
      "step": 1892
    },
    {
      "epoch": 0.1893,
      "grad_norm": 5.407393455505371,
      "learning_rate": 4.094444444444445e-05,
      "loss": 0.7222,
      "step": 1893
    },
    {
      "epoch": 0.1894,
      "grad_norm": 5.952598571777344,
      "learning_rate": 4.0939393939393946e-05,
      "loss": 0.1014,
      "step": 1894
    },
    {
      "epoch": 0.1895,
      "grad_norm": 3.1978213787078857,
      "learning_rate": 4.0934343434343435e-05,
      "loss": 0.3141,
      "step": 1895
    },
    {
      "epoch": 0.1896,
      "grad_norm": 2.5629360675811768,
      "learning_rate": 4.092929292929293e-05,
      "loss": 0.9534,
      "step": 1896
    },
    {
      "epoch": 0.1897,
      "grad_norm": 10.73979377746582,
      "learning_rate": 4.092424242424243e-05,
      "loss": 0.9745,
      "step": 1897
    },
    {
      "epoch": 0.1898,
      "grad_norm": 13.159125328063965,
      "learning_rate": 4.091919191919192e-05,
      "loss": 0.1931,
      "step": 1898
    },
    {
      "epoch": 0.1899,
      "grad_norm": 157.0120086669922,
      "learning_rate": 4.091414141414142e-05,
      "loss": 0.1102,
      "step": 1899
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.427552700042725,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 0.3387,
      "step": 1900
    },
    {
      "epoch": 0.1901,
      "grad_norm": 5.478747844696045,
      "learning_rate": 4.0904040404040404e-05,
      "loss": 0.2667,
      "step": 1901
    },
    {
      "epoch": 0.1902,
      "grad_norm": 3.7736611366271973,
      "learning_rate": 4.089898989898991e-05,
      "loss": 0.3864,
      "step": 1902
    },
    {
      "epoch": 0.1903,
      "grad_norm": 3.9276604652404785,
      "learning_rate": 4.0893939393939396e-05,
      "loss": 0.1249,
      "step": 1903
    },
    {
      "epoch": 0.1904,
      "grad_norm": 6.183051586151123,
      "learning_rate": 4.088888888888889e-05,
      "loss": 0.3251,
      "step": 1904
    },
    {
      "epoch": 0.1905,
      "grad_norm": 3.2486608028411865,
      "learning_rate": 4.088383838383839e-05,
      "loss": 0.0341,
      "step": 1905
    },
    {
      "epoch": 0.1906,
      "grad_norm": 6.189449310302734,
      "learning_rate": 4.0878787878787884e-05,
      "loss": 0.3485,
      "step": 1906
    },
    {
      "epoch": 0.1907,
      "grad_norm": 2.8000054359436035,
      "learning_rate": 4.087373737373737e-05,
      "loss": 0.1059,
      "step": 1907
    },
    {
      "epoch": 0.1908,
      "grad_norm": 11.24833869934082,
      "learning_rate": 4.0868686868686876e-05,
      "loss": 0.192,
      "step": 1908
    },
    {
      "epoch": 0.1909,
      "grad_norm": 5.519534587860107,
      "learning_rate": 4.0863636363636365e-05,
      "loss": 0.5098,
      "step": 1909
    },
    {
      "epoch": 0.191,
      "grad_norm": 8.032538414001465,
      "learning_rate": 4.085858585858586e-05,
      "loss": 0.5893,
      "step": 1910
    },
    {
      "epoch": 0.1911,
      "grad_norm": 9.812773704528809,
      "learning_rate": 4.085353535353536e-05,
      "loss": 0.2182,
      "step": 1911
    },
    {
      "epoch": 0.1912,
      "grad_norm": 8.60567855834961,
      "learning_rate": 4.084848484848485e-05,
      "loss": 0.3698,
      "step": 1912
    },
    {
      "epoch": 0.1913,
      "grad_norm": 2.7441251277923584,
      "learning_rate": 4.084343434343434e-05,
      "loss": 0.0954,
      "step": 1913
    },
    {
      "epoch": 0.1914,
      "grad_norm": 8.356961250305176,
      "learning_rate": 4.0838383838383845e-05,
      "loss": 0.3941,
      "step": 1914
    },
    {
      "epoch": 0.1915,
      "grad_norm": 14.227921485900879,
      "learning_rate": 4.0833333333333334e-05,
      "loss": 0.3012,
      "step": 1915
    },
    {
      "epoch": 0.1916,
      "grad_norm": 16.30252456665039,
      "learning_rate": 4.082828282828283e-05,
      "loss": 1.007,
      "step": 1916
    },
    {
      "epoch": 0.1917,
      "grad_norm": 7.466348171234131,
      "learning_rate": 4.0823232323232326e-05,
      "loss": 0.1815,
      "step": 1917
    },
    {
      "epoch": 0.1918,
      "grad_norm": 11.127074241638184,
      "learning_rate": 4.081818181818182e-05,
      "loss": 0.3531,
      "step": 1918
    },
    {
      "epoch": 0.1919,
      "grad_norm": 17.145444869995117,
      "learning_rate": 4.081313131313131e-05,
      "loss": 0.8989,
      "step": 1919
    },
    {
      "epoch": 0.192,
      "grad_norm": 13.225235939025879,
      "learning_rate": 4.0808080808080814e-05,
      "loss": 0.4844,
      "step": 1920
    },
    {
      "epoch": 0.1921,
      "grad_norm": 3.0128800868988037,
      "learning_rate": 4.0803030303030303e-05,
      "loss": 0.2739,
      "step": 1921
    },
    {
      "epoch": 0.1922,
      "grad_norm": 6.9734930992126465,
      "learning_rate": 4.07979797979798e-05,
      "loss": 0.4094,
      "step": 1922
    },
    {
      "epoch": 0.1923,
      "grad_norm": 17.963031768798828,
      "learning_rate": 4.0792929292929295e-05,
      "loss": 0.4688,
      "step": 1923
    },
    {
      "epoch": 0.1924,
      "grad_norm": 4.78710412979126,
      "learning_rate": 4.078787878787879e-05,
      "loss": 0.0916,
      "step": 1924
    },
    {
      "epoch": 0.1925,
      "grad_norm": 3.738210439682007,
      "learning_rate": 4.078282828282828e-05,
      "loss": 0.5273,
      "step": 1925
    },
    {
      "epoch": 0.1926,
      "grad_norm": 3.062950611114502,
      "learning_rate": 4.0777777777777783e-05,
      "loss": 0.0916,
      "step": 1926
    },
    {
      "epoch": 0.1927,
      "grad_norm": 45.00127029418945,
      "learning_rate": 4.077272727272727e-05,
      "loss": 0.2273,
      "step": 1927
    },
    {
      "epoch": 0.1928,
      "grad_norm": 2.6121277809143066,
      "learning_rate": 4.076767676767677e-05,
      "loss": 0.0701,
      "step": 1928
    },
    {
      "epoch": 0.1929,
      "grad_norm": 30.13821029663086,
      "learning_rate": 4.0762626262626265e-05,
      "loss": 0.5838,
      "step": 1929
    },
    {
      "epoch": 0.193,
      "grad_norm": 4.509305477142334,
      "learning_rate": 4.075757575757576e-05,
      "loss": 0.1052,
      "step": 1930
    },
    {
      "epoch": 0.1931,
      "grad_norm": 3.3127424716949463,
      "learning_rate": 4.0752525252525257e-05,
      "loss": 0.5069,
      "step": 1931
    },
    {
      "epoch": 0.1932,
      "grad_norm": 6.326361656188965,
      "learning_rate": 4.074747474747475e-05,
      "loss": 0.2402,
      "step": 1932
    },
    {
      "epoch": 0.1933,
      "grad_norm": 5.173768043518066,
      "learning_rate": 4.074242424242424e-05,
      "loss": 0.5923,
      "step": 1933
    },
    {
      "epoch": 0.1934,
      "grad_norm": 4.213850021362305,
      "learning_rate": 4.073737373737374e-05,
      "loss": 0.0942,
      "step": 1934
    },
    {
      "epoch": 0.1935,
      "grad_norm": 5.379908561706543,
      "learning_rate": 4.0732323232323234e-05,
      "loss": 0.3485,
      "step": 1935
    },
    {
      "epoch": 0.1936,
      "grad_norm": 4.855535507202148,
      "learning_rate": 4.072727272727273e-05,
      "loss": 0.3188,
      "step": 1936
    },
    {
      "epoch": 0.1937,
      "grad_norm": 7.013472080230713,
      "learning_rate": 4.0722222222222226e-05,
      "loss": 0.1287,
      "step": 1937
    },
    {
      "epoch": 0.1938,
      "grad_norm": 9.075186729431152,
      "learning_rate": 4.071717171717172e-05,
      "loss": 0.3236,
      "step": 1938
    },
    {
      "epoch": 0.1939,
      "grad_norm": 6.92218542098999,
      "learning_rate": 4.071212121212121e-05,
      "loss": 0.133,
      "step": 1939
    },
    {
      "epoch": 0.194,
      "grad_norm": 2.559255599975586,
      "learning_rate": 4.070707070707071e-05,
      "loss": 0.0555,
      "step": 1940
    },
    {
      "epoch": 0.1941,
      "grad_norm": 13.791166305541992,
      "learning_rate": 4.07020202020202e-05,
      "loss": 0.3152,
      "step": 1941
    },
    {
      "epoch": 0.1942,
      "grad_norm": 6.233470439910889,
      "learning_rate": 4.06969696969697e-05,
      "loss": 0.407,
      "step": 1942
    },
    {
      "epoch": 0.1943,
      "grad_norm": 13.00108528137207,
      "learning_rate": 4.0691919191919195e-05,
      "loss": 0.3138,
      "step": 1943
    },
    {
      "epoch": 0.1944,
      "grad_norm": 100.86470794677734,
      "learning_rate": 4.068686868686869e-05,
      "loss": 0.5471,
      "step": 1944
    },
    {
      "epoch": 0.1945,
      "grad_norm": 6.231204986572266,
      "learning_rate": 4.068181818181818e-05,
      "loss": 0.6711,
      "step": 1945
    },
    {
      "epoch": 0.1946,
      "grad_norm": 10.421921730041504,
      "learning_rate": 4.067676767676768e-05,
      "loss": 0.2316,
      "step": 1946
    },
    {
      "epoch": 0.1947,
      "grad_norm": 5.045609951019287,
      "learning_rate": 4.067171717171717e-05,
      "loss": 0.0728,
      "step": 1947
    },
    {
      "epoch": 0.1948,
      "grad_norm": 7.360474586486816,
      "learning_rate": 4.066666666666667e-05,
      "loss": 0.6551,
      "step": 1948
    },
    {
      "epoch": 0.1949,
      "grad_norm": 14.19039249420166,
      "learning_rate": 4.0661616161616164e-05,
      "loss": 0.3357,
      "step": 1949
    },
    {
      "epoch": 0.195,
      "grad_norm": 4.92534875869751,
      "learning_rate": 4.065656565656566e-05,
      "loss": 0.0511,
      "step": 1950
    },
    {
      "epoch": 0.1951,
      "grad_norm": 2.3557443618774414,
      "learning_rate": 4.065151515151515e-05,
      "loss": 0.1054,
      "step": 1951
    },
    {
      "epoch": 0.1952,
      "grad_norm": 36.25886917114258,
      "learning_rate": 4.064646464646465e-05,
      "loss": 0.1625,
      "step": 1952
    },
    {
      "epoch": 0.1953,
      "grad_norm": 10.686098098754883,
      "learning_rate": 4.064141414141414e-05,
      "loss": 0.8881,
      "step": 1953
    },
    {
      "epoch": 0.1954,
      "grad_norm": 4.515255451202393,
      "learning_rate": 4.063636363636364e-05,
      "loss": 0.3641,
      "step": 1954
    },
    {
      "epoch": 0.1955,
      "grad_norm": 5.347385883331299,
      "learning_rate": 4.063131313131313e-05,
      "loss": 0.2242,
      "step": 1955
    },
    {
      "epoch": 0.1956,
      "grad_norm": 7.241715908050537,
      "learning_rate": 4.062626262626263e-05,
      "loss": 0.3802,
      "step": 1956
    },
    {
      "epoch": 0.1957,
      "grad_norm": 6.659506320953369,
      "learning_rate": 4.062121212121212e-05,
      "loss": 0.0777,
      "step": 1957
    },
    {
      "epoch": 0.1958,
      "grad_norm": 3.391310453414917,
      "learning_rate": 4.061616161616162e-05,
      "loss": 0.1493,
      "step": 1958
    },
    {
      "epoch": 0.1959,
      "grad_norm": 15.8690185546875,
      "learning_rate": 4.061111111111111e-05,
      "loss": 0.2299,
      "step": 1959
    },
    {
      "epoch": 0.196,
      "grad_norm": 10.431147575378418,
      "learning_rate": 4.0606060606060606e-05,
      "loss": 0.2119,
      "step": 1960
    },
    {
      "epoch": 0.1961,
      "grad_norm": 7.428546905517578,
      "learning_rate": 4.06010101010101e-05,
      "loss": 0.278,
      "step": 1961
    },
    {
      "epoch": 0.1962,
      "grad_norm": 10.010560035705566,
      "learning_rate": 4.05959595959596e-05,
      "loss": 1.0302,
      "step": 1962
    },
    {
      "epoch": 0.1963,
      "grad_norm": 5.756954669952393,
      "learning_rate": 4.059090909090909e-05,
      "loss": 0.1336,
      "step": 1963
    },
    {
      "epoch": 0.1964,
      "grad_norm": 1.9803146123886108,
      "learning_rate": 4.058585858585859e-05,
      "loss": 0.0557,
      "step": 1964
    },
    {
      "epoch": 0.1965,
      "grad_norm": 15.698123931884766,
      "learning_rate": 4.058080808080808e-05,
      "loss": 0.4943,
      "step": 1965
    },
    {
      "epoch": 0.1966,
      "grad_norm": 4.987886905670166,
      "learning_rate": 4.0575757575757576e-05,
      "loss": 0.2915,
      "step": 1966
    },
    {
      "epoch": 0.1967,
      "grad_norm": 84.80939483642578,
      "learning_rate": 4.057070707070707e-05,
      "loss": 0.338,
      "step": 1967
    },
    {
      "epoch": 0.1968,
      "grad_norm": 5.965792179107666,
      "learning_rate": 4.056565656565657e-05,
      "loss": 0.1708,
      "step": 1968
    },
    {
      "epoch": 0.1969,
      "grad_norm": 3.558662176132202,
      "learning_rate": 4.056060606060606e-05,
      "loss": 0.1248,
      "step": 1969
    },
    {
      "epoch": 0.197,
      "grad_norm": 19.55121612548828,
      "learning_rate": 4.055555555555556e-05,
      "loss": 0.1594,
      "step": 1970
    },
    {
      "epoch": 0.1971,
      "grad_norm": 5.844959259033203,
      "learning_rate": 4.055050505050505e-05,
      "loss": 0.2877,
      "step": 1971
    },
    {
      "epoch": 0.1972,
      "grad_norm": 4.502891540527344,
      "learning_rate": 4.0545454545454545e-05,
      "loss": 0.1127,
      "step": 1972
    },
    {
      "epoch": 0.1973,
      "grad_norm": 4.380280494689941,
      "learning_rate": 4.054040404040404e-05,
      "loss": 0.1305,
      "step": 1973
    },
    {
      "epoch": 0.1974,
      "grad_norm": 4.012421131134033,
      "learning_rate": 4.053535353535354e-05,
      "loss": 0.2946,
      "step": 1974
    },
    {
      "epoch": 0.1975,
      "grad_norm": 5.641999244689941,
      "learning_rate": 4.053030303030303e-05,
      "loss": 0.3173,
      "step": 1975
    },
    {
      "epoch": 0.1976,
      "grad_norm": 8.479796409606934,
      "learning_rate": 4.052525252525253e-05,
      "loss": 0.5169,
      "step": 1976
    },
    {
      "epoch": 0.1977,
      "grad_norm": 4.419803619384766,
      "learning_rate": 4.052020202020202e-05,
      "loss": 0.1471,
      "step": 1977
    },
    {
      "epoch": 0.1978,
      "grad_norm": 10.442277908325195,
      "learning_rate": 4.0515151515151514e-05,
      "loss": 0.2742,
      "step": 1978
    },
    {
      "epoch": 0.1979,
      "grad_norm": 4.652426719665527,
      "learning_rate": 4.051010101010102e-05,
      "loss": 0.1825,
      "step": 1979
    },
    {
      "epoch": 0.198,
      "grad_norm": 3.7306864261627197,
      "learning_rate": 4.0505050505050506e-05,
      "loss": 0.1315,
      "step": 1980
    },
    {
      "epoch": 0.1981,
      "grad_norm": 5.967629432678223,
      "learning_rate": 4.05e-05,
      "loss": 0.0457,
      "step": 1981
    },
    {
      "epoch": 0.1982,
      "grad_norm": 3.7194020748138428,
      "learning_rate": 4.04949494949495e-05,
      "loss": 0.1793,
      "step": 1982
    },
    {
      "epoch": 0.1983,
      "grad_norm": 5.4104695320129395,
      "learning_rate": 4.0489898989898994e-05,
      "loss": 0.3671,
      "step": 1983
    },
    {
      "epoch": 0.1984,
      "grad_norm": 5.053973197937012,
      "learning_rate": 4.048484848484849e-05,
      "loss": 0.1786,
      "step": 1984
    },
    {
      "epoch": 0.1985,
      "grad_norm": 17.989866256713867,
      "learning_rate": 4.0479797979797986e-05,
      "loss": 0.5037,
      "step": 1985
    },
    {
      "epoch": 0.1986,
      "grad_norm": 2.449613094329834,
      "learning_rate": 4.0474747474747475e-05,
      "loss": 0.0769,
      "step": 1986
    },
    {
      "epoch": 0.1987,
      "grad_norm": 39.87034606933594,
      "learning_rate": 4.046969696969697e-05,
      "loss": 0.5174,
      "step": 1987
    },
    {
      "epoch": 0.1988,
      "grad_norm": 9.40482234954834,
      "learning_rate": 4.046464646464647e-05,
      "loss": 0.2593,
      "step": 1988
    },
    {
      "epoch": 0.1989,
      "grad_norm": 4.119239330291748,
      "learning_rate": 4.045959595959596e-05,
      "loss": 0.0875,
      "step": 1989
    },
    {
      "epoch": 0.199,
      "grad_norm": 7.91436243057251,
      "learning_rate": 4.045454545454546e-05,
      "loss": 0.2,
      "step": 1990
    },
    {
      "epoch": 0.1991,
      "grad_norm": 2.8705339431762695,
      "learning_rate": 4.0449494949494955e-05,
      "loss": 0.0487,
      "step": 1991
    },
    {
      "epoch": 0.1992,
      "grad_norm": 10.226252555847168,
      "learning_rate": 4.0444444444444444e-05,
      "loss": 0.2175,
      "step": 1992
    },
    {
      "epoch": 0.1993,
      "grad_norm": 7.312017917633057,
      "learning_rate": 4.043939393939395e-05,
      "loss": 0.2419,
      "step": 1993
    },
    {
      "epoch": 0.1994,
      "grad_norm": 7.372618675231934,
      "learning_rate": 4.0434343434343436e-05,
      "loss": 0.2985,
      "step": 1994
    },
    {
      "epoch": 0.1995,
      "grad_norm": 4.07147741317749,
      "learning_rate": 4.042929292929293e-05,
      "loss": 0.2694,
      "step": 1995
    },
    {
      "epoch": 0.1996,
      "grad_norm": 19.388896942138672,
      "learning_rate": 4.042424242424243e-05,
      "loss": 0.1201,
      "step": 1996
    },
    {
      "epoch": 0.1997,
      "grad_norm": 5.5320048332214355,
      "learning_rate": 4.0419191919191924e-05,
      "loss": 0.2039,
      "step": 1997
    },
    {
      "epoch": 0.1998,
      "grad_norm": 4.32503080368042,
      "learning_rate": 4.041414141414141e-05,
      "loss": 0.1294,
      "step": 1998
    },
    {
      "epoch": 0.1999,
      "grad_norm": 3.662264108657837,
      "learning_rate": 4.0409090909090916e-05,
      "loss": 0.2609,
      "step": 1999
    },
    {
      "epoch": 0.2,
      "grad_norm": 15.821784019470215,
      "learning_rate": 4.0404040404040405e-05,
      "loss": 0.192,
      "step": 2000
    },
    {
      "epoch": 0.2001,
      "grad_norm": 7.063655376434326,
      "learning_rate": 4.03989898989899e-05,
      "loss": 0.1025,
      "step": 2001
    },
    {
      "epoch": 0.2002,
      "grad_norm": 11.87297248840332,
      "learning_rate": 4.03939393939394e-05,
      "loss": 0.6104,
      "step": 2002
    },
    {
      "epoch": 0.2003,
      "grad_norm": 7.861557483673096,
      "learning_rate": 4.038888888888889e-05,
      "loss": 0.5232,
      "step": 2003
    },
    {
      "epoch": 0.2004,
      "grad_norm": 10.644638061523438,
      "learning_rate": 4.038383838383838e-05,
      "loss": 0.3533,
      "step": 2004
    },
    {
      "epoch": 0.2005,
      "grad_norm": 11.556401252746582,
      "learning_rate": 4.0378787878787885e-05,
      "loss": 0.5364,
      "step": 2005
    },
    {
      "epoch": 0.2006,
      "grad_norm": 5.484866619110107,
      "learning_rate": 4.0373737373737375e-05,
      "loss": 0.12,
      "step": 2006
    },
    {
      "epoch": 0.2007,
      "grad_norm": 3.7775399684906006,
      "learning_rate": 4.036868686868687e-05,
      "loss": 0.6366,
      "step": 2007
    },
    {
      "epoch": 0.2008,
      "grad_norm": 6.580745697021484,
      "learning_rate": 4.0363636363636367e-05,
      "loss": 0.6752,
      "step": 2008
    },
    {
      "epoch": 0.2009,
      "grad_norm": 3.059225559234619,
      "learning_rate": 4.035858585858586e-05,
      "loss": 0.0838,
      "step": 2009
    },
    {
      "epoch": 0.201,
      "grad_norm": 7.460925102233887,
      "learning_rate": 4.035353535353535e-05,
      "loss": 0.1204,
      "step": 2010
    },
    {
      "epoch": 0.2011,
      "grad_norm": 5.172837257385254,
      "learning_rate": 4.0348484848484854e-05,
      "loss": 0.2947,
      "step": 2011
    },
    {
      "epoch": 0.2012,
      "grad_norm": 6.3361005783081055,
      "learning_rate": 4.0343434343434344e-05,
      "loss": 0.3131,
      "step": 2012
    },
    {
      "epoch": 0.2013,
      "grad_norm": 1.972973108291626,
      "learning_rate": 4.033838383838384e-05,
      "loss": 0.0487,
      "step": 2013
    },
    {
      "epoch": 0.2014,
      "grad_norm": 8.367852210998535,
      "learning_rate": 4.0333333333333336e-05,
      "loss": 0.2177,
      "step": 2014
    },
    {
      "epoch": 0.2015,
      "grad_norm": 6.992973804473877,
      "learning_rate": 4.032828282828283e-05,
      "loss": 0.2415,
      "step": 2015
    },
    {
      "epoch": 0.2016,
      "grad_norm": 3.2700700759887695,
      "learning_rate": 4.032323232323232e-05,
      "loss": 0.1147,
      "step": 2016
    },
    {
      "epoch": 0.2017,
      "grad_norm": 4.653086185455322,
      "learning_rate": 4.0318181818181824e-05,
      "loss": 0.1621,
      "step": 2017
    },
    {
      "epoch": 0.2018,
      "grad_norm": 5.9635186195373535,
      "learning_rate": 4.031313131313131e-05,
      "loss": 0.0918,
      "step": 2018
    },
    {
      "epoch": 0.2019,
      "grad_norm": 8.61034870147705,
      "learning_rate": 4.030808080808081e-05,
      "loss": 0.1671,
      "step": 2019
    },
    {
      "epoch": 0.202,
      "grad_norm": 3.5889697074890137,
      "learning_rate": 4.0303030303030305e-05,
      "loss": 0.141,
      "step": 2020
    },
    {
      "epoch": 0.2021,
      "grad_norm": 9.18524169921875,
      "learning_rate": 4.02979797979798e-05,
      "loss": 0.1785,
      "step": 2021
    },
    {
      "epoch": 0.2022,
      "grad_norm": 6.2082695960998535,
      "learning_rate": 4.029292929292929e-05,
      "loss": 0.1376,
      "step": 2022
    },
    {
      "epoch": 0.2023,
      "grad_norm": 16.82495880126953,
      "learning_rate": 4.028787878787879e-05,
      "loss": 0.1458,
      "step": 2023
    },
    {
      "epoch": 0.2024,
      "grad_norm": 10.045440673828125,
      "learning_rate": 4.028282828282828e-05,
      "loss": 0.1949,
      "step": 2024
    },
    {
      "epoch": 0.2025,
      "grad_norm": 8.564322471618652,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.2913,
      "step": 2025
    },
    {
      "epoch": 0.2026,
      "grad_norm": 1.1002832651138306,
      "learning_rate": 4.0272727272727274e-05,
      "loss": 0.0415,
      "step": 2026
    },
    {
      "epoch": 0.2027,
      "grad_norm": 4.744013786315918,
      "learning_rate": 4.026767676767677e-05,
      "loss": 0.1355,
      "step": 2027
    },
    {
      "epoch": 0.2028,
      "grad_norm": 5.055851459503174,
      "learning_rate": 4.0262626262626266e-05,
      "loss": 0.1889,
      "step": 2028
    },
    {
      "epoch": 0.2029,
      "grad_norm": 4.1790266036987305,
      "learning_rate": 4.025757575757576e-05,
      "loss": 0.2878,
      "step": 2029
    },
    {
      "epoch": 0.203,
      "grad_norm": 6.577947616577148,
      "learning_rate": 4.025252525252525e-05,
      "loss": 0.5132,
      "step": 2030
    },
    {
      "epoch": 0.2031,
      "grad_norm": 17.318439483642578,
      "learning_rate": 4.024747474747475e-05,
      "loss": 0.1569,
      "step": 2031
    },
    {
      "epoch": 0.2032,
      "grad_norm": 8.277530670166016,
      "learning_rate": 4.024242424242424e-05,
      "loss": 0.3896,
      "step": 2032
    },
    {
      "epoch": 0.2033,
      "grad_norm": 5.609744548797607,
      "learning_rate": 4.023737373737374e-05,
      "loss": 0.3444,
      "step": 2033
    },
    {
      "epoch": 0.2034,
      "grad_norm": 36.81923294067383,
      "learning_rate": 4.0232323232323235e-05,
      "loss": 0.1075,
      "step": 2034
    },
    {
      "epoch": 0.2035,
      "grad_norm": 3.0448806285858154,
      "learning_rate": 4.022727272727273e-05,
      "loss": 0.106,
      "step": 2035
    },
    {
      "epoch": 0.2036,
      "grad_norm": 8.488776206970215,
      "learning_rate": 4.022222222222222e-05,
      "loss": 0.3114,
      "step": 2036
    },
    {
      "epoch": 0.2037,
      "grad_norm": 7.7667155265808105,
      "learning_rate": 4.021717171717172e-05,
      "loss": 0.2429,
      "step": 2037
    },
    {
      "epoch": 0.2038,
      "grad_norm": 6.912649154663086,
      "learning_rate": 4.021212121212121e-05,
      "loss": 0.2507,
      "step": 2038
    },
    {
      "epoch": 0.2039,
      "grad_norm": 11.745746612548828,
      "learning_rate": 4.020707070707071e-05,
      "loss": 0.134,
      "step": 2039
    },
    {
      "epoch": 0.204,
      "grad_norm": 6.846892833709717,
      "learning_rate": 4.0202020202020204e-05,
      "loss": 0.4963,
      "step": 2040
    },
    {
      "epoch": 0.2041,
      "grad_norm": 5.593723773956299,
      "learning_rate": 4.01969696969697e-05,
      "loss": 1.1693,
      "step": 2041
    },
    {
      "epoch": 0.2042,
      "grad_norm": 7.404541492462158,
      "learning_rate": 4.019191919191919e-05,
      "loss": 0.0901,
      "step": 2042
    },
    {
      "epoch": 0.2043,
      "grad_norm": 3.7004401683807373,
      "learning_rate": 4.018686868686869e-05,
      "loss": 0.1278,
      "step": 2043
    },
    {
      "epoch": 0.2044,
      "grad_norm": 4.623981475830078,
      "learning_rate": 4.018181818181818e-05,
      "loss": 0.1161,
      "step": 2044
    },
    {
      "epoch": 0.2045,
      "grad_norm": 2.8096542358398438,
      "learning_rate": 4.017676767676768e-05,
      "loss": 0.0802,
      "step": 2045
    },
    {
      "epoch": 0.2046,
      "grad_norm": 2.736145496368408,
      "learning_rate": 4.0171717171717173e-05,
      "loss": 0.0704,
      "step": 2046
    },
    {
      "epoch": 0.2047,
      "grad_norm": 2.3544418811798096,
      "learning_rate": 4.016666666666667e-05,
      "loss": 0.1654,
      "step": 2047
    },
    {
      "epoch": 0.2048,
      "grad_norm": 3.355750799179077,
      "learning_rate": 4.016161616161616e-05,
      "loss": 0.1484,
      "step": 2048
    },
    {
      "epoch": 0.2049,
      "grad_norm": 5.160470962524414,
      "learning_rate": 4.015656565656566e-05,
      "loss": 0.3126,
      "step": 2049
    },
    {
      "epoch": 0.205,
      "grad_norm": 3.0296823978424072,
      "learning_rate": 4.015151515151515e-05,
      "loss": 0.1161,
      "step": 2050
    },
    {
      "epoch": 0.2051,
      "grad_norm": 6.159829616546631,
      "learning_rate": 4.014646464646465e-05,
      "loss": 0.7082,
      "step": 2051
    },
    {
      "epoch": 0.2052,
      "grad_norm": 5.458632946014404,
      "learning_rate": 4.014141414141414e-05,
      "loss": 0.2403,
      "step": 2052
    },
    {
      "epoch": 0.2053,
      "grad_norm": 8.212221145629883,
      "learning_rate": 4.013636363636364e-05,
      "loss": 0.2392,
      "step": 2053
    },
    {
      "epoch": 0.2054,
      "grad_norm": 10.5684814453125,
      "learning_rate": 4.013131313131313e-05,
      "loss": 0.5578,
      "step": 2054
    },
    {
      "epoch": 0.2055,
      "grad_norm": 4.556257724761963,
      "learning_rate": 4.012626262626263e-05,
      "loss": 0.2966,
      "step": 2055
    },
    {
      "epoch": 0.2056,
      "grad_norm": 3.9723918437957764,
      "learning_rate": 4.012121212121212e-05,
      "loss": 0.6429,
      "step": 2056
    },
    {
      "epoch": 0.2057,
      "grad_norm": 3.7802038192749023,
      "learning_rate": 4.0116161616161616e-05,
      "loss": 0.3106,
      "step": 2057
    },
    {
      "epoch": 0.2058,
      "grad_norm": 2.9278202056884766,
      "learning_rate": 4.011111111111111e-05,
      "loss": 0.2702,
      "step": 2058
    },
    {
      "epoch": 0.2059,
      "grad_norm": 7.418622970581055,
      "learning_rate": 4.010606060606061e-05,
      "loss": 0.3141,
      "step": 2059
    },
    {
      "epoch": 0.206,
      "grad_norm": 3.062839984893799,
      "learning_rate": 4.01010101010101e-05,
      "loss": 0.1799,
      "step": 2060
    },
    {
      "epoch": 0.2061,
      "grad_norm": 4.795600414276123,
      "learning_rate": 4.00959595959596e-05,
      "loss": 0.1495,
      "step": 2061
    },
    {
      "epoch": 0.2062,
      "grad_norm": 5.7621073722839355,
      "learning_rate": 4.009090909090909e-05,
      "loss": 0.3194,
      "step": 2062
    },
    {
      "epoch": 0.2063,
      "grad_norm": 3.904067277908325,
      "learning_rate": 4.0085858585858585e-05,
      "loss": 0.1498,
      "step": 2063
    },
    {
      "epoch": 0.2064,
      "grad_norm": 6.181896209716797,
      "learning_rate": 4.008080808080809e-05,
      "loss": 0.0955,
      "step": 2064
    },
    {
      "epoch": 0.2065,
      "grad_norm": 3.1548633575439453,
      "learning_rate": 4.007575757575758e-05,
      "loss": 0.3157,
      "step": 2065
    },
    {
      "epoch": 0.2066,
      "grad_norm": 4.785287857055664,
      "learning_rate": 4.007070707070707e-05,
      "loss": 0.3036,
      "step": 2066
    },
    {
      "epoch": 0.2067,
      "grad_norm": 2.60996675491333,
      "learning_rate": 4.006565656565657e-05,
      "loss": 0.1329,
      "step": 2067
    },
    {
      "epoch": 0.2068,
      "grad_norm": 11.059222221374512,
      "learning_rate": 4.0060606060606065e-05,
      "loss": 0.4233,
      "step": 2068
    },
    {
      "epoch": 0.2069,
      "grad_norm": 4.6998467445373535,
      "learning_rate": 4.0055555555555554e-05,
      "loss": 0.3104,
      "step": 2069
    },
    {
      "epoch": 0.207,
      "grad_norm": 2.0738272666931152,
      "learning_rate": 4.005050505050506e-05,
      "loss": 0.0513,
      "step": 2070
    },
    {
      "epoch": 0.2071,
      "grad_norm": 2.92376708984375,
      "learning_rate": 4.0045454545454546e-05,
      "loss": 0.3191,
      "step": 2071
    },
    {
      "epoch": 0.2072,
      "grad_norm": 4.739400863647461,
      "learning_rate": 4.004040404040404e-05,
      "loss": 0.1542,
      "step": 2072
    },
    {
      "epoch": 0.2073,
      "grad_norm": 16.566913604736328,
      "learning_rate": 4.003535353535354e-05,
      "loss": 0.1839,
      "step": 2073
    },
    {
      "epoch": 0.2074,
      "grad_norm": 3.9530465602874756,
      "learning_rate": 4.0030303030303034e-05,
      "loss": 0.1486,
      "step": 2074
    },
    {
      "epoch": 0.2075,
      "grad_norm": 3.423190116882324,
      "learning_rate": 4.002525252525253e-05,
      "loss": 0.1037,
      "step": 2075
    },
    {
      "epoch": 0.2076,
      "grad_norm": 2.572650671005249,
      "learning_rate": 4.0020202020202026e-05,
      "loss": 0.0615,
      "step": 2076
    },
    {
      "epoch": 0.2077,
      "grad_norm": 5.443295955657959,
      "learning_rate": 4.0015151515151515e-05,
      "loss": 0.2097,
      "step": 2077
    },
    {
      "epoch": 0.2078,
      "grad_norm": 8.088430404663086,
      "learning_rate": 4.001010101010101e-05,
      "loss": 0.0553,
      "step": 2078
    },
    {
      "epoch": 0.2079,
      "grad_norm": 5.034331798553467,
      "learning_rate": 4.000505050505051e-05,
      "loss": 0.3327,
      "step": 2079
    },
    {
      "epoch": 0.208,
      "grad_norm": 4.841721057891846,
      "learning_rate": 4e-05,
      "loss": 0.3186,
      "step": 2080
    },
    {
      "epoch": 0.2081,
      "grad_norm": 4.854321002960205,
      "learning_rate": 3.99949494949495e-05,
      "loss": 0.167,
      "step": 2081
    },
    {
      "epoch": 0.2082,
      "grad_norm": 5.2307329177856445,
      "learning_rate": 3.9989898989898995e-05,
      "loss": 0.2936,
      "step": 2082
    },
    {
      "epoch": 0.2083,
      "grad_norm": 5.919090747833252,
      "learning_rate": 3.9984848484848484e-05,
      "loss": 0.4071,
      "step": 2083
    },
    {
      "epoch": 0.2084,
      "grad_norm": 8.547151565551758,
      "learning_rate": 3.997979797979798e-05,
      "loss": 0.1725,
      "step": 2084
    },
    {
      "epoch": 0.2085,
      "grad_norm": 7.139089584350586,
      "learning_rate": 3.9974747474747476e-05,
      "loss": 0.1979,
      "step": 2085
    },
    {
      "epoch": 0.2086,
      "grad_norm": 4.628290176391602,
      "learning_rate": 3.996969696969697e-05,
      "loss": 0.1197,
      "step": 2086
    },
    {
      "epoch": 0.2087,
      "grad_norm": 12.853306770324707,
      "learning_rate": 3.996464646464647e-05,
      "loss": 0.3206,
      "step": 2087
    },
    {
      "epoch": 0.2088,
      "grad_norm": 3.525947093963623,
      "learning_rate": 3.9959595959595964e-05,
      "loss": 0.1013,
      "step": 2088
    },
    {
      "epoch": 0.2089,
      "grad_norm": 5.58424186706543,
      "learning_rate": 3.9954545454545454e-05,
      "loss": 0.3795,
      "step": 2089
    },
    {
      "epoch": 0.209,
      "grad_norm": 4.734066486358643,
      "learning_rate": 3.9949494949494956e-05,
      "loss": 0.1794,
      "step": 2090
    },
    {
      "epoch": 0.2091,
      "grad_norm": 9.237419128417969,
      "learning_rate": 3.9944444444444446e-05,
      "loss": 0.1098,
      "step": 2091
    },
    {
      "epoch": 0.2092,
      "grad_norm": 3.0038087368011475,
      "learning_rate": 3.993939393939394e-05,
      "loss": 0.3126,
      "step": 2092
    },
    {
      "epoch": 0.2093,
      "grad_norm": 31.774837493896484,
      "learning_rate": 3.993434343434344e-05,
      "loss": 0.6796,
      "step": 2093
    },
    {
      "epoch": 0.2094,
      "grad_norm": 6.375019073486328,
      "learning_rate": 3.9929292929292934e-05,
      "loss": 0.5806,
      "step": 2094
    },
    {
      "epoch": 0.2095,
      "grad_norm": 2.3211770057678223,
      "learning_rate": 3.992424242424242e-05,
      "loss": 0.2261,
      "step": 2095
    },
    {
      "epoch": 0.2096,
      "grad_norm": 3.7230029106140137,
      "learning_rate": 3.9919191919191926e-05,
      "loss": 0.1572,
      "step": 2096
    },
    {
      "epoch": 0.2097,
      "grad_norm": 7.739403247833252,
      "learning_rate": 3.9914141414141415e-05,
      "loss": 0.4871,
      "step": 2097
    },
    {
      "epoch": 0.2098,
      "grad_norm": 2.116325855255127,
      "learning_rate": 3.990909090909091e-05,
      "loss": 0.1141,
      "step": 2098
    },
    {
      "epoch": 0.2099,
      "grad_norm": 4.866354465484619,
      "learning_rate": 3.990404040404041e-05,
      "loss": 0.3087,
      "step": 2099
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.074636459350586,
      "learning_rate": 3.98989898989899e-05,
      "loss": 0.7632,
      "step": 2100
    },
    {
      "epoch": 0.2101,
      "grad_norm": 5.727766036987305,
      "learning_rate": 3.989393939393939e-05,
      "loss": 0.3601,
      "step": 2101
    },
    {
      "epoch": 0.2102,
      "grad_norm": 5.853882312774658,
      "learning_rate": 3.9888888888888895e-05,
      "loss": 0.3713,
      "step": 2102
    },
    {
      "epoch": 0.2103,
      "grad_norm": 2.1612894535064697,
      "learning_rate": 3.9883838383838384e-05,
      "loss": 0.0495,
      "step": 2103
    },
    {
      "epoch": 0.2104,
      "grad_norm": 3.866814613342285,
      "learning_rate": 3.987878787878788e-05,
      "loss": 0.1975,
      "step": 2104
    },
    {
      "epoch": 0.2105,
      "grad_norm": 9.05514144897461,
      "learning_rate": 3.9873737373737376e-05,
      "loss": 0.1809,
      "step": 2105
    },
    {
      "epoch": 0.2106,
      "grad_norm": 4.254115581512451,
      "learning_rate": 3.986868686868687e-05,
      "loss": 0.0897,
      "step": 2106
    },
    {
      "epoch": 0.2107,
      "grad_norm": 4.0532732009887695,
      "learning_rate": 3.986363636363636e-05,
      "loss": 0.1373,
      "step": 2107
    },
    {
      "epoch": 0.2108,
      "grad_norm": 4.247585296630859,
      "learning_rate": 3.9858585858585864e-05,
      "loss": 0.1969,
      "step": 2108
    },
    {
      "epoch": 0.2109,
      "grad_norm": 19.344324111938477,
      "learning_rate": 3.985353535353535e-05,
      "loss": 0.1357,
      "step": 2109
    },
    {
      "epoch": 0.211,
      "grad_norm": 7.336690902709961,
      "learning_rate": 3.984848484848485e-05,
      "loss": 0.3236,
      "step": 2110
    },
    {
      "epoch": 0.2111,
      "grad_norm": 11.162205696105957,
      "learning_rate": 3.9843434343434345e-05,
      "loss": 0.4154,
      "step": 2111
    },
    {
      "epoch": 0.2112,
      "grad_norm": 4.800298690795898,
      "learning_rate": 3.983838383838384e-05,
      "loss": 0.226,
      "step": 2112
    },
    {
      "epoch": 0.2113,
      "grad_norm": 22.199325561523438,
      "learning_rate": 3.983333333333333e-05,
      "loss": 0.3581,
      "step": 2113
    },
    {
      "epoch": 0.2114,
      "grad_norm": 6.777432918548584,
      "learning_rate": 3.982828282828283e-05,
      "loss": 0.357,
      "step": 2114
    },
    {
      "epoch": 0.2115,
      "grad_norm": 4.218321323394775,
      "learning_rate": 3.982323232323232e-05,
      "loss": 0.2251,
      "step": 2115
    },
    {
      "epoch": 0.2116,
      "grad_norm": 8.447227478027344,
      "learning_rate": 3.981818181818182e-05,
      "loss": 0.3145,
      "step": 2116
    },
    {
      "epoch": 0.2117,
      "grad_norm": 6.087860584259033,
      "learning_rate": 3.9813131313131314e-05,
      "loss": 0.4896,
      "step": 2117
    },
    {
      "epoch": 0.2118,
      "grad_norm": 11.07656478881836,
      "learning_rate": 3.980808080808081e-05,
      "loss": 0.5734,
      "step": 2118
    },
    {
      "epoch": 0.2119,
      "grad_norm": 12.637805938720703,
      "learning_rate": 3.9803030303030306e-05,
      "loss": 0.1313,
      "step": 2119
    },
    {
      "epoch": 0.212,
      "grad_norm": 6.053125381469727,
      "learning_rate": 3.97979797979798e-05,
      "loss": 0.1986,
      "step": 2120
    },
    {
      "epoch": 0.2121,
      "grad_norm": 8.608430862426758,
      "learning_rate": 3.979292929292929e-05,
      "loss": 0.047,
      "step": 2121
    },
    {
      "epoch": 0.2122,
      "grad_norm": 2.9238576889038086,
      "learning_rate": 3.978787878787879e-05,
      "loss": 0.1053,
      "step": 2122
    },
    {
      "epoch": 0.2123,
      "grad_norm": 4.40411901473999,
      "learning_rate": 3.9782828282828283e-05,
      "loss": 0.2657,
      "step": 2123
    },
    {
      "epoch": 0.2124,
      "grad_norm": 6.634281635284424,
      "learning_rate": 3.977777777777778e-05,
      "loss": 0.5562,
      "step": 2124
    },
    {
      "epoch": 0.2125,
      "grad_norm": 5.938137531280518,
      "learning_rate": 3.9772727272727275e-05,
      "loss": 0.3592,
      "step": 2125
    },
    {
      "epoch": 0.2126,
      "grad_norm": 5.655174732208252,
      "learning_rate": 3.976767676767677e-05,
      "loss": 0.1612,
      "step": 2126
    },
    {
      "epoch": 0.2127,
      "grad_norm": 5.7029619216918945,
      "learning_rate": 3.976262626262626e-05,
      "loss": 0.2485,
      "step": 2127
    },
    {
      "epoch": 0.2128,
      "grad_norm": 7.071524143218994,
      "learning_rate": 3.975757575757576e-05,
      "loss": 0.6583,
      "step": 2128
    },
    {
      "epoch": 0.2129,
      "grad_norm": 3.4033005237579346,
      "learning_rate": 3.975252525252525e-05,
      "loss": 0.1927,
      "step": 2129
    },
    {
      "epoch": 0.213,
      "grad_norm": 4.839262962341309,
      "learning_rate": 3.974747474747475e-05,
      "loss": 0.2755,
      "step": 2130
    },
    {
      "epoch": 0.2131,
      "grad_norm": 5.8940606117248535,
      "learning_rate": 3.9742424242424245e-05,
      "loss": 0.5808,
      "step": 2131
    },
    {
      "epoch": 0.2132,
      "grad_norm": 5.1146745681762695,
      "learning_rate": 3.973737373737374e-05,
      "loss": 0.091,
      "step": 2132
    },
    {
      "epoch": 0.2133,
      "grad_norm": 7.24613094329834,
      "learning_rate": 3.973232323232323e-05,
      "loss": 0.0324,
      "step": 2133
    },
    {
      "epoch": 0.2134,
      "grad_norm": 15.009244918823242,
      "learning_rate": 3.972727272727273e-05,
      "loss": 0.2365,
      "step": 2134
    },
    {
      "epoch": 0.2135,
      "grad_norm": 6.314990520477295,
      "learning_rate": 3.972222222222222e-05,
      "loss": 0.1994,
      "step": 2135
    },
    {
      "epoch": 0.2136,
      "grad_norm": 12.681876182556152,
      "learning_rate": 3.971717171717172e-05,
      "loss": 0.1854,
      "step": 2136
    },
    {
      "epoch": 0.2137,
      "grad_norm": 5.869358539581299,
      "learning_rate": 3.9712121212121214e-05,
      "loss": 0.2734,
      "step": 2137
    },
    {
      "epoch": 0.2138,
      "grad_norm": 7.1676859855651855,
      "learning_rate": 3.970707070707071e-05,
      "loss": 0.5899,
      "step": 2138
    },
    {
      "epoch": 0.2139,
      "grad_norm": 3.511073589324951,
      "learning_rate": 3.97020202020202e-05,
      "loss": 0.2894,
      "step": 2139
    },
    {
      "epoch": 0.214,
      "grad_norm": 2.749546527862549,
      "learning_rate": 3.96969696969697e-05,
      "loss": 0.1103,
      "step": 2140
    },
    {
      "epoch": 0.2141,
      "grad_norm": 5.971930980682373,
      "learning_rate": 3.969191919191919e-05,
      "loss": 0.4708,
      "step": 2141
    },
    {
      "epoch": 0.2142,
      "grad_norm": 18.032329559326172,
      "learning_rate": 3.968686868686869e-05,
      "loss": 0.2211,
      "step": 2142
    },
    {
      "epoch": 0.2143,
      "grad_norm": 8.091100692749023,
      "learning_rate": 3.968181818181818e-05,
      "loss": 0.337,
      "step": 2143
    },
    {
      "epoch": 0.2144,
      "grad_norm": 5.9881768226623535,
      "learning_rate": 3.967676767676768e-05,
      "loss": 0.2532,
      "step": 2144
    },
    {
      "epoch": 0.2145,
      "grad_norm": 5.095917701721191,
      "learning_rate": 3.967171717171717e-05,
      "loss": 0.2659,
      "step": 2145
    },
    {
      "epoch": 0.2146,
      "grad_norm": 6.685671329498291,
      "learning_rate": 3.966666666666667e-05,
      "loss": 0.1717,
      "step": 2146
    },
    {
      "epoch": 0.2147,
      "grad_norm": 7.95535135269165,
      "learning_rate": 3.966161616161616e-05,
      "loss": 0.2368,
      "step": 2147
    },
    {
      "epoch": 0.2148,
      "grad_norm": 12.222744941711426,
      "learning_rate": 3.9656565656565656e-05,
      "loss": 0.3753,
      "step": 2148
    },
    {
      "epoch": 0.2149,
      "grad_norm": 3.405545473098755,
      "learning_rate": 3.965151515151516e-05,
      "loss": 0.1225,
      "step": 2149
    },
    {
      "epoch": 0.215,
      "grad_norm": 14.214427947998047,
      "learning_rate": 3.964646464646465e-05,
      "loss": 0.8617,
      "step": 2150
    },
    {
      "epoch": 0.2151,
      "grad_norm": 38.91141128540039,
      "learning_rate": 3.9641414141414144e-05,
      "loss": 0.3416,
      "step": 2151
    },
    {
      "epoch": 0.2152,
      "grad_norm": 4.45251989364624,
      "learning_rate": 3.963636363636364e-05,
      "loss": 0.2578,
      "step": 2152
    },
    {
      "epoch": 0.2153,
      "grad_norm": 8.443753242492676,
      "learning_rate": 3.9631313131313136e-05,
      "loss": 0.3392,
      "step": 2153
    },
    {
      "epoch": 0.2154,
      "grad_norm": 24.658077239990234,
      "learning_rate": 3.9626262626262625e-05,
      "loss": 0.394,
      "step": 2154
    },
    {
      "epoch": 0.2155,
      "grad_norm": 6.962604522705078,
      "learning_rate": 3.962121212121213e-05,
      "loss": 0.2415,
      "step": 2155
    },
    {
      "epoch": 0.2156,
      "grad_norm": 1.9039045572280884,
      "learning_rate": 3.961616161616162e-05,
      "loss": 0.1047,
      "step": 2156
    },
    {
      "epoch": 0.2157,
      "grad_norm": 1.886359453201294,
      "learning_rate": 3.961111111111111e-05,
      "loss": 0.0613,
      "step": 2157
    },
    {
      "epoch": 0.2158,
      "grad_norm": 4.1645684242248535,
      "learning_rate": 3.960606060606061e-05,
      "loss": 0.2655,
      "step": 2158
    },
    {
      "epoch": 0.2159,
      "grad_norm": 1.25278902053833,
      "learning_rate": 3.9601010101010105e-05,
      "loss": 0.0296,
      "step": 2159
    },
    {
      "epoch": 0.216,
      "grad_norm": 4.1822919845581055,
      "learning_rate": 3.9595959595959594e-05,
      "loss": 0.1804,
      "step": 2160
    },
    {
      "epoch": 0.2161,
      "grad_norm": 7.79317569732666,
      "learning_rate": 3.95909090909091e-05,
      "loss": 0.1462,
      "step": 2161
    },
    {
      "epoch": 0.2162,
      "grad_norm": 10.834098815917969,
      "learning_rate": 3.9585858585858586e-05,
      "loss": 0.284,
      "step": 2162
    },
    {
      "epoch": 0.2163,
      "grad_norm": 4.533079624176025,
      "learning_rate": 3.958080808080808e-05,
      "loss": 0.112,
      "step": 2163
    },
    {
      "epoch": 0.2164,
      "grad_norm": 2.9182145595550537,
      "learning_rate": 3.957575757575758e-05,
      "loss": 0.1537,
      "step": 2164
    },
    {
      "epoch": 0.2165,
      "grad_norm": 2.662398338317871,
      "learning_rate": 3.9570707070707074e-05,
      "loss": 0.0763,
      "step": 2165
    },
    {
      "epoch": 0.2166,
      "grad_norm": 77.88584899902344,
      "learning_rate": 3.956565656565657e-05,
      "loss": 0.085,
      "step": 2166
    },
    {
      "epoch": 0.2167,
      "grad_norm": 25.524301528930664,
      "learning_rate": 3.9560606060606066e-05,
      "loss": 0.1925,
      "step": 2167
    },
    {
      "epoch": 0.2168,
      "grad_norm": 3.134958267211914,
      "learning_rate": 3.9555555555555556e-05,
      "loss": 0.0691,
      "step": 2168
    },
    {
      "epoch": 0.2169,
      "grad_norm": 8.567543029785156,
      "learning_rate": 3.955050505050505e-05,
      "loss": 0.4458,
      "step": 2169
    },
    {
      "epoch": 0.217,
      "grad_norm": 5.2740702629089355,
      "learning_rate": 3.954545454545455e-05,
      "loss": 0.1986,
      "step": 2170
    },
    {
      "epoch": 0.2171,
      "grad_norm": 4.473512649536133,
      "learning_rate": 3.9540404040404044e-05,
      "loss": 0.1897,
      "step": 2171
    },
    {
      "epoch": 0.2172,
      "grad_norm": 4.5124831199646,
      "learning_rate": 3.953535353535354e-05,
      "loss": 0.0973,
      "step": 2172
    },
    {
      "epoch": 0.2173,
      "grad_norm": 5.43120002746582,
      "learning_rate": 3.9530303030303036e-05,
      "loss": 0.4243,
      "step": 2173
    },
    {
      "epoch": 0.2174,
      "grad_norm": 6.517044544219971,
      "learning_rate": 3.9525252525252525e-05,
      "loss": 0.3586,
      "step": 2174
    },
    {
      "epoch": 0.2175,
      "grad_norm": 5.988959789276123,
      "learning_rate": 3.952020202020202e-05,
      "loss": 0.1858,
      "step": 2175
    },
    {
      "epoch": 0.2176,
      "grad_norm": 16.660629272460938,
      "learning_rate": 3.951515151515152e-05,
      "loss": 0.6181,
      "step": 2176
    },
    {
      "epoch": 0.2177,
      "grad_norm": 5.497982025146484,
      "learning_rate": 3.951010101010101e-05,
      "loss": 0.5384,
      "step": 2177
    },
    {
      "epoch": 0.2178,
      "grad_norm": 1.7682136297225952,
      "learning_rate": 3.950505050505051e-05,
      "loss": 0.045,
      "step": 2178
    },
    {
      "epoch": 0.2179,
      "grad_norm": 2.5776848793029785,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.0788,
      "step": 2179
    },
    {
      "epoch": 0.218,
      "grad_norm": 5.817002296447754,
      "learning_rate": 3.9494949494949494e-05,
      "loss": 0.3433,
      "step": 2180
    },
    {
      "epoch": 0.2181,
      "grad_norm": 5.295860767364502,
      "learning_rate": 3.9489898989899e-05,
      "loss": 0.0876,
      "step": 2181
    },
    {
      "epoch": 0.2182,
      "grad_norm": 5.480366230010986,
      "learning_rate": 3.9484848484848486e-05,
      "loss": 0.8536,
      "step": 2182
    },
    {
      "epoch": 0.2183,
      "grad_norm": 61.91639709472656,
      "learning_rate": 3.947979797979798e-05,
      "loss": 0.5412,
      "step": 2183
    },
    {
      "epoch": 0.2184,
      "grad_norm": 3.715311288833618,
      "learning_rate": 3.947474747474748e-05,
      "loss": 0.2149,
      "step": 2184
    },
    {
      "epoch": 0.2185,
      "grad_norm": 12.916150093078613,
      "learning_rate": 3.9469696969696974e-05,
      "loss": 0.3937,
      "step": 2185
    },
    {
      "epoch": 0.2186,
      "grad_norm": 2.627157688140869,
      "learning_rate": 3.946464646464646e-05,
      "loss": 0.0755,
      "step": 2186
    },
    {
      "epoch": 0.2187,
      "grad_norm": 7.8726983070373535,
      "learning_rate": 3.9459595959595966e-05,
      "loss": 0.5027,
      "step": 2187
    },
    {
      "epoch": 0.2188,
      "grad_norm": 10.285513877868652,
      "learning_rate": 3.9454545454545455e-05,
      "loss": 0.5318,
      "step": 2188
    },
    {
      "epoch": 0.2189,
      "grad_norm": 2.626086473464966,
      "learning_rate": 3.944949494949495e-05,
      "loss": 0.0611,
      "step": 2189
    },
    {
      "epoch": 0.219,
      "grad_norm": 13.008635520935059,
      "learning_rate": 3.944444444444445e-05,
      "loss": 0.2481,
      "step": 2190
    },
    {
      "epoch": 0.2191,
      "grad_norm": 8.275407791137695,
      "learning_rate": 3.943939393939394e-05,
      "loss": 0.3051,
      "step": 2191
    },
    {
      "epoch": 0.2192,
      "grad_norm": 4.908443450927734,
      "learning_rate": 3.943434343434343e-05,
      "loss": 0.2322,
      "step": 2192
    },
    {
      "epoch": 0.2193,
      "grad_norm": 8.2615327835083,
      "learning_rate": 3.9429292929292935e-05,
      "loss": 0.1661,
      "step": 2193
    },
    {
      "epoch": 0.2194,
      "grad_norm": 7.726133823394775,
      "learning_rate": 3.9424242424242424e-05,
      "loss": 0.4552,
      "step": 2194
    },
    {
      "epoch": 0.2195,
      "grad_norm": 6.402275085449219,
      "learning_rate": 3.941919191919192e-05,
      "loss": 0.2007,
      "step": 2195
    },
    {
      "epoch": 0.2196,
      "grad_norm": 3.6310763359069824,
      "learning_rate": 3.9414141414141416e-05,
      "loss": 0.1197,
      "step": 2196
    },
    {
      "epoch": 0.2197,
      "grad_norm": 7.451692581176758,
      "learning_rate": 3.940909090909091e-05,
      "loss": 0.5189,
      "step": 2197
    },
    {
      "epoch": 0.2198,
      "grad_norm": 8.142746925354004,
      "learning_rate": 3.94040404040404e-05,
      "loss": 0.1636,
      "step": 2198
    },
    {
      "epoch": 0.2199,
      "grad_norm": 6.098875522613525,
      "learning_rate": 3.9398989898989904e-05,
      "loss": 0.2096,
      "step": 2199
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.027121067047119,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.6645,
      "step": 2200
    },
    {
      "epoch": 0.2201,
      "grad_norm": 3.005505084991455,
      "learning_rate": 3.938888888888889e-05,
      "loss": 0.1563,
      "step": 2201
    },
    {
      "epoch": 0.2202,
      "grad_norm": 6.054770469665527,
      "learning_rate": 3.9383838383838385e-05,
      "loss": 0.1114,
      "step": 2202
    },
    {
      "epoch": 0.2203,
      "grad_norm": 20.819005966186523,
      "learning_rate": 3.937878787878788e-05,
      "loss": 0.4171,
      "step": 2203
    },
    {
      "epoch": 0.2204,
      "grad_norm": 6.533499240875244,
      "learning_rate": 3.937373737373737e-05,
      "loss": 0.3377,
      "step": 2204
    },
    {
      "epoch": 0.2205,
      "grad_norm": 2.1322929859161377,
      "learning_rate": 3.936868686868687e-05,
      "loss": 0.0689,
      "step": 2205
    },
    {
      "epoch": 0.2206,
      "grad_norm": 2.9361965656280518,
      "learning_rate": 3.936363636363636e-05,
      "loss": 0.0637,
      "step": 2206
    },
    {
      "epoch": 0.2207,
      "grad_norm": 4.279067039489746,
      "learning_rate": 3.935858585858586e-05,
      "loss": 0.1049,
      "step": 2207
    },
    {
      "epoch": 0.2208,
      "grad_norm": 11.759444236755371,
      "learning_rate": 3.9353535353535355e-05,
      "loss": 0.6495,
      "step": 2208
    },
    {
      "epoch": 0.2209,
      "grad_norm": 2.731576681137085,
      "learning_rate": 3.934848484848485e-05,
      "loss": 0.1116,
      "step": 2209
    },
    {
      "epoch": 0.221,
      "grad_norm": 8.184845924377441,
      "learning_rate": 3.9343434343434346e-05,
      "loss": 0.3292,
      "step": 2210
    },
    {
      "epoch": 0.2211,
      "grad_norm": 2.64194655418396,
      "learning_rate": 3.933838383838384e-05,
      "loss": 0.1236,
      "step": 2211
    },
    {
      "epoch": 0.2212,
      "grad_norm": 2.478940963745117,
      "learning_rate": 3.933333333333333e-05,
      "loss": 0.0723,
      "step": 2212
    },
    {
      "epoch": 0.2213,
      "grad_norm": 8.653010368347168,
      "learning_rate": 3.932828282828283e-05,
      "loss": 0.2007,
      "step": 2213
    },
    {
      "epoch": 0.2214,
      "grad_norm": 2.567578077316284,
      "learning_rate": 3.9323232323232324e-05,
      "loss": 0.0636,
      "step": 2214
    },
    {
      "epoch": 0.2215,
      "grad_norm": 31.440185546875,
      "learning_rate": 3.931818181818182e-05,
      "loss": 0.2122,
      "step": 2215
    },
    {
      "epoch": 0.2216,
      "grad_norm": 3.7536299228668213,
      "learning_rate": 3.9313131313131316e-05,
      "loss": 0.0857,
      "step": 2216
    },
    {
      "epoch": 0.2217,
      "grad_norm": 3.021869421005249,
      "learning_rate": 3.930808080808081e-05,
      "loss": 0.0928,
      "step": 2217
    },
    {
      "epoch": 0.2218,
      "grad_norm": 26.685773849487305,
      "learning_rate": 3.93030303030303e-05,
      "loss": 0.2532,
      "step": 2218
    },
    {
      "epoch": 0.2219,
      "grad_norm": 3.79645037651062,
      "learning_rate": 3.9297979797979804e-05,
      "loss": 0.1169,
      "step": 2219
    },
    {
      "epoch": 0.222,
      "grad_norm": 12.777544021606445,
      "learning_rate": 3.929292929292929e-05,
      "loss": 0.1915,
      "step": 2220
    },
    {
      "epoch": 0.2221,
      "grad_norm": 4.820925235748291,
      "learning_rate": 3.928787878787879e-05,
      "loss": 0.1282,
      "step": 2221
    },
    {
      "epoch": 0.2222,
      "grad_norm": 5.0469889640808105,
      "learning_rate": 3.9282828282828285e-05,
      "loss": 0.2828,
      "step": 2222
    },
    {
      "epoch": 0.2223,
      "grad_norm": 6.211517810821533,
      "learning_rate": 3.927777777777778e-05,
      "loss": 0.3846,
      "step": 2223
    },
    {
      "epoch": 0.2224,
      "grad_norm": 7.464432239532471,
      "learning_rate": 3.927272727272727e-05,
      "loss": 1.3219,
      "step": 2224
    },
    {
      "epoch": 0.2225,
      "grad_norm": 6.107182502746582,
      "learning_rate": 3.926767676767677e-05,
      "loss": 0.6309,
      "step": 2225
    },
    {
      "epoch": 0.2226,
      "grad_norm": 1.768639326095581,
      "learning_rate": 3.926262626262626e-05,
      "loss": 0.0498,
      "step": 2226
    },
    {
      "epoch": 0.2227,
      "grad_norm": 12.687746047973633,
      "learning_rate": 3.925757575757576e-05,
      "loss": 0.2421,
      "step": 2227
    },
    {
      "epoch": 0.2228,
      "grad_norm": 16.378339767456055,
      "learning_rate": 3.9252525252525254e-05,
      "loss": 0.4311,
      "step": 2228
    },
    {
      "epoch": 0.2229,
      "grad_norm": 2.5767080783843994,
      "learning_rate": 3.924747474747475e-05,
      "loss": 0.2143,
      "step": 2229
    },
    {
      "epoch": 0.223,
      "grad_norm": 5.5981831550598145,
      "learning_rate": 3.924242424242424e-05,
      "loss": 0.2201,
      "step": 2230
    },
    {
      "epoch": 0.2231,
      "grad_norm": 2.9194414615631104,
      "learning_rate": 3.923737373737374e-05,
      "loss": 0.1283,
      "step": 2231
    },
    {
      "epoch": 0.2232,
      "grad_norm": 9.780818939208984,
      "learning_rate": 3.923232323232323e-05,
      "loss": 0.1346,
      "step": 2232
    },
    {
      "epoch": 0.2233,
      "grad_norm": 4.53388786315918,
      "learning_rate": 3.922727272727273e-05,
      "loss": 0.0895,
      "step": 2233
    },
    {
      "epoch": 0.2234,
      "grad_norm": 3.3282973766326904,
      "learning_rate": 3.922222222222223e-05,
      "loss": 0.0448,
      "step": 2234
    },
    {
      "epoch": 0.2235,
      "grad_norm": 7.081636905670166,
      "learning_rate": 3.921717171717172e-05,
      "loss": 0.6351,
      "step": 2235
    },
    {
      "epoch": 0.2236,
      "grad_norm": 4.572004318237305,
      "learning_rate": 3.9212121212121215e-05,
      "loss": 0.4099,
      "step": 2236
    },
    {
      "epoch": 0.2237,
      "grad_norm": 8.729022979736328,
      "learning_rate": 3.920707070707071e-05,
      "loss": 0.4203,
      "step": 2237
    },
    {
      "epoch": 0.2238,
      "grad_norm": 2.1789095401763916,
      "learning_rate": 3.920202020202021e-05,
      "loss": 0.2143,
      "step": 2238
    },
    {
      "epoch": 0.2239,
      "grad_norm": 7.239274978637695,
      "learning_rate": 3.9196969696969696e-05,
      "loss": 0.4151,
      "step": 2239
    },
    {
      "epoch": 0.224,
      "grad_norm": 5.593113422393799,
      "learning_rate": 3.91919191919192e-05,
      "loss": 0.3732,
      "step": 2240
    },
    {
      "epoch": 0.2241,
      "grad_norm": 7.207128524780273,
      "learning_rate": 3.918686868686869e-05,
      "loss": 0.3853,
      "step": 2241
    },
    {
      "epoch": 0.2242,
      "grad_norm": 8.09017276763916,
      "learning_rate": 3.9181818181818184e-05,
      "loss": 0.4908,
      "step": 2242
    },
    {
      "epoch": 0.2243,
      "grad_norm": 7.856698036193848,
      "learning_rate": 3.917676767676768e-05,
      "loss": 0.1466,
      "step": 2243
    },
    {
      "epoch": 0.2244,
      "grad_norm": 7.705386638641357,
      "learning_rate": 3.9171717171717176e-05,
      "loss": 0.4325,
      "step": 2244
    },
    {
      "epoch": 0.2245,
      "grad_norm": 7.76754903793335,
      "learning_rate": 3.9166666666666665e-05,
      "loss": 0.3629,
      "step": 2245
    },
    {
      "epoch": 0.2246,
      "grad_norm": 8.111959457397461,
      "learning_rate": 3.916161616161617e-05,
      "loss": 0.3594,
      "step": 2246
    },
    {
      "epoch": 0.2247,
      "grad_norm": 25.89757537841797,
      "learning_rate": 3.915656565656566e-05,
      "loss": 0.4447,
      "step": 2247
    },
    {
      "epoch": 0.2248,
      "grad_norm": 2.0617072582244873,
      "learning_rate": 3.9151515151515153e-05,
      "loss": 0.0353,
      "step": 2248
    },
    {
      "epoch": 0.2249,
      "grad_norm": 6.6421966552734375,
      "learning_rate": 3.914646464646465e-05,
      "loss": 0.0383,
      "step": 2249
    },
    {
      "epoch": 0.225,
      "grad_norm": 79.7930679321289,
      "learning_rate": 3.9141414141414145e-05,
      "loss": 0.1476,
      "step": 2250
    },
    {
      "epoch": 0.2251,
      "grad_norm": 5.3072404861450195,
      "learning_rate": 3.9136363636363635e-05,
      "loss": 0.3275,
      "step": 2251
    },
    {
      "epoch": 0.2252,
      "grad_norm": 3.8362131118774414,
      "learning_rate": 3.913131313131314e-05,
      "loss": 0.1471,
      "step": 2252
    },
    {
      "epoch": 0.2253,
      "grad_norm": 8.34691333770752,
      "learning_rate": 3.912626262626263e-05,
      "loss": 0.4485,
      "step": 2253
    },
    {
      "epoch": 0.2254,
      "grad_norm": 2.220384120941162,
      "learning_rate": 3.912121212121212e-05,
      "loss": 0.0387,
      "step": 2254
    },
    {
      "epoch": 0.2255,
      "grad_norm": 9.277286529541016,
      "learning_rate": 3.911616161616162e-05,
      "loss": 0.757,
      "step": 2255
    },
    {
      "epoch": 0.2256,
      "grad_norm": 5.809612274169922,
      "learning_rate": 3.9111111111111115e-05,
      "loss": 0.2934,
      "step": 2256
    },
    {
      "epoch": 0.2257,
      "grad_norm": 5.315961837768555,
      "learning_rate": 3.9106060606060604e-05,
      "loss": 0.0732,
      "step": 2257
    },
    {
      "epoch": 0.2258,
      "grad_norm": 6.099401950836182,
      "learning_rate": 3.9101010101010107e-05,
      "loss": 0.3908,
      "step": 2258
    },
    {
      "epoch": 0.2259,
      "grad_norm": 2.4850637912750244,
      "learning_rate": 3.9095959595959596e-05,
      "loss": 0.0721,
      "step": 2259
    },
    {
      "epoch": 0.226,
      "grad_norm": 14.918560981750488,
      "learning_rate": 3.909090909090909e-05,
      "loss": 0.4987,
      "step": 2260
    },
    {
      "epoch": 0.2261,
      "grad_norm": 2.3083674907684326,
      "learning_rate": 3.908585858585859e-05,
      "loss": 0.0509,
      "step": 2261
    },
    {
      "epoch": 0.2262,
      "grad_norm": 4.710660934448242,
      "learning_rate": 3.9080808080808084e-05,
      "loss": 0.175,
      "step": 2262
    },
    {
      "epoch": 0.2263,
      "grad_norm": 5.036222457885742,
      "learning_rate": 3.907575757575758e-05,
      "loss": 0.3045,
      "step": 2263
    },
    {
      "epoch": 0.2264,
      "grad_norm": 16.565387725830078,
      "learning_rate": 3.9070707070707076e-05,
      "loss": 0.3171,
      "step": 2264
    },
    {
      "epoch": 0.2265,
      "grad_norm": 6.332075595855713,
      "learning_rate": 3.9065656565656565e-05,
      "loss": 0.1422,
      "step": 2265
    },
    {
      "epoch": 0.2266,
      "grad_norm": 9.84310531616211,
      "learning_rate": 3.906060606060606e-05,
      "loss": 0.5074,
      "step": 2266
    },
    {
      "epoch": 0.2267,
      "grad_norm": 4.491461277008057,
      "learning_rate": 3.905555555555556e-05,
      "loss": 0.2831,
      "step": 2267
    },
    {
      "epoch": 0.2268,
      "grad_norm": 2.807532787322998,
      "learning_rate": 3.905050505050505e-05,
      "loss": 0.0542,
      "step": 2268
    },
    {
      "epoch": 0.2269,
      "grad_norm": 9.863914489746094,
      "learning_rate": 3.904545454545455e-05,
      "loss": 0.4103,
      "step": 2269
    },
    {
      "epoch": 0.227,
      "grad_norm": 5.0995025634765625,
      "learning_rate": 3.9040404040404045e-05,
      "loss": 0.1997,
      "step": 2270
    },
    {
      "epoch": 0.2271,
      "grad_norm": 1.394687294960022,
      "learning_rate": 3.9035353535353534e-05,
      "loss": 0.0242,
      "step": 2271
    },
    {
      "epoch": 0.2272,
      "grad_norm": 3.3676815032958984,
      "learning_rate": 3.903030303030304e-05,
      "loss": 0.0666,
      "step": 2272
    },
    {
      "epoch": 0.2273,
      "grad_norm": 8.832927703857422,
      "learning_rate": 3.9025252525252526e-05,
      "loss": 0.1398,
      "step": 2273
    },
    {
      "epoch": 0.2274,
      "grad_norm": 13.582135200500488,
      "learning_rate": 3.902020202020202e-05,
      "loss": 0.3667,
      "step": 2274
    },
    {
      "epoch": 0.2275,
      "grad_norm": 6.817351818084717,
      "learning_rate": 3.901515151515152e-05,
      "loss": 0.2836,
      "step": 2275
    },
    {
      "epoch": 0.2276,
      "grad_norm": 5.799720287322998,
      "learning_rate": 3.9010101010101014e-05,
      "loss": 0.1952,
      "step": 2276
    },
    {
      "epoch": 0.2277,
      "grad_norm": 4.5082855224609375,
      "learning_rate": 3.90050505050505e-05,
      "loss": 0.3718,
      "step": 2277
    },
    {
      "epoch": 0.2278,
      "grad_norm": 27.854509353637695,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 1.334,
      "step": 2278
    },
    {
      "epoch": 0.2279,
      "grad_norm": 6.946311950683594,
      "learning_rate": 3.8994949494949495e-05,
      "loss": 0.3526,
      "step": 2279
    },
    {
      "epoch": 0.228,
      "grad_norm": 39.11675262451172,
      "learning_rate": 3.898989898989899e-05,
      "loss": 0.436,
      "step": 2280
    },
    {
      "epoch": 0.2281,
      "grad_norm": 4.7964396476745605,
      "learning_rate": 3.898484848484849e-05,
      "loss": 0.1215,
      "step": 2281
    },
    {
      "epoch": 0.2282,
      "grad_norm": 3.5151922702789307,
      "learning_rate": 3.897979797979798e-05,
      "loss": 0.3236,
      "step": 2282
    },
    {
      "epoch": 0.2283,
      "grad_norm": 1.7429280281066895,
      "learning_rate": 3.897474747474747e-05,
      "loss": 0.0284,
      "step": 2283
    },
    {
      "epoch": 0.2284,
      "grad_norm": 5.563189506530762,
      "learning_rate": 3.8969696969696975e-05,
      "loss": 0.1098,
      "step": 2284
    },
    {
      "epoch": 0.2285,
      "grad_norm": 11.093931198120117,
      "learning_rate": 3.8964646464646464e-05,
      "loss": 1.3181,
      "step": 2285
    },
    {
      "epoch": 0.2286,
      "grad_norm": 4.659689426422119,
      "learning_rate": 3.895959595959596e-05,
      "loss": 0.1676,
      "step": 2286
    },
    {
      "epoch": 0.2287,
      "grad_norm": 5.190449237823486,
      "learning_rate": 3.8954545454545456e-05,
      "loss": 0.3781,
      "step": 2287
    },
    {
      "epoch": 0.2288,
      "grad_norm": 1.9286428689956665,
      "learning_rate": 3.894949494949495e-05,
      "loss": 0.0429,
      "step": 2288
    },
    {
      "epoch": 0.2289,
      "grad_norm": 5.397378444671631,
      "learning_rate": 3.894444444444444e-05,
      "loss": 0.3528,
      "step": 2289
    },
    {
      "epoch": 0.229,
      "grad_norm": 5.092958450317383,
      "learning_rate": 3.8939393939393944e-05,
      "loss": 0.2062,
      "step": 2290
    },
    {
      "epoch": 0.2291,
      "grad_norm": 6.338027477264404,
      "learning_rate": 3.8934343434343434e-05,
      "loss": 0.455,
      "step": 2291
    },
    {
      "epoch": 0.2292,
      "grad_norm": 19.089134216308594,
      "learning_rate": 3.892929292929293e-05,
      "loss": 0.3151,
      "step": 2292
    },
    {
      "epoch": 0.2293,
      "grad_norm": 5.876231670379639,
      "learning_rate": 3.8924242424242426e-05,
      "loss": 0.3459,
      "step": 2293
    },
    {
      "epoch": 0.2294,
      "grad_norm": 2.8735616207122803,
      "learning_rate": 3.891919191919192e-05,
      "loss": 0.3291,
      "step": 2294
    },
    {
      "epoch": 0.2295,
      "grad_norm": 4.163005828857422,
      "learning_rate": 3.891414141414141e-05,
      "loss": 0.079,
      "step": 2295
    },
    {
      "epoch": 0.2296,
      "grad_norm": 8.826871871948242,
      "learning_rate": 3.8909090909090914e-05,
      "loss": 0.3,
      "step": 2296
    },
    {
      "epoch": 0.2297,
      "grad_norm": 4.5849151611328125,
      "learning_rate": 3.89040404040404e-05,
      "loss": 0.3389,
      "step": 2297
    },
    {
      "epoch": 0.2298,
      "grad_norm": 16.761735916137695,
      "learning_rate": 3.88989898989899e-05,
      "loss": 0.3203,
      "step": 2298
    },
    {
      "epoch": 0.2299,
      "grad_norm": 10.47524642944336,
      "learning_rate": 3.8893939393939395e-05,
      "loss": 0.6606,
      "step": 2299
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.217395782470703,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.3137,
      "step": 2300
    },
    {
      "epoch": 0.2301,
      "grad_norm": 2.5194954872131348,
      "learning_rate": 3.888383838383839e-05,
      "loss": 0.1137,
      "step": 2301
    },
    {
      "epoch": 0.2302,
      "grad_norm": 2.825117349624634,
      "learning_rate": 3.887878787878788e-05,
      "loss": 0.1102,
      "step": 2302
    },
    {
      "epoch": 0.2303,
      "grad_norm": 4.630329608917236,
      "learning_rate": 3.887373737373737e-05,
      "loss": 0.2334,
      "step": 2303
    },
    {
      "epoch": 0.2304,
      "grad_norm": 2.1351699829101562,
      "learning_rate": 3.886868686868687e-05,
      "loss": 0.0684,
      "step": 2304
    },
    {
      "epoch": 0.2305,
      "grad_norm": 4.107018947601318,
      "learning_rate": 3.8863636363636364e-05,
      "loss": 0.2621,
      "step": 2305
    },
    {
      "epoch": 0.2306,
      "grad_norm": 6.122448921203613,
      "learning_rate": 3.885858585858586e-05,
      "loss": 0.5003,
      "step": 2306
    },
    {
      "epoch": 0.2307,
      "grad_norm": 8.221659660339355,
      "learning_rate": 3.8853535353535356e-05,
      "loss": 0.3885,
      "step": 2307
    },
    {
      "epoch": 0.2308,
      "grad_norm": 2.0784051418304443,
      "learning_rate": 3.884848484848485e-05,
      "loss": 0.085,
      "step": 2308
    },
    {
      "epoch": 0.2309,
      "grad_norm": 5.845290184020996,
      "learning_rate": 3.884343434343434e-05,
      "loss": 0.5684,
      "step": 2309
    },
    {
      "epoch": 0.231,
      "grad_norm": 6.107038974761963,
      "learning_rate": 3.8838383838383844e-05,
      "loss": 0.5415,
      "step": 2310
    },
    {
      "epoch": 0.2311,
      "grad_norm": 3.4729831218719482,
      "learning_rate": 3.883333333333333e-05,
      "loss": 0.1878,
      "step": 2311
    },
    {
      "epoch": 0.2312,
      "grad_norm": 2.0538697242736816,
      "learning_rate": 3.882828282828283e-05,
      "loss": 0.0266,
      "step": 2312
    },
    {
      "epoch": 0.2313,
      "grad_norm": 6.513246536254883,
      "learning_rate": 3.8823232323232325e-05,
      "loss": 0.3815,
      "step": 2313
    },
    {
      "epoch": 0.2314,
      "grad_norm": 3.9161901473999023,
      "learning_rate": 3.881818181818182e-05,
      "loss": 0.3816,
      "step": 2314
    },
    {
      "epoch": 0.2315,
      "grad_norm": 7.129893779754639,
      "learning_rate": 3.881313131313131e-05,
      "loss": 0.3423,
      "step": 2315
    },
    {
      "epoch": 0.2316,
      "grad_norm": 4.423102378845215,
      "learning_rate": 3.880808080808081e-05,
      "loss": 0.2122,
      "step": 2316
    },
    {
      "epoch": 0.2317,
      "grad_norm": 4.2824554443359375,
      "learning_rate": 3.88030303030303e-05,
      "loss": 0.2774,
      "step": 2317
    },
    {
      "epoch": 0.2318,
      "grad_norm": 4.655766010284424,
      "learning_rate": 3.87979797979798e-05,
      "loss": 0.2783,
      "step": 2318
    },
    {
      "epoch": 0.2319,
      "grad_norm": 5.215950012207031,
      "learning_rate": 3.8792929292929294e-05,
      "loss": 0.5694,
      "step": 2319
    },
    {
      "epoch": 0.232,
      "grad_norm": 3.7669498920440674,
      "learning_rate": 3.878787878787879e-05,
      "loss": 0.0892,
      "step": 2320
    },
    {
      "epoch": 0.2321,
      "grad_norm": 5.290677070617676,
      "learning_rate": 3.8782828282828286e-05,
      "loss": 0.705,
      "step": 2321
    },
    {
      "epoch": 0.2322,
      "grad_norm": 1.343183994293213,
      "learning_rate": 3.877777777777778e-05,
      "loss": 0.0503,
      "step": 2322
    },
    {
      "epoch": 0.2323,
      "grad_norm": 5.854480266571045,
      "learning_rate": 3.877272727272728e-05,
      "loss": 0.3112,
      "step": 2323
    },
    {
      "epoch": 0.2324,
      "grad_norm": 6.5727009773254395,
      "learning_rate": 3.876767676767677e-05,
      "loss": 0.248,
      "step": 2324
    },
    {
      "epoch": 0.2325,
      "grad_norm": 4.799156188964844,
      "learning_rate": 3.876262626262627e-05,
      "loss": 0.09,
      "step": 2325
    },
    {
      "epoch": 0.2326,
      "grad_norm": 5.1615891456604,
      "learning_rate": 3.875757575757576e-05,
      "loss": 0.6655,
      "step": 2326
    },
    {
      "epoch": 0.2327,
      "grad_norm": 4.465076446533203,
      "learning_rate": 3.8752525252525255e-05,
      "loss": 0.2625,
      "step": 2327
    },
    {
      "epoch": 0.2328,
      "grad_norm": 6.803846836090088,
      "learning_rate": 3.874747474747475e-05,
      "loss": 0.4334,
      "step": 2328
    },
    {
      "epoch": 0.2329,
      "grad_norm": 3.3307337760925293,
      "learning_rate": 3.874242424242425e-05,
      "loss": 0.1362,
      "step": 2329
    },
    {
      "epoch": 0.233,
      "grad_norm": 3.4149105548858643,
      "learning_rate": 3.8737373737373737e-05,
      "loss": 0.2662,
      "step": 2330
    },
    {
      "epoch": 0.2331,
      "grad_norm": 46.15982437133789,
      "learning_rate": 3.873232323232324e-05,
      "loss": 0.3049,
      "step": 2331
    },
    {
      "epoch": 0.2332,
      "grad_norm": 3.688364028930664,
      "learning_rate": 3.872727272727273e-05,
      "loss": 0.3021,
      "step": 2332
    },
    {
      "epoch": 0.2333,
      "grad_norm": 4.053861618041992,
      "learning_rate": 3.8722222222222225e-05,
      "loss": 0.1002,
      "step": 2333
    },
    {
      "epoch": 0.2334,
      "grad_norm": 13.495458602905273,
      "learning_rate": 3.871717171717172e-05,
      "loss": 0.2904,
      "step": 2334
    },
    {
      "epoch": 0.2335,
      "grad_norm": 4.4630866050720215,
      "learning_rate": 3.8712121212121217e-05,
      "loss": 0.3986,
      "step": 2335
    },
    {
      "epoch": 0.2336,
      "grad_norm": 5.9839630126953125,
      "learning_rate": 3.8707070707070706e-05,
      "loss": 0.1598,
      "step": 2336
    },
    {
      "epoch": 0.2337,
      "grad_norm": 4.3925299644470215,
      "learning_rate": 3.870202020202021e-05,
      "loss": 0.3389,
      "step": 2337
    },
    {
      "epoch": 0.2338,
      "grad_norm": 3.6000115871429443,
      "learning_rate": 3.86969696969697e-05,
      "loss": 0.1718,
      "step": 2338
    },
    {
      "epoch": 0.2339,
      "grad_norm": 4.3442583084106445,
      "learning_rate": 3.8691919191919194e-05,
      "loss": 0.1697,
      "step": 2339
    },
    {
      "epoch": 0.234,
      "grad_norm": 6.62675142288208,
      "learning_rate": 3.868686868686869e-05,
      "loss": 0.5787,
      "step": 2340
    },
    {
      "epoch": 0.2341,
      "grad_norm": 5.0971245765686035,
      "learning_rate": 3.8681818181818186e-05,
      "loss": 0.1013,
      "step": 2341
    },
    {
      "epoch": 0.2342,
      "grad_norm": 7.6519083976745605,
      "learning_rate": 3.8676767676767675e-05,
      "loss": 0.2139,
      "step": 2342
    },
    {
      "epoch": 0.2343,
      "grad_norm": 13.956546783447266,
      "learning_rate": 3.867171717171718e-05,
      "loss": 0.3853,
      "step": 2343
    },
    {
      "epoch": 0.2344,
      "grad_norm": 6.406900882720947,
      "learning_rate": 3.866666666666667e-05,
      "loss": 0.1257,
      "step": 2344
    },
    {
      "epoch": 0.2345,
      "grad_norm": 5.25191068649292,
      "learning_rate": 3.866161616161616e-05,
      "loss": 0.2102,
      "step": 2345
    },
    {
      "epoch": 0.2346,
      "grad_norm": 13.34771728515625,
      "learning_rate": 3.865656565656566e-05,
      "loss": 0.3227,
      "step": 2346
    },
    {
      "epoch": 0.2347,
      "grad_norm": 3.6030585765838623,
      "learning_rate": 3.8651515151515155e-05,
      "loss": 0.1569,
      "step": 2347
    },
    {
      "epoch": 0.2348,
      "grad_norm": 4.301846504211426,
      "learning_rate": 3.8646464646464644e-05,
      "loss": 0.1981,
      "step": 2348
    },
    {
      "epoch": 0.2349,
      "grad_norm": 4.496871471405029,
      "learning_rate": 3.864141414141415e-05,
      "loss": 0.1528,
      "step": 2349
    },
    {
      "epoch": 0.235,
      "grad_norm": 1.6852949857711792,
      "learning_rate": 3.8636363636363636e-05,
      "loss": 0.0234,
      "step": 2350
    },
    {
      "epoch": 0.2351,
      "grad_norm": 8.438323974609375,
      "learning_rate": 3.863131313131313e-05,
      "loss": 0.2455,
      "step": 2351
    },
    {
      "epoch": 0.2352,
      "grad_norm": 5.329364776611328,
      "learning_rate": 3.862626262626263e-05,
      "loss": 0.2144,
      "step": 2352
    },
    {
      "epoch": 0.2353,
      "grad_norm": 3.3149898052215576,
      "learning_rate": 3.8621212121212124e-05,
      "loss": 0.1899,
      "step": 2353
    },
    {
      "epoch": 0.2354,
      "grad_norm": 7.02733039855957,
      "learning_rate": 3.861616161616162e-05,
      "loss": 0.195,
      "step": 2354
    },
    {
      "epoch": 0.2355,
      "grad_norm": 7.401096343994141,
      "learning_rate": 3.8611111111111116e-05,
      "loss": 0.3314,
      "step": 2355
    },
    {
      "epoch": 0.2356,
      "grad_norm": 3.1484389305114746,
      "learning_rate": 3.8606060606060605e-05,
      "loss": 0.2961,
      "step": 2356
    },
    {
      "epoch": 0.2357,
      "grad_norm": 8.124480247497559,
      "learning_rate": 3.86010101010101e-05,
      "loss": 0.2551,
      "step": 2357
    },
    {
      "epoch": 0.2358,
      "grad_norm": 4.8340163230896,
      "learning_rate": 3.85959595959596e-05,
      "loss": 0.2124,
      "step": 2358
    },
    {
      "epoch": 0.2359,
      "grad_norm": 5.086196422576904,
      "learning_rate": 3.859090909090909e-05,
      "loss": 0.0572,
      "step": 2359
    },
    {
      "epoch": 0.236,
      "grad_norm": 32.38016128540039,
      "learning_rate": 3.858585858585859e-05,
      "loss": 0.2666,
      "step": 2360
    },
    {
      "epoch": 0.2361,
      "grad_norm": 6.9705986976623535,
      "learning_rate": 3.8580808080808085e-05,
      "loss": 0.4681,
      "step": 2361
    },
    {
      "epoch": 0.2362,
      "grad_norm": 3.4867215156555176,
      "learning_rate": 3.8575757575757574e-05,
      "loss": 0.1231,
      "step": 2362
    },
    {
      "epoch": 0.2363,
      "grad_norm": 8.118420600891113,
      "learning_rate": 3.857070707070708e-05,
      "loss": 0.3169,
      "step": 2363
    },
    {
      "epoch": 0.2364,
      "grad_norm": 5.347795009613037,
      "learning_rate": 3.8565656565656566e-05,
      "loss": 0.242,
      "step": 2364
    },
    {
      "epoch": 0.2365,
      "grad_norm": 4.808307647705078,
      "learning_rate": 3.856060606060606e-05,
      "loss": 0.3187,
      "step": 2365
    },
    {
      "epoch": 0.2366,
      "grad_norm": 3.9997098445892334,
      "learning_rate": 3.855555555555556e-05,
      "loss": 0.4037,
      "step": 2366
    },
    {
      "epoch": 0.2367,
      "grad_norm": 8.050063133239746,
      "learning_rate": 3.8550505050505054e-05,
      "loss": 0.8123,
      "step": 2367
    },
    {
      "epoch": 0.2368,
      "grad_norm": 2.6655683517456055,
      "learning_rate": 3.8545454545454544e-05,
      "loss": 0.233,
      "step": 2368
    },
    {
      "epoch": 0.2369,
      "grad_norm": 2.7267282009124756,
      "learning_rate": 3.8540404040404046e-05,
      "loss": 0.062,
      "step": 2369
    },
    {
      "epoch": 0.237,
      "grad_norm": 5.089876174926758,
      "learning_rate": 3.8535353535353536e-05,
      "loss": 0.7399,
      "step": 2370
    },
    {
      "epoch": 0.2371,
      "grad_norm": 14.358963012695312,
      "learning_rate": 3.853030303030303e-05,
      "loss": 0.2333,
      "step": 2371
    },
    {
      "epoch": 0.2372,
      "grad_norm": 2.421065092086792,
      "learning_rate": 3.852525252525253e-05,
      "loss": 0.1219,
      "step": 2372
    },
    {
      "epoch": 0.2373,
      "grad_norm": 3.759335994720459,
      "learning_rate": 3.8520202020202024e-05,
      "loss": 0.1961,
      "step": 2373
    },
    {
      "epoch": 0.2374,
      "grad_norm": 2.9975595474243164,
      "learning_rate": 3.851515151515151e-05,
      "loss": 0.0745,
      "step": 2374
    },
    {
      "epoch": 0.2375,
      "grad_norm": 3.3878772258758545,
      "learning_rate": 3.8510101010101015e-05,
      "loss": 0.314,
      "step": 2375
    },
    {
      "epoch": 0.2376,
      "grad_norm": 17.541996002197266,
      "learning_rate": 3.8505050505050505e-05,
      "loss": 0.3078,
      "step": 2376
    },
    {
      "epoch": 0.2377,
      "grad_norm": 3.999610185623169,
      "learning_rate": 3.85e-05,
      "loss": 0.0852,
      "step": 2377
    },
    {
      "epoch": 0.2378,
      "grad_norm": 4.405918598175049,
      "learning_rate": 3.84949494949495e-05,
      "loss": 0.3009,
      "step": 2378
    },
    {
      "epoch": 0.2379,
      "grad_norm": 11.225872039794922,
      "learning_rate": 3.848989898989899e-05,
      "loss": 0.1496,
      "step": 2379
    },
    {
      "epoch": 0.238,
      "grad_norm": 5.572174549102783,
      "learning_rate": 3.848484848484848e-05,
      "loss": 0.2906,
      "step": 2380
    },
    {
      "epoch": 0.2381,
      "grad_norm": 2.690718412399292,
      "learning_rate": 3.8479797979797985e-05,
      "loss": 0.2948,
      "step": 2381
    },
    {
      "epoch": 0.2382,
      "grad_norm": 18.919660568237305,
      "learning_rate": 3.8474747474747474e-05,
      "loss": 0.3575,
      "step": 2382
    },
    {
      "epoch": 0.2383,
      "grad_norm": 11.264904022216797,
      "learning_rate": 3.846969696969697e-05,
      "loss": 0.6081,
      "step": 2383
    },
    {
      "epoch": 0.2384,
      "grad_norm": 6.143326282501221,
      "learning_rate": 3.8464646464646466e-05,
      "loss": 0.7274,
      "step": 2384
    },
    {
      "epoch": 0.2385,
      "grad_norm": 10.243568420410156,
      "learning_rate": 3.845959595959596e-05,
      "loss": 0.1465,
      "step": 2385
    },
    {
      "epoch": 0.2386,
      "grad_norm": 6.898720741271973,
      "learning_rate": 3.845454545454545e-05,
      "loss": 0.5135,
      "step": 2386
    },
    {
      "epoch": 0.2387,
      "grad_norm": 4.384650230407715,
      "learning_rate": 3.8449494949494954e-05,
      "loss": 0.2246,
      "step": 2387
    },
    {
      "epoch": 0.2388,
      "grad_norm": 7.3950910568237305,
      "learning_rate": 3.844444444444444e-05,
      "loss": 0.5014,
      "step": 2388
    },
    {
      "epoch": 0.2389,
      "grad_norm": 5.206550598144531,
      "learning_rate": 3.843939393939394e-05,
      "loss": 0.1446,
      "step": 2389
    },
    {
      "epoch": 0.239,
      "grad_norm": 4.702919960021973,
      "learning_rate": 3.8434343434343435e-05,
      "loss": 0.4254,
      "step": 2390
    },
    {
      "epoch": 0.2391,
      "grad_norm": 12.128808975219727,
      "learning_rate": 3.842929292929293e-05,
      "loss": 0.4639,
      "step": 2391
    },
    {
      "epoch": 0.2392,
      "grad_norm": 11.71992015838623,
      "learning_rate": 3.842424242424243e-05,
      "loss": 0.8505,
      "step": 2392
    },
    {
      "epoch": 0.2393,
      "grad_norm": 4.31128454208374,
      "learning_rate": 3.841919191919192e-05,
      "loss": 0.3356,
      "step": 2393
    },
    {
      "epoch": 0.2394,
      "grad_norm": 2.6367876529693604,
      "learning_rate": 3.841414141414141e-05,
      "loss": 0.2674,
      "step": 2394
    },
    {
      "epoch": 0.2395,
      "grad_norm": 13.3275146484375,
      "learning_rate": 3.840909090909091e-05,
      "loss": 0.0951,
      "step": 2395
    },
    {
      "epoch": 0.2396,
      "grad_norm": 23.82736587524414,
      "learning_rate": 3.8404040404040404e-05,
      "loss": 0.2539,
      "step": 2396
    },
    {
      "epoch": 0.2397,
      "grad_norm": 5.193596839904785,
      "learning_rate": 3.83989898989899e-05,
      "loss": 0.035,
      "step": 2397
    },
    {
      "epoch": 0.2398,
      "grad_norm": 5.115236759185791,
      "learning_rate": 3.8393939393939396e-05,
      "loss": 0.0616,
      "step": 2398
    },
    {
      "epoch": 0.2399,
      "grad_norm": 6.033630847930908,
      "learning_rate": 3.838888888888889e-05,
      "loss": 0.0937,
      "step": 2399
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.201953887939453,
      "learning_rate": 3.838383838383838e-05,
      "loss": 0.3198,
      "step": 2400
    },
    {
      "epoch": 0.2401,
      "grad_norm": 2.540592908859253,
      "learning_rate": 3.837878787878788e-05,
      "loss": 0.0468,
      "step": 2401
    },
    {
      "epoch": 0.2402,
      "grad_norm": 1.4675401449203491,
      "learning_rate": 3.837373737373737e-05,
      "loss": 0.0491,
      "step": 2402
    },
    {
      "epoch": 0.2403,
      "grad_norm": 4.843230724334717,
      "learning_rate": 3.836868686868687e-05,
      "loss": 0.7481,
      "step": 2403
    },
    {
      "epoch": 0.2404,
      "grad_norm": 7.305363178253174,
      "learning_rate": 3.8363636363636365e-05,
      "loss": 0.1006,
      "step": 2404
    },
    {
      "epoch": 0.2405,
      "grad_norm": 3.5613269805908203,
      "learning_rate": 3.835858585858586e-05,
      "loss": 0.1173,
      "step": 2405
    },
    {
      "epoch": 0.2406,
      "grad_norm": 7.079519748687744,
      "learning_rate": 3.835353535353536e-05,
      "loss": 0.7437,
      "step": 2406
    },
    {
      "epoch": 0.2407,
      "grad_norm": 11.570077896118164,
      "learning_rate": 3.834848484848485e-05,
      "loss": 0.1469,
      "step": 2407
    },
    {
      "epoch": 0.2408,
      "grad_norm": 9.624913215637207,
      "learning_rate": 3.834343434343435e-05,
      "loss": 0.1166,
      "step": 2408
    },
    {
      "epoch": 0.2409,
      "grad_norm": 4.492124080657959,
      "learning_rate": 3.833838383838384e-05,
      "loss": 0.2492,
      "step": 2409
    },
    {
      "epoch": 0.241,
      "grad_norm": 3.948408603668213,
      "learning_rate": 3.8333333333333334e-05,
      "loss": 0.0907,
      "step": 2410
    },
    {
      "epoch": 0.2411,
      "grad_norm": 2.9169349670410156,
      "learning_rate": 3.832828282828283e-05,
      "loss": 0.0753,
      "step": 2411
    },
    {
      "epoch": 0.2412,
      "grad_norm": 3.1813066005706787,
      "learning_rate": 3.8323232323232326e-05,
      "loss": 0.3008,
      "step": 2412
    },
    {
      "epoch": 0.2413,
      "grad_norm": 3.0776524543762207,
      "learning_rate": 3.831818181818182e-05,
      "loss": 0.236,
      "step": 2413
    },
    {
      "epoch": 0.2414,
      "grad_norm": 10.647998809814453,
      "learning_rate": 3.831313131313132e-05,
      "loss": 0.9187,
      "step": 2414
    },
    {
      "epoch": 0.2415,
      "grad_norm": 6.605848789215088,
      "learning_rate": 3.830808080808081e-05,
      "loss": 0.4337,
      "step": 2415
    },
    {
      "epoch": 0.2416,
      "grad_norm": 5.979429244995117,
      "learning_rate": 3.830303030303031e-05,
      "loss": 0.1063,
      "step": 2416
    },
    {
      "epoch": 0.2417,
      "grad_norm": 8.172574996948242,
      "learning_rate": 3.82979797979798e-05,
      "loss": 0.1727,
      "step": 2417
    },
    {
      "epoch": 0.2418,
      "grad_norm": 1.8699843883514404,
      "learning_rate": 3.8292929292929296e-05,
      "loss": 0.0409,
      "step": 2418
    },
    {
      "epoch": 0.2419,
      "grad_norm": 7.610767364501953,
      "learning_rate": 3.828787878787879e-05,
      "loss": 0.4635,
      "step": 2419
    },
    {
      "epoch": 0.242,
      "grad_norm": 3.7038087844848633,
      "learning_rate": 3.828282828282829e-05,
      "loss": 0.1013,
      "step": 2420
    },
    {
      "epoch": 0.2421,
      "grad_norm": 2.0791513919830322,
      "learning_rate": 3.827777777777778e-05,
      "loss": 0.0755,
      "step": 2421
    },
    {
      "epoch": 0.2422,
      "grad_norm": 7.182448863983154,
      "learning_rate": 3.827272727272728e-05,
      "loss": 0.2429,
      "step": 2422
    },
    {
      "epoch": 0.2423,
      "grad_norm": 3.8985812664031982,
      "learning_rate": 3.826767676767677e-05,
      "loss": 0.123,
      "step": 2423
    },
    {
      "epoch": 0.2424,
      "grad_norm": 5.164766788482666,
      "learning_rate": 3.8262626262626265e-05,
      "loss": 0.1987,
      "step": 2424
    },
    {
      "epoch": 0.2425,
      "grad_norm": 4.0978169441223145,
      "learning_rate": 3.825757575757576e-05,
      "loss": 0.1148,
      "step": 2425
    },
    {
      "epoch": 0.2426,
      "grad_norm": 23.116600036621094,
      "learning_rate": 3.825252525252526e-05,
      "loss": 0.3104,
      "step": 2426
    },
    {
      "epoch": 0.2427,
      "grad_norm": 2.522035598754883,
      "learning_rate": 3.8247474747474746e-05,
      "loss": 0.0416,
      "step": 2427
    },
    {
      "epoch": 0.2428,
      "grad_norm": 5.793736457824707,
      "learning_rate": 3.824242424242425e-05,
      "loss": 0.2236,
      "step": 2428
    },
    {
      "epoch": 0.2429,
      "grad_norm": 3.8881609439849854,
      "learning_rate": 3.823737373737374e-05,
      "loss": 0.1157,
      "step": 2429
    },
    {
      "epoch": 0.243,
      "grad_norm": 2.540576696395874,
      "learning_rate": 3.8232323232323234e-05,
      "loss": 0.0804,
      "step": 2430
    },
    {
      "epoch": 0.2431,
      "grad_norm": 10.609162330627441,
      "learning_rate": 3.822727272727273e-05,
      "loss": 0.2712,
      "step": 2431
    },
    {
      "epoch": 0.2432,
      "grad_norm": 4.54896879196167,
      "learning_rate": 3.8222222222222226e-05,
      "loss": 0.2424,
      "step": 2432
    },
    {
      "epoch": 0.2433,
      "grad_norm": 9.99466609954834,
      "learning_rate": 3.8217171717171715e-05,
      "loss": 0.1914,
      "step": 2433
    },
    {
      "epoch": 0.2434,
      "grad_norm": 9.474411010742188,
      "learning_rate": 3.821212121212122e-05,
      "loss": 0.5273,
      "step": 2434
    },
    {
      "epoch": 0.2435,
      "grad_norm": 8.312530517578125,
      "learning_rate": 3.820707070707071e-05,
      "loss": 0.0861,
      "step": 2435
    },
    {
      "epoch": 0.2436,
      "grad_norm": 3.5466530323028564,
      "learning_rate": 3.82020202020202e-05,
      "loss": 0.3079,
      "step": 2436
    },
    {
      "epoch": 0.2437,
      "grad_norm": 5.638923168182373,
      "learning_rate": 3.81969696969697e-05,
      "loss": 0.2231,
      "step": 2437
    },
    {
      "epoch": 0.2438,
      "grad_norm": 5.811923980712891,
      "learning_rate": 3.8191919191919195e-05,
      "loss": 0.2089,
      "step": 2438
    },
    {
      "epoch": 0.2439,
      "grad_norm": 4.05006742477417,
      "learning_rate": 3.8186868686868684e-05,
      "loss": 0.3248,
      "step": 2439
    },
    {
      "epoch": 0.244,
      "grad_norm": 6.085812091827393,
      "learning_rate": 3.818181818181819e-05,
      "loss": 0.2215,
      "step": 2440
    },
    {
      "epoch": 0.2441,
      "grad_norm": 6.730982780456543,
      "learning_rate": 3.8176767676767676e-05,
      "loss": 0.287,
      "step": 2441
    },
    {
      "epoch": 0.2442,
      "grad_norm": 6.257259845733643,
      "learning_rate": 3.817171717171717e-05,
      "loss": 0.4048,
      "step": 2442
    },
    {
      "epoch": 0.2443,
      "grad_norm": 2.1789309978485107,
      "learning_rate": 3.816666666666667e-05,
      "loss": 0.0609,
      "step": 2443
    },
    {
      "epoch": 0.2444,
      "grad_norm": 3.0877444744110107,
      "learning_rate": 3.8161616161616164e-05,
      "loss": 0.0786,
      "step": 2444
    },
    {
      "epoch": 0.2445,
      "grad_norm": 42.50786590576172,
      "learning_rate": 3.815656565656566e-05,
      "loss": 0.2594,
      "step": 2445
    },
    {
      "epoch": 0.2446,
      "grad_norm": 11.88132095336914,
      "learning_rate": 3.8151515151515156e-05,
      "loss": 0.158,
      "step": 2446
    },
    {
      "epoch": 0.2447,
      "grad_norm": 4.8791303634643555,
      "learning_rate": 3.8146464646464645e-05,
      "loss": 0.2461,
      "step": 2447
    },
    {
      "epoch": 0.2448,
      "grad_norm": 5.376856327056885,
      "learning_rate": 3.814141414141414e-05,
      "loss": 0.1254,
      "step": 2448
    },
    {
      "epoch": 0.2449,
      "grad_norm": 2.9270811080932617,
      "learning_rate": 3.813636363636364e-05,
      "loss": 0.0652,
      "step": 2449
    },
    {
      "epoch": 0.245,
      "grad_norm": 5.220025539398193,
      "learning_rate": 3.8131313131313133e-05,
      "loss": 0.1225,
      "step": 2450
    },
    {
      "epoch": 0.2451,
      "grad_norm": 1.7911967039108276,
      "learning_rate": 3.812626262626263e-05,
      "loss": 0.0527,
      "step": 2451
    },
    {
      "epoch": 0.2452,
      "grad_norm": 3.46407151222229,
      "learning_rate": 3.8121212121212125e-05,
      "loss": 0.063,
      "step": 2452
    },
    {
      "epoch": 0.2453,
      "grad_norm": 4.593059062957764,
      "learning_rate": 3.8116161616161615e-05,
      "loss": 0.1262,
      "step": 2453
    },
    {
      "epoch": 0.2454,
      "grad_norm": 6.412887096405029,
      "learning_rate": 3.811111111111112e-05,
      "loss": 0.1662,
      "step": 2454
    },
    {
      "epoch": 0.2455,
      "grad_norm": 5.737298965454102,
      "learning_rate": 3.810606060606061e-05,
      "loss": 0.157,
      "step": 2455
    },
    {
      "epoch": 0.2456,
      "grad_norm": 27.898012161254883,
      "learning_rate": 3.81010101010101e-05,
      "loss": 0.2885,
      "step": 2456
    },
    {
      "epoch": 0.2457,
      "grad_norm": 7.0346174240112305,
      "learning_rate": 3.80959595959596e-05,
      "loss": 0.2874,
      "step": 2457
    },
    {
      "epoch": 0.2458,
      "grad_norm": 7.050204277038574,
      "learning_rate": 3.8090909090909095e-05,
      "loss": 0.1417,
      "step": 2458
    },
    {
      "epoch": 0.2459,
      "grad_norm": 2.0922489166259766,
      "learning_rate": 3.8085858585858584e-05,
      "loss": 0.025,
      "step": 2459
    },
    {
      "epoch": 0.246,
      "grad_norm": 11.11543083190918,
      "learning_rate": 3.8080808080808087e-05,
      "loss": 0.2362,
      "step": 2460
    },
    {
      "epoch": 0.2461,
      "grad_norm": 2.149506092071533,
      "learning_rate": 3.8075757575757576e-05,
      "loss": 0.0614,
      "step": 2461
    },
    {
      "epoch": 0.2462,
      "grad_norm": 3.5523030757904053,
      "learning_rate": 3.807070707070707e-05,
      "loss": 0.3691,
      "step": 2462
    },
    {
      "epoch": 0.2463,
      "grad_norm": 2.856654167175293,
      "learning_rate": 3.806565656565657e-05,
      "loss": 0.2851,
      "step": 2463
    },
    {
      "epoch": 0.2464,
      "grad_norm": 8.724230766296387,
      "learning_rate": 3.8060606060606064e-05,
      "loss": 0.2232,
      "step": 2464
    },
    {
      "epoch": 0.2465,
      "grad_norm": 8.642350196838379,
      "learning_rate": 3.805555555555555e-05,
      "loss": 0.2405,
      "step": 2465
    },
    {
      "epoch": 0.2466,
      "grad_norm": 2.184345245361328,
      "learning_rate": 3.8050505050505056e-05,
      "loss": 0.108,
      "step": 2466
    },
    {
      "epoch": 0.2467,
      "grad_norm": 1.371071457862854,
      "learning_rate": 3.8045454545454545e-05,
      "loss": 0.0347,
      "step": 2467
    },
    {
      "epoch": 0.2468,
      "grad_norm": 5.3519206047058105,
      "learning_rate": 3.804040404040404e-05,
      "loss": 0.7824,
      "step": 2468
    },
    {
      "epoch": 0.2469,
      "grad_norm": 5.468597412109375,
      "learning_rate": 3.803535353535354e-05,
      "loss": 0.2572,
      "step": 2469
    },
    {
      "epoch": 0.247,
      "grad_norm": 5.627052307128906,
      "learning_rate": 3.803030303030303e-05,
      "loss": 0.32,
      "step": 2470
    },
    {
      "epoch": 0.2471,
      "grad_norm": 7.940445423126221,
      "learning_rate": 3.802525252525252e-05,
      "loss": 0.5318,
      "step": 2471
    },
    {
      "epoch": 0.2472,
      "grad_norm": 10.461563110351562,
      "learning_rate": 3.8020202020202025e-05,
      "loss": 0.549,
      "step": 2472
    },
    {
      "epoch": 0.2473,
      "grad_norm": 5.235066890716553,
      "learning_rate": 3.8015151515151514e-05,
      "loss": 0.3204,
      "step": 2473
    },
    {
      "epoch": 0.2474,
      "grad_norm": 3.690664529800415,
      "learning_rate": 3.801010101010101e-05,
      "loss": 0.186,
      "step": 2474
    },
    {
      "epoch": 0.2475,
      "grad_norm": 8.429404258728027,
      "learning_rate": 3.8005050505050506e-05,
      "loss": 0.4032,
      "step": 2475
    },
    {
      "epoch": 0.2476,
      "grad_norm": 8.903319358825684,
      "learning_rate": 3.8e-05,
      "loss": 0.3344,
      "step": 2476
    },
    {
      "epoch": 0.2477,
      "grad_norm": 5.661881446838379,
      "learning_rate": 3.799494949494949e-05,
      "loss": 0.0217,
      "step": 2477
    },
    {
      "epoch": 0.2478,
      "grad_norm": 4.819735050201416,
      "learning_rate": 3.7989898989898994e-05,
      "loss": 0.5974,
      "step": 2478
    },
    {
      "epoch": 0.2479,
      "grad_norm": 7.977259159088135,
      "learning_rate": 3.798484848484848e-05,
      "loss": 0.0679,
      "step": 2479
    },
    {
      "epoch": 0.248,
      "grad_norm": 5.492610454559326,
      "learning_rate": 3.797979797979798e-05,
      "loss": 0.5374,
      "step": 2480
    },
    {
      "epoch": 0.2481,
      "grad_norm": 14.194738388061523,
      "learning_rate": 3.7974747474747475e-05,
      "loss": 0.2722,
      "step": 2481
    },
    {
      "epoch": 0.2482,
      "grad_norm": 3.974897623062134,
      "learning_rate": 3.796969696969697e-05,
      "loss": 0.17,
      "step": 2482
    },
    {
      "epoch": 0.2483,
      "grad_norm": 3.3003768920898438,
      "learning_rate": 3.796464646464647e-05,
      "loss": 0.0679,
      "step": 2483
    },
    {
      "epoch": 0.2484,
      "grad_norm": 3.1876766681671143,
      "learning_rate": 3.795959595959596e-05,
      "loss": 0.1046,
      "step": 2484
    },
    {
      "epoch": 0.2485,
      "grad_norm": 2.9874651432037354,
      "learning_rate": 3.795454545454545e-05,
      "loss": 0.0829,
      "step": 2485
    },
    {
      "epoch": 0.2486,
      "grad_norm": 10.73110294342041,
      "learning_rate": 3.794949494949495e-05,
      "loss": 0.1409,
      "step": 2486
    },
    {
      "epoch": 0.2487,
      "grad_norm": 2.869986057281494,
      "learning_rate": 3.7944444444444444e-05,
      "loss": 0.0112,
      "step": 2487
    },
    {
      "epoch": 0.2488,
      "grad_norm": 4.1134161949157715,
      "learning_rate": 3.793939393939394e-05,
      "loss": 0.1002,
      "step": 2488
    },
    {
      "epoch": 0.2489,
      "grad_norm": 6.884200572967529,
      "learning_rate": 3.7934343434343436e-05,
      "loss": 0.4302,
      "step": 2489
    },
    {
      "epoch": 0.249,
      "grad_norm": 7.3377251625061035,
      "learning_rate": 3.792929292929293e-05,
      "loss": 0.3858,
      "step": 2490
    },
    {
      "epoch": 0.2491,
      "grad_norm": 6.829806804656982,
      "learning_rate": 3.792424242424243e-05,
      "loss": 0.439,
      "step": 2491
    },
    {
      "epoch": 0.2492,
      "grad_norm": 3.750896453857422,
      "learning_rate": 3.791919191919192e-05,
      "loss": 0.1523,
      "step": 2492
    },
    {
      "epoch": 0.2493,
      "grad_norm": 1.7535009384155273,
      "learning_rate": 3.791414141414142e-05,
      "loss": 0.0498,
      "step": 2493
    },
    {
      "epoch": 0.2494,
      "grad_norm": 7.8987717628479,
      "learning_rate": 3.790909090909091e-05,
      "loss": 0.5508,
      "step": 2494
    },
    {
      "epoch": 0.2495,
      "grad_norm": 5.698704242706299,
      "learning_rate": 3.7904040404040406e-05,
      "loss": 0.2902,
      "step": 2495
    },
    {
      "epoch": 0.2496,
      "grad_norm": 5.336981773376465,
      "learning_rate": 3.78989898989899e-05,
      "loss": 0.3074,
      "step": 2496
    },
    {
      "epoch": 0.2497,
      "grad_norm": 3.807997465133667,
      "learning_rate": 3.78939393939394e-05,
      "loss": 0.3955,
      "step": 2497
    },
    {
      "epoch": 0.2498,
      "grad_norm": 2.7299749851226807,
      "learning_rate": 3.7888888888888894e-05,
      "loss": 0.1154,
      "step": 2498
    },
    {
      "epoch": 0.2499,
      "grad_norm": 3.6687254905700684,
      "learning_rate": 3.788383838383839e-05,
      "loss": 0.1628,
      "step": 2499
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.6052374839782715,
      "learning_rate": 3.787878787878788e-05,
      "loss": 0.3631,
      "step": 2500
    },
    {
      "epoch": 0.2501,
      "grad_norm": 3.633329391479492,
      "learning_rate": 3.7873737373737375e-05,
      "loss": 0.2326,
      "step": 2501
    },
    {
      "epoch": 0.2502,
      "grad_norm": 3.9573826789855957,
      "learning_rate": 3.786868686868687e-05,
      "loss": 0.1159,
      "step": 2502
    },
    {
      "epoch": 0.2503,
      "grad_norm": 3.9282114505767822,
      "learning_rate": 3.786363636363637e-05,
      "loss": 0.3193,
      "step": 2503
    },
    {
      "epoch": 0.2504,
      "grad_norm": 1.812964916229248,
      "learning_rate": 3.785858585858586e-05,
      "loss": 0.067,
      "step": 2504
    },
    {
      "epoch": 0.2505,
      "grad_norm": 5.184354782104492,
      "learning_rate": 3.785353535353536e-05,
      "loss": 0.3671,
      "step": 2505
    },
    {
      "epoch": 0.2506,
      "grad_norm": 3.2129504680633545,
      "learning_rate": 3.784848484848485e-05,
      "loss": 0.1217,
      "step": 2506
    },
    {
      "epoch": 0.2507,
      "grad_norm": 7.275712490081787,
      "learning_rate": 3.784343434343435e-05,
      "loss": 0.5419,
      "step": 2507
    },
    {
      "epoch": 0.2508,
      "grad_norm": 9.881250381469727,
      "learning_rate": 3.783838383838384e-05,
      "loss": 0.5213,
      "step": 2508
    },
    {
      "epoch": 0.2509,
      "grad_norm": 4.441324234008789,
      "learning_rate": 3.7833333333333336e-05,
      "loss": 0.3626,
      "step": 2509
    },
    {
      "epoch": 0.251,
      "grad_norm": 3.759693145751953,
      "learning_rate": 3.782828282828283e-05,
      "loss": 0.3382,
      "step": 2510
    },
    {
      "epoch": 0.2511,
      "grad_norm": 30.58207130432129,
      "learning_rate": 3.782323232323233e-05,
      "loss": 0.541,
      "step": 2511
    },
    {
      "epoch": 0.2512,
      "grad_norm": 4.918605804443359,
      "learning_rate": 3.781818181818182e-05,
      "loss": 0.1479,
      "step": 2512
    },
    {
      "epoch": 0.2513,
      "grad_norm": 5.019792079925537,
      "learning_rate": 3.781313131313132e-05,
      "loss": 0.3267,
      "step": 2513
    },
    {
      "epoch": 0.2514,
      "grad_norm": 6.868767261505127,
      "learning_rate": 3.780808080808081e-05,
      "loss": 0.5743,
      "step": 2514
    },
    {
      "epoch": 0.2515,
      "grad_norm": 3.8031582832336426,
      "learning_rate": 3.7803030303030305e-05,
      "loss": 0.1346,
      "step": 2515
    },
    {
      "epoch": 0.2516,
      "grad_norm": 3.4726181030273438,
      "learning_rate": 3.77979797979798e-05,
      "loss": 0.3615,
      "step": 2516
    },
    {
      "epoch": 0.2517,
      "grad_norm": 14.304558753967285,
      "learning_rate": 3.77929292929293e-05,
      "loss": 0.5607,
      "step": 2517
    },
    {
      "epoch": 0.2518,
      "grad_norm": 1.481461524963379,
      "learning_rate": 3.7787878787878786e-05,
      "loss": 0.0644,
      "step": 2518
    },
    {
      "epoch": 0.2519,
      "grad_norm": 4.621582508087158,
      "learning_rate": 3.778282828282829e-05,
      "loss": 0.3556,
      "step": 2519
    },
    {
      "epoch": 0.252,
      "grad_norm": 4.640269756317139,
      "learning_rate": 3.777777777777778e-05,
      "loss": 0.2782,
      "step": 2520
    },
    {
      "epoch": 0.2521,
      "grad_norm": 5.257242679595947,
      "learning_rate": 3.7772727272727274e-05,
      "loss": 0.3708,
      "step": 2521
    },
    {
      "epoch": 0.2522,
      "grad_norm": 3.2071194648742676,
      "learning_rate": 3.776767676767677e-05,
      "loss": 0.1201,
      "step": 2522
    },
    {
      "epoch": 0.2523,
      "grad_norm": 1.2259949445724487,
      "learning_rate": 3.7762626262626266e-05,
      "loss": 0.0416,
      "step": 2523
    },
    {
      "epoch": 0.2524,
      "grad_norm": 7.904854774475098,
      "learning_rate": 3.7757575757575755e-05,
      "loss": 0.1315,
      "step": 2524
    },
    {
      "epoch": 0.2525,
      "grad_norm": 10.247503280639648,
      "learning_rate": 3.775252525252526e-05,
      "loss": 0.5098,
      "step": 2525
    },
    {
      "epoch": 0.2526,
      "grad_norm": 3.4723827838897705,
      "learning_rate": 3.774747474747475e-05,
      "loss": 0.3035,
      "step": 2526
    },
    {
      "epoch": 0.2527,
      "grad_norm": 5.701580047607422,
      "learning_rate": 3.7742424242424243e-05,
      "loss": 0.4262,
      "step": 2527
    },
    {
      "epoch": 0.2528,
      "grad_norm": 10.670646667480469,
      "learning_rate": 3.773737373737374e-05,
      "loss": 0.0849,
      "step": 2528
    },
    {
      "epoch": 0.2529,
      "grad_norm": 2.908656358718872,
      "learning_rate": 3.7732323232323235e-05,
      "loss": 0.1741,
      "step": 2529
    },
    {
      "epoch": 0.253,
      "grad_norm": 9.526105880737305,
      "learning_rate": 3.7727272727272725e-05,
      "loss": 0.8764,
      "step": 2530
    },
    {
      "epoch": 0.2531,
      "grad_norm": 5.797723293304443,
      "learning_rate": 3.772222222222223e-05,
      "loss": 0.2458,
      "step": 2531
    },
    {
      "epoch": 0.2532,
      "grad_norm": 18.90694808959961,
      "learning_rate": 3.7717171717171717e-05,
      "loss": 0.1224,
      "step": 2532
    },
    {
      "epoch": 0.2533,
      "grad_norm": 2.3851711750030518,
      "learning_rate": 3.771212121212121e-05,
      "loss": 0.1262,
      "step": 2533
    },
    {
      "epoch": 0.2534,
      "grad_norm": 3.6586787700653076,
      "learning_rate": 3.770707070707071e-05,
      "loss": 0.0632,
      "step": 2534
    },
    {
      "epoch": 0.2535,
      "grad_norm": 2.5098912715911865,
      "learning_rate": 3.7702020202020205e-05,
      "loss": 0.0888,
      "step": 2535
    },
    {
      "epoch": 0.2536,
      "grad_norm": 4.493457794189453,
      "learning_rate": 3.76969696969697e-05,
      "loss": 0.0899,
      "step": 2536
    },
    {
      "epoch": 0.2537,
      "grad_norm": 2.9035980701446533,
      "learning_rate": 3.7691919191919197e-05,
      "loss": 0.1122,
      "step": 2537
    },
    {
      "epoch": 0.2538,
      "grad_norm": 2.8000340461730957,
      "learning_rate": 3.7686868686868686e-05,
      "loss": 0.0441,
      "step": 2538
    },
    {
      "epoch": 0.2539,
      "grad_norm": 1.285501480102539,
      "learning_rate": 3.768181818181818e-05,
      "loss": 0.0457,
      "step": 2539
    },
    {
      "epoch": 0.254,
      "grad_norm": 10.27978801727295,
      "learning_rate": 3.767676767676768e-05,
      "loss": 0.8906,
      "step": 2540
    },
    {
      "epoch": 0.2541,
      "grad_norm": 7.343108654022217,
      "learning_rate": 3.7671717171717174e-05,
      "loss": 0.4553,
      "step": 2541
    },
    {
      "epoch": 0.2542,
      "grad_norm": 1.2424449920654297,
      "learning_rate": 3.766666666666667e-05,
      "loss": 0.0693,
      "step": 2542
    },
    {
      "epoch": 0.2543,
      "grad_norm": 2.4560940265655518,
      "learning_rate": 3.7661616161616166e-05,
      "loss": 0.047,
      "step": 2543
    },
    {
      "epoch": 0.2544,
      "grad_norm": 10.09820556640625,
      "learning_rate": 3.7656565656565655e-05,
      "loss": 0.3152,
      "step": 2544
    },
    {
      "epoch": 0.2545,
      "grad_norm": 9.929243087768555,
      "learning_rate": 3.765151515151516e-05,
      "loss": 0.7116,
      "step": 2545
    },
    {
      "epoch": 0.2546,
      "grad_norm": 6.968231678009033,
      "learning_rate": 3.764646464646465e-05,
      "loss": 0.4125,
      "step": 2546
    },
    {
      "epoch": 0.2547,
      "grad_norm": 4.3772735595703125,
      "learning_rate": 3.764141414141414e-05,
      "loss": 0.1695,
      "step": 2547
    },
    {
      "epoch": 0.2548,
      "grad_norm": 10.159887313842773,
      "learning_rate": 3.763636363636364e-05,
      "loss": 1.0599,
      "step": 2548
    },
    {
      "epoch": 0.2549,
      "grad_norm": 5.133492946624756,
      "learning_rate": 3.7631313131313135e-05,
      "loss": 0.3563,
      "step": 2549
    },
    {
      "epoch": 0.255,
      "grad_norm": 4.317134380340576,
      "learning_rate": 3.7626262626262624e-05,
      "loss": 0.0761,
      "step": 2550
    },
    {
      "epoch": 0.2551,
      "grad_norm": 7.300112247467041,
      "learning_rate": 3.762121212121213e-05,
      "loss": 0.8203,
      "step": 2551
    },
    {
      "epoch": 0.2552,
      "grad_norm": 7.015380382537842,
      "learning_rate": 3.7616161616161616e-05,
      "loss": 0.2714,
      "step": 2552
    },
    {
      "epoch": 0.2553,
      "grad_norm": 11.204684257507324,
      "learning_rate": 3.761111111111111e-05,
      "loss": 0.5518,
      "step": 2553
    },
    {
      "epoch": 0.2554,
      "grad_norm": 10.623847007751465,
      "learning_rate": 3.760606060606061e-05,
      "loss": 0.5774,
      "step": 2554
    },
    {
      "epoch": 0.2555,
      "grad_norm": 22.857961654663086,
      "learning_rate": 3.7601010101010104e-05,
      "loss": 0.5922,
      "step": 2555
    },
    {
      "epoch": 0.2556,
      "grad_norm": 7.142025947570801,
      "learning_rate": 3.759595959595959e-05,
      "loss": 0.4438,
      "step": 2556
    },
    {
      "epoch": 0.2557,
      "grad_norm": 6.021360874176025,
      "learning_rate": 3.7590909090909096e-05,
      "loss": 0.169,
      "step": 2557
    },
    {
      "epoch": 0.2558,
      "grad_norm": 15.914841651916504,
      "learning_rate": 3.7585858585858585e-05,
      "loss": 0.4434,
      "step": 2558
    },
    {
      "epoch": 0.2559,
      "grad_norm": 5.046176433563232,
      "learning_rate": 3.758080808080808e-05,
      "loss": 0.1799,
      "step": 2559
    },
    {
      "epoch": 0.256,
      "grad_norm": 3.413914680480957,
      "learning_rate": 3.757575757575758e-05,
      "loss": 0.3138,
      "step": 2560
    },
    {
      "epoch": 0.2561,
      "grad_norm": 6.782406806945801,
      "learning_rate": 3.757070707070707e-05,
      "loss": 0.2808,
      "step": 2561
    },
    {
      "epoch": 0.2562,
      "grad_norm": 6.586654186248779,
      "learning_rate": 3.756565656565656e-05,
      "loss": 0.2094,
      "step": 2562
    },
    {
      "epoch": 0.2563,
      "grad_norm": 9.250658988952637,
      "learning_rate": 3.7560606060606065e-05,
      "loss": 0.5571,
      "step": 2563
    },
    {
      "epoch": 0.2564,
      "grad_norm": 7.606903076171875,
      "learning_rate": 3.7555555555555554e-05,
      "loss": 0.3318,
      "step": 2564
    },
    {
      "epoch": 0.2565,
      "grad_norm": 15.011571884155273,
      "learning_rate": 3.755050505050505e-05,
      "loss": 0.1669,
      "step": 2565
    },
    {
      "epoch": 0.2566,
      "grad_norm": 5.045196533203125,
      "learning_rate": 3.7545454545454546e-05,
      "loss": 0.2213,
      "step": 2566
    },
    {
      "epoch": 0.2567,
      "grad_norm": 3.573671340942383,
      "learning_rate": 3.754040404040404e-05,
      "loss": 0.1684,
      "step": 2567
    },
    {
      "epoch": 0.2568,
      "grad_norm": 8.765237808227539,
      "learning_rate": 3.753535353535353e-05,
      "loss": 0.1804,
      "step": 2568
    },
    {
      "epoch": 0.2569,
      "grad_norm": 5.151146411895752,
      "learning_rate": 3.7530303030303034e-05,
      "loss": 0.199,
      "step": 2569
    },
    {
      "epoch": 0.257,
      "grad_norm": 11.225188255310059,
      "learning_rate": 3.7525252525252524e-05,
      "loss": 0.2668,
      "step": 2570
    },
    {
      "epoch": 0.2571,
      "grad_norm": 0.9829931855201721,
      "learning_rate": 3.752020202020202e-05,
      "loss": 0.0168,
      "step": 2571
    },
    {
      "epoch": 0.2572,
      "grad_norm": 23.863903045654297,
      "learning_rate": 3.7515151515151516e-05,
      "loss": 0.2284,
      "step": 2572
    },
    {
      "epoch": 0.2573,
      "grad_norm": 4.925906658172607,
      "learning_rate": 3.751010101010101e-05,
      "loss": 0.1516,
      "step": 2573
    },
    {
      "epoch": 0.2574,
      "grad_norm": 10.38662338256836,
      "learning_rate": 3.75050505050505e-05,
      "loss": 0.2632,
      "step": 2574
    },
    {
      "epoch": 0.2575,
      "grad_norm": 8.911972045898438,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.2125,
      "step": 2575
    },
    {
      "epoch": 0.2576,
      "grad_norm": 3.5548484325408936,
      "learning_rate": 3.74949494949495e-05,
      "loss": 0.0851,
      "step": 2576
    },
    {
      "epoch": 0.2577,
      "grad_norm": 7.991683006286621,
      "learning_rate": 3.748989898989899e-05,
      "loss": 0.2701,
      "step": 2577
    },
    {
      "epoch": 0.2578,
      "grad_norm": 6.159383296966553,
      "learning_rate": 3.748484848484849e-05,
      "loss": 0.2923,
      "step": 2578
    },
    {
      "epoch": 0.2579,
      "grad_norm": 8.813909530639648,
      "learning_rate": 3.747979797979798e-05,
      "loss": 0.2304,
      "step": 2579
    },
    {
      "epoch": 0.258,
      "grad_norm": 3.2570931911468506,
      "learning_rate": 3.747474747474748e-05,
      "loss": 0.1311,
      "step": 2580
    },
    {
      "epoch": 0.2581,
      "grad_norm": 3.328310012817383,
      "learning_rate": 3.746969696969697e-05,
      "loss": 0.0942,
      "step": 2581
    },
    {
      "epoch": 0.2582,
      "grad_norm": 3.688584804534912,
      "learning_rate": 3.746464646464647e-05,
      "loss": 0.0985,
      "step": 2582
    },
    {
      "epoch": 0.2583,
      "grad_norm": 17.62662696838379,
      "learning_rate": 3.745959595959596e-05,
      "loss": 0.2091,
      "step": 2583
    },
    {
      "epoch": 0.2584,
      "grad_norm": 6.381011486053467,
      "learning_rate": 3.745454545454546e-05,
      "loss": 0.2356,
      "step": 2584
    },
    {
      "epoch": 0.2585,
      "grad_norm": 8.809460639953613,
      "learning_rate": 3.744949494949495e-05,
      "loss": 0.24,
      "step": 2585
    },
    {
      "epoch": 0.2586,
      "grad_norm": 7.465897083282471,
      "learning_rate": 3.7444444444444446e-05,
      "loss": 0.2813,
      "step": 2586
    },
    {
      "epoch": 0.2587,
      "grad_norm": 50.68670654296875,
      "learning_rate": 3.743939393939394e-05,
      "loss": 0.3432,
      "step": 2587
    },
    {
      "epoch": 0.2588,
      "grad_norm": 14.941656112670898,
      "learning_rate": 3.743434343434344e-05,
      "loss": 0.6174,
      "step": 2588
    },
    {
      "epoch": 0.2589,
      "grad_norm": 12.479443550109863,
      "learning_rate": 3.7429292929292934e-05,
      "loss": 0.1557,
      "step": 2589
    },
    {
      "epoch": 0.259,
      "grad_norm": 3.718564987182617,
      "learning_rate": 3.742424242424243e-05,
      "loss": 0.162,
      "step": 2590
    },
    {
      "epoch": 0.2591,
      "grad_norm": 15.67428970336914,
      "learning_rate": 3.741919191919192e-05,
      "loss": 0.3725,
      "step": 2591
    },
    {
      "epoch": 0.2592,
      "grad_norm": 17.447277069091797,
      "learning_rate": 3.7414141414141415e-05,
      "loss": 0.2298,
      "step": 2592
    },
    {
      "epoch": 0.2593,
      "grad_norm": 7.383298873901367,
      "learning_rate": 3.740909090909091e-05,
      "loss": 0.1201,
      "step": 2593
    },
    {
      "epoch": 0.2594,
      "grad_norm": 6.067961692810059,
      "learning_rate": 3.740404040404041e-05,
      "loss": 0.3574,
      "step": 2594
    },
    {
      "epoch": 0.2595,
      "grad_norm": 7.24599027633667,
      "learning_rate": 3.73989898989899e-05,
      "loss": 0.3,
      "step": 2595
    },
    {
      "epoch": 0.2596,
      "grad_norm": 16.44220733642578,
      "learning_rate": 3.73939393939394e-05,
      "loss": 0.2091,
      "step": 2596
    },
    {
      "epoch": 0.2597,
      "grad_norm": 6.298063278198242,
      "learning_rate": 3.738888888888889e-05,
      "loss": 0.0716,
      "step": 2597
    },
    {
      "epoch": 0.2598,
      "grad_norm": 8.184462547302246,
      "learning_rate": 3.738383838383839e-05,
      "loss": 0.3749,
      "step": 2598
    },
    {
      "epoch": 0.2599,
      "grad_norm": 2.725050449371338,
      "learning_rate": 3.737878787878788e-05,
      "loss": 0.1048,
      "step": 2599
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.904613733291626,
      "learning_rate": 3.7373737373737376e-05,
      "loss": 0.1162,
      "step": 2600
    },
    {
      "epoch": 0.2601,
      "grad_norm": 4.033769607543945,
      "learning_rate": 3.736868686868687e-05,
      "loss": 0.135,
      "step": 2601
    },
    {
      "epoch": 0.2602,
      "grad_norm": 3.5412299633026123,
      "learning_rate": 3.736363636363637e-05,
      "loss": 0.1691,
      "step": 2602
    },
    {
      "epoch": 0.2603,
      "grad_norm": 4.707581996917725,
      "learning_rate": 3.735858585858586e-05,
      "loss": 0.189,
      "step": 2603
    },
    {
      "epoch": 0.2604,
      "grad_norm": 3.3441050052642822,
      "learning_rate": 3.735353535353536e-05,
      "loss": 0.1462,
      "step": 2604
    },
    {
      "epoch": 0.2605,
      "grad_norm": 5.514741897583008,
      "learning_rate": 3.734848484848485e-05,
      "loss": 0.0972,
      "step": 2605
    },
    {
      "epoch": 0.2606,
      "grad_norm": 3.4280900955200195,
      "learning_rate": 3.7343434343434345e-05,
      "loss": 0.1091,
      "step": 2606
    },
    {
      "epoch": 0.2607,
      "grad_norm": 2.104914903640747,
      "learning_rate": 3.733838383838384e-05,
      "loss": 0.0631,
      "step": 2607
    },
    {
      "epoch": 0.2608,
      "grad_norm": 4.163397312164307,
      "learning_rate": 3.733333333333334e-05,
      "loss": 0.3216,
      "step": 2608
    },
    {
      "epoch": 0.2609,
      "grad_norm": 4.653924465179443,
      "learning_rate": 3.7328282828282827e-05,
      "loss": 0.1004,
      "step": 2609
    },
    {
      "epoch": 0.261,
      "grad_norm": 4.302547931671143,
      "learning_rate": 3.732323232323233e-05,
      "loss": 0.0873,
      "step": 2610
    },
    {
      "epoch": 0.2611,
      "grad_norm": 4.646703720092773,
      "learning_rate": 3.731818181818182e-05,
      "loss": 0.1266,
      "step": 2611
    },
    {
      "epoch": 0.2612,
      "grad_norm": 1.8911460638046265,
      "learning_rate": 3.7313131313131314e-05,
      "loss": 0.0381,
      "step": 2612
    },
    {
      "epoch": 0.2613,
      "grad_norm": 17.941682815551758,
      "learning_rate": 3.730808080808081e-05,
      "loss": 0.318,
      "step": 2613
    },
    {
      "epoch": 0.2614,
      "grad_norm": 13.040088653564453,
      "learning_rate": 3.7303030303030306e-05,
      "loss": 0.3412,
      "step": 2614
    },
    {
      "epoch": 0.2615,
      "grad_norm": 4.166217803955078,
      "learning_rate": 3.7297979797979796e-05,
      "loss": 0.0492,
      "step": 2615
    },
    {
      "epoch": 0.2616,
      "grad_norm": 8.978730201721191,
      "learning_rate": 3.72929292929293e-05,
      "loss": 0.458,
      "step": 2616
    },
    {
      "epoch": 0.2617,
      "grad_norm": 5.099554538726807,
      "learning_rate": 3.728787878787879e-05,
      "loss": 0.2897,
      "step": 2617
    },
    {
      "epoch": 0.2618,
      "grad_norm": 2.6540472507476807,
      "learning_rate": 3.7282828282828284e-05,
      "loss": 0.1224,
      "step": 2618
    },
    {
      "epoch": 0.2619,
      "grad_norm": 5.016241550445557,
      "learning_rate": 3.727777777777778e-05,
      "loss": 0.0381,
      "step": 2619
    },
    {
      "epoch": 0.262,
      "grad_norm": 5.281260967254639,
      "learning_rate": 3.7272727272727276e-05,
      "loss": 0.2779,
      "step": 2620
    },
    {
      "epoch": 0.2621,
      "grad_norm": 2.1882247924804688,
      "learning_rate": 3.7267676767676765e-05,
      "loss": 0.1106,
      "step": 2621
    },
    {
      "epoch": 0.2622,
      "grad_norm": 4.3451948165893555,
      "learning_rate": 3.726262626262627e-05,
      "loss": 0.1705,
      "step": 2622
    },
    {
      "epoch": 0.2623,
      "grad_norm": 6.787327289581299,
      "learning_rate": 3.725757575757576e-05,
      "loss": 0.0704,
      "step": 2623
    },
    {
      "epoch": 0.2624,
      "grad_norm": 11.07645034790039,
      "learning_rate": 3.725252525252525e-05,
      "loss": 0.3576,
      "step": 2624
    },
    {
      "epoch": 0.2625,
      "grad_norm": 6.848848819732666,
      "learning_rate": 3.724747474747475e-05,
      "loss": 0.2807,
      "step": 2625
    },
    {
      "epoch": 0.2626,
      "grad_norm": 3.02716064453125,
      "learning_rate": 3.7242424242424245e-05,
      "loss": 0.0309,
      "step": 2626
    },
    {
      "epoch": 0.2627,
      "grad_norm": 9.12061595916748,
      "learning_rate": 3.723737373737374e-05,
      "loss": 0.0682,
      "step": 2627
    },
    {
      "epoch": 0.2628,
      "grad_norm": 6.9251885414123535,
      "learning_rate": 3.723232323232324e-05,
      "loss": 0.3565,
      "step": 2628
    },
    {
      "epoch": 0.2629,
      "grad_norm": 2.3400938510894775,
      "learning_rate": 3.7227272727272726e-05,
      "loss": 0.0224,
      "step": 2629
    },
    {
      "epoch": 0.263,
      "grad_norm": 2.232347011566162,
      "learning_rate": 3.722222222222222e-05,
      "loss": 0.1287,
      "step": 2630
    },
    {
      "epoch": 0.2631,
      "grad_norm": 3.6365950107574463,
      "learning_rate": 3.721717171717172e-05,
      "loss": 0.2308,
      "step": 2631
    },
    {
      "epoch": 0.2632,
      "grad_norm": 8.775969505310059,
      "learning_rate": 3.7212121212121214e-05,
      "loss": 0.4635,
      "step": 2632
    },
    {
      "epoch": 0.2633,
      "grad_norm": 1.3536441326141357,
      "learning_rate": 3.720707070707071e-05,
      "loss": 0.0403,
      "step": 2633
    },
    {
      "epoch": 0.2634,
      "grad_norm": 10.622013092041016,
      "learning_rate": 3.7202020202020206e-05,
      "loss": 0.7014,
      "step": 2634
    },
    {
      "epoch": 0.2635,
      "grad_norm": 4.288541316986084,
      "learning_rate": 3.7196969696969695e-05,
      "loss": 0.2311,
      "step": 2635
    },
    {
      "epoch": 0.2636,
      "grad_norm": 1.5616534948349,
      "learning_rate": 3.719191919191919e-05,
      "loss": 0.024,
      "step": 2636
    },
    {
      "epoch": 0.2637,
      "grad_norm": 2.6443464756011963,
      "learning_rate": 3.718686868686869e-05,
      "loss": 0.1084,
      "step": 2637
    },
    {
      "epoch": 0.2638,
      "grad_norm": 54.99334716796875,
      "learning_rate": 3.718181818181818e-05,
      "loss": 0.3015,
      "step": 2638
    },
    {
      "epoch": 0.2639,
      "grad_norm": 10.006834030151367,
      "learning_rate": 3.717676767676768e-05,
      "loss": 0.6562,
      "step": 2639
    },
    {
      "epoch": 0.264,
      "grad_norm": 3.195986032485962,
      "learning_rate": 3.7171717171717175e-05,
      "loss": 0.0652,
      "step": 2640
    },
    {
      "epoch": 0.2641,
      "grad_norm": 3.0788705348968506,
      "learning_rate": 3.7166666666666664e-05,
      "loss": 0.1025,
      "step": 2641
    },
    {
      "epoch": 0.2642,
      "grad_norm": 4.25254487991333,
      "learning_rate": 3.716161616161617e-05,
      "loss": 0.1045,
      "step": 2642
    },
    {
      "epoch": 0.2643,
      "grad_norm": 5.172402858734131,
      "learning_rate": 3.7156565656565656e-05,
      "loss": 0.1004,
      "step": 2643
    },
    {
      "epoch": 0.2644,
      "grad_norm": 9.845352172851562,
      "learning_rate": 3.715151515151515e-05,
      "loss": 0.4852,
      "step": 2644
    },
    {
      "epoch": 0.2645,
      "grad_norm": 7.054336071014404,
      "learning_rate": 3.714646464646465e-05,
      "loss": 0.3791,
      "step": 2645
    },
    {
      "epoch": 0.2646,
      "grad_norm": 2.7373409271240234,
      "learning_rate": 3.7141414141414144e-05,
      "loss": 0.0948,
      "step": 2646
    },
    {
      "epoch": 0.2647,
      "grad_norm": 11.166977882385254,
      "learning_rate": 3.7136363636363633e-05,
      "loss": 0.8422,
      "step": 2647
    },
    {
      "epoch": 0.2648,
      "grad_norm": 14.717024803161621,
      "learning_rate": 3.7131313131313136e-05,
      "loss": 0.0522,
      "step": 2648
    },
    {
      "epoch": 0.2649,
      "grad_norm": 2.602900743484497,
      "learning_rate": 3.7126262626262625e-05,
      "loss": 0.1457,
      "step": 2649
    },
    {
      "epoch": 0.265,
      "grad_norm": 10.732046127319336,
      "learning_rate": 3.712121212121212e-05,
      "loss": 0.4913,
      "step": 2650
    },
    {
      "epoch": 0.2651,
      "grad_norm": 8.339414596557617,
      "learning_rate": 3.711616161616162e-05,
      "loss": 0.4423,
      "step": 2651
    },
    {
      "epoch": 0.2652,
      "grad_norm": 4.317476272583008,
      "learning_rate": 3.7111111111111113e-05,
      "loss": 0.1506,
      "step": 2652
    },
    {
      "epoch": 0.2653,
      "grad_norm": 3.151411294937134,
      "learning_rate": 3.71060606060606e-05,
      "loss": 0.1232,
      "step": 2653
    },
    {
      "epoch": 0.2654,
      "grad_norm": 3.356527805328369,
      "learning_rate": 3.7101010101010105e-05,
      "loss": 0.0869,
      "step": 2654
    },
    {
      "epoch": 0.2655,
      "grad_norm": 5.004629611968994,
      "learning_rate": 3.7095959595959595e-05,
      "loss": 0.0797,
      "step": 2655
    },
    {
      "epoch": 0.2656,
      "grad_norm": 4.777871608734131,
      "learning_rate": 3.709090909090909e-05,
      "loss": 0.1911,
      "step": 2656
    },
    {
      "epoch": 0.2657,
      "grad_norm": 4.3874831199646,
      "learning_rate": 3.7085858585858587e-05,
      "loss": 0.2823,
      "step": 2657
    },
    {
      "epoch": 0.2658,
      "grad_norm": 7.243968486785889,
      "learning_rate": 3.708080808080808e-05,
      "loss": 0.3916,
      "step": 2658
    },
    {
      "epoch": 0.2659,
      "grad_norm": 3.9555346965789795,
      "learning_rate": 3.707575757575757e-05,
      "loss": 0.0894,
      "step": 2659
    },
    {
      "epoch": 0.266,
      "grad_norm": 9.602887153625488,
      "learning_rate": 3.7070707070707075e-05,
      "loss": 0.7856,
      "step": 2660
    },
    {
      "epoch": 0.2661,
      "grad_norm": 4.997343063354492,
      "learning_rate": 3.706565656565657e-05,
      "loss": 0.2069,
      "step": 2661
    },
    {
      "epoch": 0.2662,
      "grad_norm": 14.014642715454102,
      "learning_rate": 3.706060606060606e-05,
      "loss": 0.36,
      "step": 2662
    },
    {
      "epoch": 0.2663,
      "grad_norm": 9.724322319030762,
      "learning_rate": 3.705555555555556e-05,
      "loss": 0.0684,
      "step": 2663
    },
    {
      "epoch": 0.2664,
      "grad_norm": 6.705308437347412,
      "learning_rate": 3.705050505050505e-05,
      "loss": 0.2173,
      "step": 2664
    },
    {
      "epoch": 0.2665,
      "grad_norm": 16.96195411682129,
      "learning_rate": 3.704545454545455e-05,
      "loss": 0.2451,
      "step": 2665
    },
    {
      "epoch": 0.2666,
      "grad_norm": 2.5108509063720703,
      "learning_rate": 3.7040404040404044e-05,
      "loss": 0.0804,
      "step": 2666
    },
    {
      "epoch": 0.2667,
      "grad_norm": 7.62470006942749,
      "learning_rate": 3.703535353535354e-05,
      "loss": 0.2041,
      "step": 2667
    },
    {
      "epoch": 0.2668,
      "grad_norm": 7.759490966796875,
      "learning_rate": 3.703030303030303e-05,
      "loss": 0.2575,
      "step": 2668
    },
    {
      "epoch": 0.2669,
      "grad_norm": 16.121610641479492,
      "learning_rate": 3.702525252525253e-05,
      "loss": 0.3171,
      "step": 2669
    },
    {
      "epoch": 0.267,
      "grad_norm": 7.1227641105651855,
      "learning_rate": 3.702020202020202e-05,
      "loss": 0.2651,
      "step": 2670
    },
    {
      "epoch": 0.2671,
      "grad_norm": 7.040735721588135,
      "learning_rate": 3.701515151515152e-05,
      "loss": 0.5903,
      "step": 2671
    },
    {
      "epoch": 0.2672,
      "grad_norm": 11.586650848388672,
      "learning_rate": 3.701010101010101e-05,
      "loss": 0.6662,
      "step": 2672
    },
    {
      "epoch": 0.2673,
      "grad_norm": 5.11433744430542,
      "learning_rate": 3.700505050505051e-05,
      "loss": 0.1814,
      "step": 2673
    },
    {
      "epoch": 0.2674,
      "grad_norm": 7.755463123321533,
      "learning_rate": 3.7e-05,
      "loss": 0.0952,
      "step": 2674
    },
    {
      "epoch": 0.2675,
      "grad_norm": 8.785964965820312,
      "learning_rate": 3.69949494949495e-05,
      "loss": 0.3463,
      "step": 2675
    },
    {
      "epoch": 0.2676,
      "grad_norm": 39.36437225341797,
      "learning_rate": 3.698989898989899e-05,
      "loss": 0.7789,
      "step": 2676
    },
    {
      "epoch": 0.2677,
      "grad_norm": 25.625558853149414,
      "learning_rate": 3.6984848484848486e-05,
      "loss": 0.3817,
      "step": 2677
    },
    {
      "epoch": 0.2678,
      "grad_norm": 5.592584609985352,
      "learning_rate": 3.697979797979798e-05,
      "loss": 0.2361,
      "step": 2678
    },
    {
      "epoch": 0.2679,
      "grad_norm": 8.243741989135742,
      "learning_rate": 3.697474747474748e-05,
      "loss": 0.4659,
      "step": 2679
    },
    {
      "epoch": 0.268,
      "grad_norm": 9.804769515991211,
      "learning_rate": 3.6969696969696974e-05,
      "loss": 0.1433,
      "step": 2680
    },
    {
      "epoch": 0.2681,
      "grad_norm": 4.059830665588379,
      "learning_rate": 3.696464646464647e-05,
      "loss": 0.1724,
      "step": 2681
    },
    {
      "epoch": 0.2682,
      "grad_norm": 10.261710166931152,
      "learning_rate": 3.695959595959596e-05,
      "loss": 0.4043,
      "step": 2682
    },
    {
      "epoch": 0.2683,
      "grad_norm": 4.198519229888916,
      "learning_rate": 3.6954545454545455e-05,
      "loss": 0.241,
      "step": 2683
    },
    {
      "epoch": 0.2684,
      "grad_norm": 2.4639840126037598,
      "learning_rate": 3.694949494949495e-05,
      "loss": 0.0697,
      "step": 2684
    },
    {
      "epoch": 0.2685,
      "grad_norm": 4.386950969696045,
      "learning_rate": 3.694444444444445e-05,
      "loss": 0.17,
      "step": 2685
    },
    {
      "epoch": 0.2686,
      "grad_norm": 30.131675720214844,
      "learning_rate": 3.693939393939394e-05,
      "loss": 0.4085,
      "step": 2686
    },
    {
      "epoch": 0.2687,
      "grad_norm": 2.2709295749664307,
      "learning_rate": 3.693434343434344e-05,
      "loss": 0.0686,
      "step": 2687
    },
    {
      "epoch": 0.2688,
      "grad_norm": 3.01920223236084,
      "learning_rate": 3.692929292929293e-05,
      "loss": 0.0992,
      "step": 2688
    },
    {
      "epoch": 0.2689,
      "grad_norm": 4.122603416442871,
      "learning_rate": 3.692424242424243e-05,
      "loss": 0.1018,
      "step": 2689
    },
    {
      "epoch": 0.269,
      "grad_norm": 7.591348171234131,
      "learning_rate": 3.691919191919192e-05,
      "loss": 0.45,
      "step": 2690
    },
    {
      "epoch": 0.2691,
      "grad_norm": 7.751885890960693,
      "learning_rate": 3.6914141414141416e-05,
      "loss": 0.4082,
      "step": 2691
    },
    {
      "epoch": 0.2692,
      "grad_norm": 3.6076364517211914,
      "learning_rate": 3.690909090909091e-05,
      "loss": 0.0506,
      "step": 2692
    },
    {
      "epoch": 0.2693,
      "grad_norm": 10.021163940429688,
      "learning_rate": 3.690404040404041e-05,
      "loss": 0.6506,
      "step": 2693
    },
    {
      "epoch": 0.2694,
      "grad_norm": 2.865628242492676,
      "learning_rate": 3.68989898989899e-05,
      "loss": 0.1037,
      "step": 2694
    },
    {
      "epoch": 0.2695,
      "grad_norm": 7.257493019104004,
      "learning_rate": 3.68939393939394e-05,
      "loss": 0.3149,
      "step": 2695
    },
    {
      "epoch": 0.2696,
      "grad_norm": 4.9068379402160645,
      "learning_rate": 3.688888888888889e-05,
      "loss": 0.3117,
      "step": 2696
    },
    {
      "epoch": 0.2697,
      "grad_norm": 3.992971181869507,
      "learning_rate": 3.6883838383838386e-05,
      "loss": 0.1448,
      "step": 2697
    },
    {
      "epoch": 0.2698,
      "grad_norm": 29.156503677368164,
      "learning_rate": 3.687878787878788e-05,
      "loss": 0.2645,
      "step": 2698
    },
    {
      "epoch": 0.2699,
      "grad_norm": 4.404965877532959,
      "learning_rate": 3.687373737373738e-05,
      "loss": 0.2043,
      "step": 2699
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.020088195800781,
      "learning_rate": 3.686868686868687e-05,
      "loss": 0.2864,
      "step": 2700
    },
    {
      "epoch": 0.2701,
      "grad_norm": 9.632144927978516,
      "learning_rate": 3.686363636363637e-05,
      "loss": 0.251,
      "step": 2701
    },
    {
      "epoch": 0.2702,
      "grad_norm": 6.249451637268066,
      "learning_rate": 3.685858585858586e-05,
      "loss": 0.3698,
      "step": 2702
    },
    {
      "epoch": 0.2703,
      "grad_norm": 4.373552322387695,
      "learning_rate": 3.6853535353535355e-05,
      "loss": 0.2001,
      "step": 2703
    },
    {
      "epoch": 0.2704,
      "grad_norm": 11.288521766662598,
      "learning_rate": 3.684848484848485e-05,
      "loss": 0.4468,
      "step": 2704
    },
    {
      "epoch": 0.2705,
      "grad_norm": 4.66939640045166,
      "learning_rate": 3.684343434343435e-05,
      "loss": 0.1107,
      "step": 2705
    },
    {
      "epoch": 0.2706,
      "grad_norm": 11.752008438110352,
      "learning_rate": 3.6838383838383836e-05,
      "loss": 0.2199,
      "step": 2706
    },
    {
      "epoch": 0.2707,
      "grad_norm": 7.3769001960754395,
      "learning_rate": 3.683333333333334e-05,
      "loss": 0.2205,
      "step": 2707
    },
    {
      "epoch": 0.2708,
      "grad_norm": 6.40695858001709,
      "learning_rate": 3.682828282828283e-05,
      "loss": 0.4304,
      "step": 2708
    },
    {
      "epoch": 0.2709,
      "grad_norm": 10.439391136169434,
      "learning_rate": 3.6823232323232324e-05,
      "loss": 0.3397,
      "step": 2709
    },
    {
      "epoch": 0.271,
      "grad_norm": 5.225971221923828,
      "learning_rate": 3.681818181818182e-05,
      "loss": 0.2819,
      "step": 2710
    },
    {
      "epoch": 0.2711,
      "grad_norm": 13.576817512512207,
      "learning_rate": 3.6813131313131316e-05,
      "loss": 0.4869,
      "step": 2711
    },
    {
      "epoch": 0.2712,
      "grad_norm": 4.117631912231445,
      "learning_rate": 3.6808080808080805e-05,
      "loss": 0.1719,
      "step": 2712
    },
    {
      "epoch": 0.2713,
      "grad_norm": 3.9639689922332764,
      "learning_rate": 3.680303030303031e-05,
      "loss": 0.1022,
      "step": 2713
    },
    {
      "epoch": 0.2714,
      "grad_norm": 7.069735527038574,
      "learning_rate": 3.67979797979798e-05,
      "loss": 0.2595,
      "step": 2714
    },
    {
      "epoch": 0.2715,
      "grad_norm": 3.696424961090088,
      "learning_rate": 3.679292929292929e-05,
      "loss": 0.1486,
      "step": 2715
    },
    {
      "epoch": 0.2716,
      "grad_norm": 6.0260114669799805,
      "learning_rate": 3.678787878787879e-05,
      "loss": 0.1539,
      "step": 2716
    },
    {
      "epoch": 0.2717,
      "grad_norm": 4.6880784034729,
      "learning_rate": 3.6782828282828285e-05,
      "loss": 0.1849,
      "step": 2717
    },
    {
      "epoch": 0.2718,
      "grad_norm": 3.8330957889556885,
      "learning_rate": 3.677777777777778e-05,
      "loss": 0.2032,
      "step": 2718
    },
    {
      "epoch": 0.2719,
      "grad_norm": 7.9575371742248535,
      "learning_rate": 3.677272727272728e-05,
      "loss": 0.3307,
      "step": 2719
    },
    {
      "epoch": 0.272,
      "grad_norm": 4.1370158195495605,
      "learning_rate": 3.6767676767676766e-05,
      "loss": 0.0599,
      "step": 2720
    },
    {
      "epoch": 0.2721,
      "grad_norm": 8.720464706420898,
      "learning_rate": 3.676262626262626e-05,
      "loss": 0.4719,
      "step": 2721
    },
    {
      "epoch": 0.2722,
      "grad_norm": 4.014006614685059,
      "learning_rate": 3.675757575757576e-05,
      "loss": 0.1289,
      "step": 2722
    },
    {
      "epoch": 0.2723,
      "grad_norm": 7.20210075378418,
      "learning_rate": 3.6752525252525254e-05,
      "loss": 0.1261,
      "step": 2723
    },
    {
      "epoch": 0.2724,
      "grad_norm": 5.1863203048706055,
      "learning_rate": 3.674747474747475e-05,
      "loss": 0.2676,
      "step": 2724
    },
    {
      "epoch": 0.2725,
      "grad_norm": 3.0748846530914307,
      "learning_rate": 3.6742424242424246e-05,
      "loss": 0.1148,
      "step": 2725
    },
    {
      "epoch": 0.2726,
      "grad_norm": 10.67739486694336,
      "learning_rate": 3.6737373737373735e-05,
      "loss": 0.3034,
      "step": 2726
    },
    {
      "epoch": 0.2727,
      "grad_norm": 4.8485026359558105,
      "learning_rate": 3.673232323232323e-05,
      "loss": 0.226,
      "step": 2727
    },
    {
      "epoch": 0.2728,
      "grad_norm": 5.091731548309326,
      "learning_rate": 3.672727272727273e-05,
      "loss": 0.2036,
      "step": 2728
    },
    {
      "epoch": 0.2729,
      "grad_norm": 47.320640563964844,
      "learning_rate": 3.672222222222222e-05,
      "loss": 0.2974,
      "step": 2729
    },
    {
      "epoch": 0.273,
      "grad_norm": 2.8441884517669678,
      "learning_rate": 3.671717171717172e-05,
      "loss": 0.0688,
      "step": 2730
    },
    {
      "epoch": 0.2731,
      "grad_norm": 2.55962872505188,
      "learning_rate": 3.6712121212121215e-05,
      "loss": 0.1028,
      "step": 2731
    },
    {
      "epoch": 0.2732,
      "grad_norm": 3.524423360824585,
      "learning_rate": 3.6707070707070705e-05,
      "loss": 0.2219,
      "step": 2732
    },
    {
      "epoch": 0.2733,
      "grad_norm": 8.386990547180176,
      "learning_rate": 3.670202020202021e-05,
      "loss": 0.3236,
      "step": 2733
    },
    {
      "epoch": 0.2734,
      "grad_norm": 7.080157279968262,
      "learning_rate": 3.6696969696969697e-05,
      "loss": 0.1821,
      "step": 2734
    },
    {
      "epoch": 0.2735,
      "grad_norm": 8.14339542388916,
      "learning_rate": 3.669191919191919e-05,
      "loss": 0.2161,
      "step": 2735
    },
    {
      "epoch": 0.2736,
      "grad_norm": 5.467625141143799,
      "learning_rate": 3.668686868686869e-05,
      "loss": 0.3186,
      "step": 2736
    },
    {
      "epoch": 0.2737,
      "grad_norm": 6.065621852874756,
      "learning_rate": 3.6681818181818185e-05,
      "loss": 0.1445,
      "step": 2737
    },
    {
      "epoch": 0.2738,
      "grad_norm": 20.918283462524414,
      "learning_rate": 3.6676767676767674e-05,
      "loss": 0.3601,
      "step": 2738
    },
    {
      "epoch": 0.2739,
      "grad_norm": 6.843874931335449,
      "learning_rate": 3.6671717171717176e-05,
      "loss": 0.2656,
      "step": 2739
    },
    {
      "epoch": 0.274,
      "grad_norm": 2.987595558166504,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.1406,
      "step": 2740
    },
    {
      "epoch": 0.2741,
      "grad_norm": 5.246522426605225,
      "learning_rate": 3.666161616161616e-05,
      "loss": 0.4225,
      "step": 2741
    },
    {
      "epoch": 0.2742,
      "grad_norm": 2.7914671897888184,
      "learning_rate": 3.665656565656566e-05,
      "loss": 0.199,
      "step": 2742
    },
    {
      "epoch": 0.2743,
      "grad_norm": 2.714428663253784,
      "learning_rate": 3.6651515151515154e-05,
      "loss": 0.0859,
      "step": 2743
    },
    {
      "epoch": 0.2744,
      "grad_norm": 3.3410136699676514,
      "learning_rate": 3.664646464646464e-05,
      "loss": 0.1133,
      "step": 2744
    },
    {
      "epoch": 0.2745,
      "grad_norm": 5.175949573516846,
      "learning_rate": 3.6641414141414146e-05,
      "loss": 0.337,
      "step": 2745
    },
    {
      "epoch": 0.2746,
      "grad_norm": 8.428905487060547,
      "learning_rate": 3.663636363636364e-05,
      "loss": 0.7956,
      "step": 2746
    },
    {
      "epoch": 0.2747,
      "grad_norm": 3.6135120391845703,
      "learning_rate": 3.663131313131313e-05,
      "loss": 0.2336,
      "step": 2747
    },
    {
      "epoch": 0.2748,
      "grad_norm": 3.0360162258148193,
      "learning_rate": 3.6626262626262634e-05,
      "loss": 0.1179,
      "step": 2748
    },
    {
      "epoch": 0.2749,
      "grad_norm": 3.8040690422058105,
      "learning_rate": 3.662121212121212e-05,
      "loss": 0.1096,
      "step": 2749
    },
    {
      "epoch": 0.275,
      "grad_norm": 3.0988097190856934,
      "learning_rate": 3.661616161616162e-05,
      "loss": 0.1859,
      "step": 2750
    },
    {
      "epoch": 0.2751,
      "grad_norm": 4.444920063018799,
      "learning_rate": 3.6611111111111115e-05,
      "loss": 0.1132,
      "step": 2751
    },
    {
      "epoch": 0.2752,
      "grad_norm": 9.45478630065918,
      "learning_rate": 3.660606060606061e-05,
      "loss": 0.1539,
      "step": 2752
    },
    {
      "epoch": 0.2753,
      "grad_norm": 2.5391154289245605,
      "learning_rate": 3.66010101010101e-05,
      "loss": 0.068,
      "step": 2753
    },
    {
      "epoch": 0.2754,
      "grad_norm": 8.779529571533203,
      "learning_rate": 3.65959595959596e-05,
      "loss": 0.4102,
      "step": 2754
    },
    {
      "epoch": 0.2755,
      "grad_norm": 4.626233100891113,
      "learning_rate": 3.659090909090909e-05,
      "loss": 0.2847,
      "step": 2755
    },
    {
      "epoch": 0.2756,
      "grad_norm": 3.561530351638794,
      "learning_rate": 3.658585858585859e-05,
      "loss": 0.2053,
      "step": 2756
    },
    {
      "epoch": 0.2757,
      "grad_norm": 3.547919273376465,
      "learning_rate": 3.6580808080808084e-05,
      "loss": 0.0887,
      "step": 2757
    },
    {
      "epoch": 0.2758,
      "grad_norm": 7.211075782775879,
      "learning_rate": 3.657575757575758e-05,
      "loss": 0.1467,
      "step": 2758
    },
    {
      "epoch": 0.2759,
      "grad_norm": 1.534916877746582,
      "learning_rate": 3.657070707070707e-05,
      "loss": 0.0378,
      "step": 2759
    },
    {
      "epoch": 0.276,
      "grad_norm": 4.708143711090088,
      "learning_rate": 3.656565656565657e-05,
      "loss": 0.1909,
      "step": 2760
    },
    {
      "epoch": 0.2761,
      "grad_norm": 13.150456428527832,
      "learning_rate": 3.656060606060606e-05,
      "loss": 0.1065,
      "step": 2761
    },
    {
      "epoch": 0.2762,
      "grad_norm": 9.875165939331055,
      "learning_rate": 3.655555555555556e-05,
      "loss": 0.3208,
      "step": 2762
    },
    {
      "epoch": 0.2763,
      "grad_norm": 5.298358917236328,
      "learning_rate": 3.655050505050505e-05,
      "loss": 0.2919,
      "step": 2763
    },
    {
      "epoch": 0.2764,
      "grad_norm": 11.670745849609375,
      "learning_rate": 3.654545454545455e-05,
      "loss": 0.3721,
      "step": 2764
    },
    {
      "epoch": 0.2765,
      "grad_norm": 3.2027056217193604,
      "learning_rate": 3.654040404040404e-05,
      "loss": 0.0596,
      "step": 2765
    },
    {
      "epoch": 0.2766,
      "grad_norm": 6.597387313842773,
      "learning_rate": 3.653535353535354e-05,
      "loss": 0.231,
      "step": 2766
    },
    {
      "epoch": 0.2767,
      "grad_norm": 8.52628231048584,
      "learning_rate": 3.653030303030303e-05,
      "loss": 0.2062,
      "step": 2767
    },
    {
      "epoch": 0.2768,
      "grad_norm": 2.2531020641326904,
      "learning_rate": 3.6525252525252526e-05,
      "loss": 0.0895,
      "step": 2768
    },
    {
      "epoch": 0.2769,
      "grad_norm": 5.021322727203369,
      "learning_rate": 3.652020202020202e-05,
      "loss": 0.1538,
      "step": 2769
    },
    {
      "epoch": 0.277,
      "grad_norm": 5.442748546600342,
      "learning_rate": 3.651515151515152e-05,
      "loss": 0.104,
      "step": 2770
    },
    {
      "epoch": 0.2771,
      "grad_norm": 6.182654857635498,
      "learning_rate": 3.6510101010101014e-05,
      "loss": 0.1514,
      "step": 2771
    },
    {
      "epoch": 0.2772,
      "grad_norm": 2.49722957611084,
      "learning_rate": 3.650505050505051e-05,
      "loss": 0.1614,
      "step": 2772
    },
    {
      "epoch": 0.2773,
      "grad_norm": 3.983530044555664,
      "learning_rate": 3.65e-05,
      "loss": 0.2832,
      "step": 2773
    },
    {
      "epoch": 0.2774,
      "grad_norm": 9.43697452545166,
      "learning_rate": 3.6494949494949496e-05,
      "loss": 0.4113,
      "step": 2774
    },
    {
      "epoch": 0.2775,
      "grad_norm": 8.170114517211914,
      "learning_rate": 3.648989898989899e-05,
      "loss": 0.2706,
      "step": 2775
    },
    {
      "epoch": 0.2776,
      "grad_norm": 1.4621193408966064,
      "learning_rate": 3.648484848484849e-05,
      "loss": 0.034,
      "step": 2776
    },
    {
      "epoch": 0.2777,
      "grad_norm": 4.752368927001953,
      "learning_rate": 3.6479797979797983e-05,
      "loss": 0.1316,
      "step": 2777
    },
    {
      "epoch": 0.2778,
      "grad_norm": 5.432509422302246,
      "learning_rate": 3.647474747474748e-05,
      "loss": 0.3012,
      "step": 2778
    },
    {
      "epoch": 0.2779,
      "grad_norm": 3.8270668983459473,
      "learning_rate": 3.646969696969697e-05,
      "loss": 0.2836,
      "step": 2779
    },
    {
      "epoch": 0.278,
      "grad_norm": 2.699828863143921,
      "learning_rate": 3.6464646464646465e-05,
      "loss": 0.0854,
      "step": 2780
    },
    {
      "epoch": 0.2781,
      "grad_norm": 5.483888149261475,
      "learning_rate": 3.645959595959596e-05,
      "loss": 0.4212,
      "step": 2781
    },
    {
      "epoch": 0.2782,
      "grad_norm": 2.5818824768066406,
      "learning_rate": 3.645454545454546e-05,
      "loss": 0.0964,
      "step": 2782
    },
    {
      "epoch": 0.2783,
      "grad_norm": 6.472258567810059,
      "learning_rate": 3.644949494949495e-05,
      "loss": 0.3745,
      "step": 2783
    },
    {
      "epoch": 0.2784,
      "grad_norm": 4.311099529266357,
      "learning_rate": 3.644444444444445e-05,
      "loss": 0.2369,
      "step": 2784
    },
    {
      "epoch": 0.2785,
      "grad_norm": 3.1781678199768066,
      "learning_rate": 3.643939393939394e-05,
      "loss": 0.1108,
      "step": 2785
    },
    {
      "epoch": 0.2786,
      "grad_norm": 8.411627769470215,
      "learning_rate": 3.643434343434344e-05,
      "loss": 0.5554,
      "step": 2786
    },
    {
      "epoch": 0.2787,
      "grad_norm": 6.281577110290527,
      "learning_rate": 3.642929292929293e-05,
      "loss": 0.1361,
      "step": 2787
    },
    {
      "epoch": 0.2788,
      "grad_norm": 5.234624862670898,
      "learning_rate": 3.6424242424242426e-05,
      "loss": 0.2627,
      "step": 2788
    },
    {
      "epoch": 0.2789,
      "grad_norm": 1.8878819942474365,
      "learning_rate": 3.641919191919192e-05,
      "loss": 0.0387,
      "step": 2789
    },
    {
      "epoch": 0.279,
      "grad_norm": 9.071844100952148,
      "learning_rate": 3.641414141414142e-05,
      "loss": 0.373,
      "step": 2790
    },
    {
      "epoch": 0.2791,
      "grad_norm": 4.354456901550293,
      "learning_rate": 3.640909090909091e-05,
      "loss": 0.2222,
      "step": 2791
    },
    {
      "epoch": 0.2792,
      "grad_norm": 7.214076995849609,
      "learning_rate": 3.640404040404041e-05,
      "loss": 0.349,
      "step": 2792
    },
    {
      "epoch": 0.2793,
      "grad_norm": 3.4228625297546387,
      "learning_rate": 3.63989898989899e-05,
      "loss": 0.1334,
      "step": 2793
    },
    {
      "epoch": 0.2794,
      "grad_norm": 4.05573844909668,
      "learning_rate": 3.6393939393939395e-05,
      "loss": 0.2664,
      "step": 2794
    },
    {
      "epoch": 0.2795,
      "grad_norm": 6.419083595275879,
      "learning_rate": 3.638888888888889e-05,
      "loss": 0.2733,
      "step": 2795
    },
    {
      "epoch": 0.2796,
      "grad_norm": 2.1894149780273438,
      "learning_rate": 3.638383838383839e-05,
      "loss": 0.1012,
      "step": 2796
    },
    {
      "epoch": 0.2797,
      "grad_norm": 3.2633187770843506,
      "learning_rate": 3.6378787878787876e-05,
      "loss": 0.1569,
      "step": 2797
    },
    {
      "epoch": 0.2798,
      "grad_norm": 5.96645975112915,
      "learning_rate": 3.637373737373738e-05,
      "loss": 0.252,
      "step": 2798
    },
    {
      "epoch": 0.2799,
      "grad_norm": 7.876717567443848,
      "learning_rate": 3.636868686868687e-05,
      "loss": 0.4246,
      "step": 2799
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.644723892211914,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.2522,
      "step": 2800
    },
    {
      "epoch": 0.2801,
      "grad_norm": 4.443268299102783,
      "learning_rate": 3.635858585858586e-05,
      "loss": 0.0968,
      "step": 2801
    },
    {
      "epoch": 0.2802,
      "grad_norm": 3.4851369857788086,
      "learning_rate": 3.6353535353535356e-05,
      "loss": 0.2275,
      "step": 2802
    },
    {
      "epoch": 0.2803,
      "grad_norm": 2.664074420928955,
      "learning_rate": 3.6348484848484845e-05,
      "loss": 0.1239,
      "step": 2803
    },
    {
      "epoch": 0.2804,
      "grad_norm": 4.564184188842773,
      "learning_rate": 3.634343434343435e-05,
      "loss": 0.3318,
      "step": 2804
    },
    {
      "epoch": 0.2805,
      "grad_norm": 5.896300792694092,
      "learning_rate": 3.633838383838384e-05,
      "loss": 0.2147,
      "step": 2805
    },
    {
      "epoch": 0.2806,
      "grad_norm": 2.9874603748321533,
      "learning_rate": 3.633333333333333e-05,
      "loss": 0.0616,
      "step": 2806
    },
    {
      "epoch": 0.2807,
      "grad_norm": 7.987407207489014,
      "learning_rate": 3.632828282828283e-05,
      "loss": 0.339,
      "step": 2807
    },
    {
      "epoch": 0.2808,
      "grad_norm": 14.153170585632324,
      "learning_rate": 3.6323232323232325e-05,
      "loss": 0.522,
      "step": 2808
    },
    {
      "epoch": 0.2809,
      "grad_norm": 1.7062127590179443,
      "learning_rate": 3.6318181818181815e-05,
      "loss": 0.0383,
      "step": 2809
    },
    {
      "epoch": 0.281,
      "grad_norm": 3.375035524368286,
      "learning_rate": 3.631313131313132e-05,
      "loss": 0.1517,
      "step": 2810
    },
    {
      "epoch": 0.2811,
      "grad_norm": 5.104559898376465,
      "learning_rate": 3.6308080808080806e-05,
      "loss": 0.1245,
      "step": 2811
    },
    {
      "epoch": 0.2812,
      "grad_norm": 3.268700361251831,
      "learning_rate": 3.63030303030303e-05,
      "loss": 0.1881,
      "step": 2812
    },
    {
      "epoch": 0.2813,
      "grad_norm": 7.1892266273498535,
      "learning_rate": 3.62979797979798e-05,
      "loss": 0.249,
      "step": 2813
    },
    {
      "epoch": 0.2814,
      "grad_norm": 17.740577697753906,
      "learning_rate": 3.6292929292929294e-05,
      "loss": 0.3993,
      "step": 2814
    },
    {
      "epoch": 0.2815,
      "grad_norm": 12.161538124084473,
      "learning_rate": 3.628787878787879e-05,
      "loss": 0.4015,
      "step": 2815
    },
    {
      "epoch": 0.2816,
      "grad_norm": 3.6203296184539795,
      "learning_rate": 3.6282828282828286e-05,
      "loss": 0.1294,
      "step": 2816
    },
    {
      "epoch": 0.2817,
      "grad_norm": 4.860653877258301,
      "learning_rate": 3.6277777777777776e-05,
      "loss": 0.2209,
      "step": 2817
    },
    {
      "epoch": 0.2818,
      "grad_norm": 4.714880466461182,
      "learning_rate": 3.627272727272727e-05,
      "loss": 0.1208,
      "step": 2818
    },
    {
      "epoch": 0.2819,
      "grad_norm": 2.694425106048584,
      "learning_rate": 3.626767676767677e-05,
      "loss": 0.1013,
      "step": 2819
    },
    {
      "epoch": 0.282,
      "grad_norm": 1.7608321905136108,
      "learning_rate": 3.6262626262626264e-05,
      "loss": 0.0493,
      "step": 2820
    },
    {
      "epoch": 0.2821,
      "grad_norm": 4.527198791503906,
      "learning_rate": 3.625757575757576e-05,
      "loss": 0.1844,
      "step": 2821
    },
    {
      "epoch": 0.2822,
      "grad_norm": 6.424966812133789,
      "learning_rate": 3.6252525252525256e-05,
      "loss": 0.314,
      "step": 2822
    },
    {
      "epoch": 0.2823,
      "grad_norm": 8.853665351867676,
      "learning_rate": 3.6247474747474745e-05,
      "loss": 0.2802,
      "step": 2823
    },
    {
      "epoch": 0.2824,
      "grad_norm": 2.4401707649230957,
      "learning_rate": 3.624242424242425e-05,
      "loss": 0.0376,
      "step": 2824
    },
    {
      "epoch": 0.2825,
      "grad_norm": 7.713212490081787,
      "learning_rate": 3.623737373737374e-05,
      "loss": 0.2533,
      "step": 2825
    },
    {
      "epoch": 0.2826,
      "grad_norm": 5.539959907531738,
      "learning_rate": 3.623232323232323e-05,
      "loss": 0.4816,
      "step": 2826
    },
    {
      "epoch": 0.2827,
      "grad_norm": 1.5715097188949585,
      "learning_rate": 3.622727272727273e-05,
      "loss": 0.0475,
      "step": 2827
    },
    {
      "epoch": 0.2828,
      "grad_norm": 8.640216827392578,
      "learning_rate": 3.6222222222222225e-05,
      "loss": 0.1475,
      "step": 2828
    },
    {
      "epoch": 0.2829,
      "grad_norm": 8.835668563842773,
      "learning_rate": 3.6217171717171714e-05,
      "loss": 0.4131,
      "step": 2829
    },
    {
      "epoch": 0.283,
      "grad_norm": 2.6396405696868896,
      "learning_rate": 3.621212121212122e-05,
      "loss": 0.2801,
      "step": 2830
    },
    {
      "epoch": 0.2831,
      "grad_norm": 4.339224338531494,
      "learning_rate": 3.6207070707070706e-05,
      "loss": 0.1365,
      "step": 2831
    },
    {
      "epoch": 0.2832,
      "grad_norm": 6.7571258544921875,
      "learning_rate": 3.62020202020202e-05,
      "loss": 0.2764,
      "step": 2832
    },
    {
      "epoch": 0.2833,
      "grad_norm": 5.121927738189697,
      "learning_rate": 3.6196969696969705e-05,
      "loss": 0.229,
      "step": 2833
    },
    {
      "epoch": 0.2834,
      "grad_norm": 7.551070690155029,
      "learning_rate": 3.6191919191919194e-05,
      "loss": 0.453,
      "step": 2834
    },
    {
      "epoch": 0.2835,
      "grad_norm": 3.2799010276794434,
      "learning_rate": 3.618686868686869e-05,
      "loss": 0.1435,
      "step": 2835
    },
    {
      "epoch": 0.2836,
      "grad_norm": 18.209197998046875,
      "learning_rate": 3.6181818181818186e-05,
      "loss": 0.1291,
      "step": 2836
    },
    {
      "epoch": 0.2837,
      "grad_norm": 6.371016979217529,
      "learning_rate": 3.617676767676768e-05,
      "loss": 0.1886,
      "step": 2837
    },
    {
      "epoch": 0.2838,
      "grad_norm": 9.331805229187012,
      "learning_rate": 3.617171717171717e-05,
      "loss": 0.2651,
      "step": 2838
    },
    {
      "epoch": 0.2839,
      "grad_norm": 21.582754135131836,
      "learning_rate": 3.6166666666666674e-05,
      "loss": 0.5847,
      "step": 2839
    },
    {
      "epoch": 0.284,
      "grad_norm": 7.174103260040283,
      "learning_rate": 3.616161616161616e-05,
      "loss": 0.346,
      "step": 2840
    },
    {
      "epoch": 0.2841,
      "grad_norm": 9.75558090209961,
      "learning_rate": 3.615656565656566e-05,
      "loss": 0.2755,
      "step": 2841
    },
    {
      "epoch": 0.2842,
      "grad_norm": 13.086472511291504,
      "learning_rate": 3.6151515151515155e-05,
      "loss": 0.3855,
      "step": 2842
    },
    {
      "epoch": 0.2843,
      "grad_norm": 9.72380256652832,
      "learning_rate": 3.614646464646465e-05,
      "loss": 0.4689,
      "step": 2843
    },
    {
      "epoch": 0.2844,
      "grad_norm": 13.664664268493652,
      "learning_rate": 3.614141414141414e-05,
      "loss": 0.314,
      "step": 2844
    },
    {
      "epoch": 0.2845,
      "grad_norm": 2.5307445526123047,
      "learning_rate": 3.613636363636364e-05,
      "loss": 0.1119,
      "step": 2845
    },
    {
      "epoch": 0.2846,
      "grad_norm": 6.491151332855225,
      "learning_rate": 3.613131313131313e-05,
      "loss": 0.2987,
      "step": 2846
    },
    {
      "epoch": 0.2847,
      "grad_norm": 22.319679260253906,
      "learning_rate": 3.612626262626263e-05,
      "loss": 0.296,
      "step": 2847
    },
    {
      "epoch": 0.2848,
      "grad_norm": 9.574606895446777,
      "learning_rate": 3.6121212121212124e-05,
      "loss": 0.5007,
      "step": 2848
    },
    {
      "epoch": 0.2849,
      "grad_norm": 3.5193159580230713,
      "learning_rate": 3.611616161616162e-05,
      "loss": 0.1412,
      "step": 2849
    },
    {
      "epoch": 0.285,
      "grad_norm": 3.1531929969787598,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.0731,
      "step": 2850
    },
    {
      "epoch": 0.2851,
      "grad_norm": 6.34915828704834,
      "learning_rate": 3.610606060606061e-05,
      "loss": 0.2388,
      "step": 2851
    },
    {
      "epoch": 0.2852,
      "grad_norm": 5.553352355957031,
      "learning_rate": 3.61010101010101e-05,
      "loss": 0.2977,
      "step": 2852
    },
    {
      "epoch": 0.2853,
      "grad_norm": 6.095126628875732,
      "learning_rate": 3.60959595959596e-05,
      "loss": 0.2572,
      "step": 2853
    },
    {
      "epoch": 0.2854,
      "grad_norm": 5.254753112792969,
      "learning_rate": 3.6090909090909093e-05,
      "loss": 0.2766,
      "step": 2854
    },
    {
      "epoch": 0.2855,
      "grad_norm": 3.9254090785980225,
      "learning_rate": 3.608585858585859e-05,
      "loss": 0.2382,
      "step": 2855
    },
    {
      "epoch": 0.2856,
      "grad_norm": 7.1781535148620605,
      "learning_rate": 3.608080808080808e-05,
      "loss": 0.2297,
      "step": 2856
    },
    {
      "epoch": 0.2857,
      "grad_norm": 8.532951354980469,
      "learning_rate": 3.607575757575758e-05,
      "loss": 0.3649,
      "step": 2857
    },
    {
      "epoch": 0.2858,
      "grad_norm": 8.744199752807617,
      "learning_rate": 3.607070707070707e-05,
      "loss": 0.2441,
      "step": 2858
    },
    {
      "epoch": 0.2859,
      "grad_norm": 4.5447163581848145,
      "learning_rate": 3.6065656565656567e-05,
      "loss": 0.1597,
      "step": 2859
    },
    {
      "epoch": 0.286,
      "grad_norm": 5.915641784667969,
      "learning_rate": 3.606060606060606e-05,
      "loss": 0.4122,
      "step": 2860
    },
    {
      "epoch": 0.2861,
      "grad_norm": 3.2519009113311768,
      "learning_rate": 3.605555555555556e-05,
      "loss": 0.1101,
      "step": 2861
    },
    {
      "epoch": 0.2862,
      "grad_norm": 4.105710506439209,
      "learning_rate": 3.6050505050505055e-05,
      "loss": 0.0933,
      "step": 2862
    },
    {
      "epoch": 0.2863,
      "grad_norm": 5.70635461807251,
      "learning_rate": 3.604545454545455e-05,
      "loss": 0.0996,
      "step": 2863
    },
    {
      "epoch": 0.2864,
      "grad_norm": 5.487968444824219,
      "learning_rate": 3.604040404040404e-05,
      "loss": 0.2789,
      "step": 2864
    },
    {
      "epoch": 0.2865,
      "grad_norm": 17.727506637573242,
      "learning_rate": 3.6035353535353536e-05,
      "loss": 0.1892,
      "step": 2865
    },
    {
      "epoch": 0.2866,
      "grad_norm": 5.726678371429443,
      "learning_rate": 3.603030303030303e-05,
      "loss": 0.3118,
      "step": 2866
    },
    {
      "epoch": 0.2867,
      "grad_norm": 3.2467610836029053,
      "learning_rate": 3.602525252525253e-05,
      "loss": 0.1023,
      "step": 2867
    },
    {
      "epoch": 0.2868,
      "grad_norm": 14.315835952758789,
      "learning_rate": 3.6020202020202024e-05,
      "loss": 0.2812,
      "step": 2868
    },
    {
      "epoch": 0.2869,
      "grad_norm": 1.9476101398468018,
      "learning_rate": 3.601515151515152e-05,
      "loss": 0.0611,
      "step": 2869
    },
    {
      "epoch": 0.287,
      "grad_norm": 113.3505630493164,
      "learning_rate": 3.601010101010101e-05,
      "loss": 0.1503,
      "step": 2870
    },
    {
      "epoch": 0.2871,
      "grad_norm": 4.474056243896484,
      "learning_rate": 3.6005050505050505e-05,
      "loss": 0.1102,
      "step": 2871
    },
    {
      "epoch": 0.2872,
      "grad_norm": 8.77238941192627,
      "learning_rate": 3.6e-05,
      "loss": 0.2094,
      "step": 2872
    },
    {
      "epoch": 0.2873,
      "grad_norm": 13.2750883102417,
      "learning_rate": 3.59949494949495e-05,
      "loss": 0.7397,
      "step": 2873
    },
    {
      "epoch": 0.2874,
      "grad_norm": 13.909192085266113,
      "learning_rate": 3.598989898989899e-05,
      "loss": 0.2786,
      "step": 2874
    },
    {
      "epoch": 0.2875,
      "grad_norm": 86.79112243652344,
      "learning_rate": 3.598484848484849e-05,
      "loss": 0.1776,
      "step": 2875
    },
    {
      "epoch": 0.2876,
      "grad_norm": 3.8309268951416016,
      "learning_rate": 3.597979797979798e-05,
      "loss": 0.1472,
      "step": 2876
    },
    {
      "epoch": 0.2877,
      "grad_norm": 2.0050745010375977,
      "learning_rate": 3.597474747474748e-05,
      "loss": 0.0514,
      "step": 2877
    },
    {
      "epoch": 0.2878,
      "grad_norm": 11.831098556518555,
      "learning_rate": 3.596969696969697e-05,
      "loss": 0.3507,
      "step": 2878
    },
    {
      "epoch": 0.2879,
      "grad_norm": 8.085896492004395,
      "learning_rate": 3.5964646464646466e-05,
      "loss": 0.2158,
      "step": 2879
    },
    {
      "epoch": 0.288,
      "grad_norm": 10.20850658416748,
      "learning_rate": 3.595959595959596e-05,
      "loss": 0.1979,
      "step": 2880
    },
    {
      "epoch": 0.2881,
      "grad_norm": 4.080511093139648,
      "learning_rate": 3.595454545454546e-05,
      "loss": 0.2304,
      "step": 2881
    },
    {
      "epoch": 0.2882,
      "grad_norm": 9.234663963317871,
      "learning_rate": 3.594949494949495e-05,
      "loss": 0.3462,
      "step": 2882
    },
    {
      "epoch": 0.2883,
      "grad_norm": 4.143762111663818,
      "learning_rate": 3.594444444444445e-05,
      "loss": 0.132,
      "step": 2883
    },
    {
      "epoch": 0.2884,
      "grad_norm": 25.886266708374023,
      "learning_rate": 3.593939393939394e-05,
      "loss": 0.5915,
      "step": 2884
    },
    {
      "epoch": 0.2885,
      "grad_norm": 6.432650089263916,
      "learning_rate": 3.5934343434343435e-05,
      "loss": 0.1781,
      "step": 2885
    },
    {
      "epoch": 0.2886,
      "grad_norm": 2.721238136291504,
      "learning_rate": 3.592929292929293e-05,
      "loss": 0.0795,
      "step": 2886
    },
    {
      "epoch": 0.2887,
      "grad_norm": 14.722451210021973,
      "learning_rate": 3.592424242424243e-05,
      "loss": 0.475,
      "step": 2887
    },
    {
      "epoch": 0.2888,
      "grad_norm": 8.040976524353027,
      "learning_rate": 3.5919191919191916e-05,
      "loss": 0.3912,
      "step": 2888
    },
    {
      "epoch": 0.2889,
      "grad_norm": 2.8723084926605225,
      "learning_rate": 3.591414141414142e-05,
      "loss": 0.0815,
      "step": 2889
    },
    {
      "epoch": 0.289,
      "grad_norm": 5.609866619110107,
      "learning_rate": 3.590909090909091e-05,
      "loss": 0.1102,
      "step": 2890
    },
    {
      "epoch": 0.2891,
      "grad_norm": 2.0935781002044678,
      "learning_rate": 3.5904040404040404e-05,
      "loss": 0.0768,
      "step": 2891
    },
    {
      "epoch": 0.2892,
      "grad_norm": 5.019315242767334,
      "learning_rate": 3.58989898989899e-05,
      "loss": 0.0853,
      "step": 2892
    },
    {
      "epoch": 0.2893,
      "grad_norm": 6.17024564743042,
      "learning_rate": 3.5893939393939396e-05,
      "loss": 0.2562,
      "step": 2893
    },
    {
      "epoch": 0.2894,
      "grad_norm": 2.5924198627471924,
      "learning_rate": 3.5888888888888886e-05,
      "loss": 0.0841,
      "step": 2894
    },
    {
      "epoch": 0.2895,
      "grad_norm": 2.20220947265625,
      "learning_rate": 3.588383838383839e-05,
      "loss": 0.0662,
      "step": 2895
    },
    {
      "epoch": 0.2896,
      "grad_norm": 18.189489364624023,
      "learning_rate": 3.587878787878788e-05,
      "loss": 0.8917,
      "step": 2896
    },
    {
      "epoch": 0.2897,
      "grad_norm": 6.725922584533691,
      "learning_rate": 3.5873737373737374e-05,
      "loss": 0.3673,
      "step": 2897
    },
    {
      "epoch": 0.2898,
      "grad_norm": 11.6707763671875,
      "learning_rate": 3.586868686868687e-05,
      "loss": 0.6808,
      "step": 2898
    },
    {
      "epoch": 0.2899,
      "grad_norm": 82.02326202392578,
      "learning_rate": 3.5863636363636366e-05,
      "loss": 0.478,
      "step": 2899
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.898562908172607,
      "learning_rate": 3.5858585858585855e-05,
      "loss": 0.1737,
      "step": 2900
    },
    {
      "epoch": 0.2901,
      "grad_norm": 3.235586643218994,
      "learning_rate": 3.585353535353536e-05,
      "loss": 0.0703,
      "step": 2901
    },
    {
      "epoch": 0.2902,
      "grad_norm": 4.866949558258057,
      "learning_rate": 3.584848484848485e-05,
      "loss": 0.1365,
      "step": 2902
    },
    {
      "epoch": 0.2903,
      "grad_norm": 6.758153438568115,
      "learning_rate": 3.584343434343434e-05,
      "loss": 0.3416,
      "step": 2903
    },
    {
      "epoch": 0.2904,
      "grad_norm": 13.77380657196045,
      "learning_rate": 3.583838383838384e-05,
      "loss": 0.0876,
      "step": 2904
    },
    {
      "epoch": 0.2905,
      "grad_norm": 3.9092023372650146,
      "learning_rate": 3.5833333333333335e-05,
      "loss": 0.1882,
      "step": 2905
    },
    {
      "epoch": 0.2906,
      "grad_norm": 9.776119232177734,
      "learning_rate": 3.582828282828283e-05,
      "loss": 0.2175,
      "step": 2906
    },
    {
      "epoch": 0.2907,
      "grad_norm": 3.662838935852051,
      "learning_rate": 3.582323232323233e-05,
      "loss": 0.1807,
      "step": 2907
    },
    {
      "epoch": 0.2908,
      "grad_norm": 3.4327540397644043,
      "learning_rate": 3.5818181818181816e-05,
      "loss": 0.137,
      "step": 2908
    },
    {
      "epoch": 0.2909,
      "grad_norm": 4.252387046813965,
      "learning_rate": 3.581313131313131e-05,
      "loss": 0.294,
      "step": 2909
    },
    {
      "epoch": 0.291,
      "grad_norm": 8.210935592651367,
      "learning_rate": 3.580808080808081e-05,
      "loss": 0.4069,
      "step": 2910
    },
    {
      "epoch": 0.2911,
      "grad_norm": 6.389494895935059,
      "learning_rate": 3.5803030303030304e-05,
      "loss": 0.1664,
      "step": 2911
    },
    {
      "epoch": 0.2912,
      "grad_norm": 9.605997085571289,
      "learning_rate": 3.57979797979798e-05,
      "loss": 0.3134,
      "step": 2912
    },
    {
      "epoch": 0.2913,
      "grad_norm": 7.919416904449463,
      "learning_rate": 3.5792929292929296e-05,
      "loss": 0.237,
      "step": 2913
    },
    {
      "epoch": 0.2914,
      "grad_norm": 8.580288887023926,
      "learning_rate": 3.5787878787878785e-05,
      "loss": 0.3141,
      "step": 2914
    },
    {
      "epoch": 0.2915,
      "grad_norm": 4.480668067932129,
      "learning_rate": 3.578282828282829e-05,
      "loss": 0.2407,
      "step": 2915
    },
    {
      "epoch": 0.2916,
      "grad_norm": 10.687233924865723,
      "learning_rate": 3.577777777777778e-05,
      "loss": 0.4018,
      "step": 2916
    },
    {
      "epoch": 0.2917,
      "grad_norm": 5.5972065925598145,
      "learning_rate": 3.577272727272727e-05,
      "loss": 0.2644,
      "step": 2917
    },
    {
      "epoch": 0.2918,
      "grad_norm": 9.610940933227539,
      "learning_rate": 3.576767676767677e-05,
      "loss": 0.3024,
      "step": 2918
    },
    {
      "epoch": 0.2919,
      "grad_norm": 1.961254596710205,
      "learning_rate": 3.5762626262626265e-05,
      "loss": 0.0413,
      "step": 2919
    },
    {
      "epoch": 0.292,
      "grad_norm": 3.2429757118225098,
      "learning_rate": 3.575757575757576e-05,
      "loss": 0.0866,
      "step": 2920
    },
    {
      "epoch": 0.2921,
      "grad_norm": 8.123971939086914,
      "learning_rate": 3.575252525252526e-05,
      "loss": 0.6057,
      "step": 2921
    },
    {
      "epoch": 0.2922,
      "grad_norm": 5.701267719268799,
      "learning_rate": 3.574747474747475e-05,
      "loss": 0.2168,
      "step": 2922
    },
    {
      "epoch": 0.2923,
      "grad_norm": 5.903127193450928,
      "learning_rate": 3.574242424242424e-05,
      "loss": 0.2595,
      "step": 2923
    },
    {
      "epoch": 0.2924,
      "grad_norm": 16.755722045898438,
      "learning_rate": 3.5737373737373745e-05,
      "loss": 0.2892,
      "step": 2924
    },
    {
      "epoch": 0.2925,
      "grad_norm": 4.172173976898193,
      "learning_rate": 3.5732323232323234e-05,
      "loss": 0.1599,
      "step": 2925
    },
    {
      "epoch": 0.2926,
      "grad_norm": 8.117311477661133,
      "learning_rate": 3.572727272727273e-05,
      "loss": 0.211,
      "step": 2926
    },
    {
      "epoch": 0.2927,
      "grad_norm": 6.599339008331299,
      "learning_rate": 3.5722222222222226e-05,
      "loss": 0.3577,
      "step": 2927
    },
    {
      "epoch": 0.2928,
      "grad_norm": 6.557584762573242,
      "learning_rate": 3.571717171717172e-05,
      "loss": 0.1597,
      "step": 2928
    },
    {
      "epoch": 0.2929,
      "grad_norm": 5.491583824157715,
      "learning_rate": 3.571212121212121e-05,
      "loss": 0.1175,
      "step": 2929
    },
    {
      "epoch": 0.293,
      "grad_norm": 4.8435211181640625,
      "learning_rate": 3.5707070707070714e-05,
      "loss": 0.0528,
      "step": 2930
    },
    {
      "epoch": 0.2931,
      "grad_norm": 6.789720058441162,
      "learning_rate": 3.57020202020202e-05,
      "loss": 0.331,
      "step": 2931
    },
    {
      "epoch": 0.2932,
      "grad_norm": 3.2500879764556885,
      "learning_rate": 3.56969696969697e-05,
      "loss": 0.0877,
      "step": 2932
    },
    {
      "epoch": 0.2933,
      "grad_norm": 2.3641889095306396,
      "learning_rate": 3.5691919191919195e-05,
      "loss": 0.0472,
      "step": 2933
    },
    {
      "epoch": 0.2934,
      "grad_norm": 4.001527309417725,
      "learning_rate": 3.568686868686869e-05,
      "loss": 0.1066,
      "step": 2934
    },
    {
      "epoch": 0.2935,
      "grad_norm": 8.062759399414062,
      "learning_rate": 3.568181818181818e-05,
      "loss": 0.3039,
      "step": 2935
    },
    {
      "epoch": 0.2936,
      "grad_norm": 2.7392563819885254,
      "learning_rate": 3.567676767676768e-05,
      "loss": 0.1077,
      "step": 2936
    },
    {
      "epoch": 0.2937,
      "grad_norm": 4.542890548706055,
      "learning_rate": 3.567171717171717e-05,
      "loss": 0.1995,
      "step": 2937
    },
    {
      "epoch": 0.2938,
      "grad_norm": 4.781312465667725,
      "learning_rate": 3.566666666666667e-05,
      "loss": 0.1603,
      "step": 2938
    },
    {
      "epoch": 0.2939,
      "grad_norm": 21.559219360351562,
      "learning_rate": 3.5661616161616165e-05,
      "loss": 0.6527,
      "step": 2939
    },
    {
      "epoch": 0.294,
      "grad_norm": 3.3237550258636475,
      "learning_rate": 3.565656565656566e-05,
      "loss": 0.0784,
      "step": 2940
    },
    {
      "epoch": 0.2941,
      "grad_norm": 6.1138691902160645,
      "learning_rate": 3.565151515151515e-05,
      "loss": 0.1831,
      "step": 2941
    },
    {
      "epoch": 0.2942,
      "grad_norm": 4.324044704437256,
      "learning_rate": 3.564646464646465e-05,
      "loss": 0.1537,
      "step": 2942
    },
    {
      "epoch": 0.2943,
      "grad_norm": 15.74575424194336,
      "learning_rate": 3.564141414141414e-05,
      "loss": 0.4167,
      "step": 2943
    },
    {
      "epoch": 0.2944,
      "grad_norm": 4.497015953063965,
      "learning_rate": 3.563636363636364e-05,
      "loss": 0.0824,
      "step": 2944
    },
    {
      "epoch": 0.2945,
      "grad_norm": 3.1983940601348877,
      "learning_rate": 3.5631313131313134e-05,
      "loss": 0.063,
      "step": 2945
    },
    {
      "epoch": 0.2946,
      "grad_norm": 5.0391621589660645,
      "learning_rate": 3.562626262626263e-05,
      "loss": 0.3303,
      "step": 2946
    },
    {
      "epoch": 0.2947,
      "grad_norm": 13.85533332824707,
      "learning_rate": 3.562121212121212e-05,
      "loss": 0.6431,
      "step": 2947
    },
    {
      "epoch": 0.2948,
      "grad_norm": 4.49905252456665,
      "learning_rate": 3.561616161616162e-05,
      "loss": 0.2192,
      "step": 2948
    },
    {
      "epoch": 0.2949,
      "grad_norm": 6.767737865447998,
      "learning_rate": 3.561111111111111e-05,
      "loss": 0.3914,
      "step": 2949
    },
    {
      "epoch": 0.295,
      "grad_norm": 7.522176742553711,
      "learning_rate": 3.560606060606061e-05,
      "loss": 0.4876,
      "step": 2950
    },
    {
      "epoch": 0.2951,
      "grad_norm": 13.55602741241455,
      "learning_rate": 3.56010101010101e-05,
      "loss": 0.653,
      "step": 2951
    },
    {
      "epoch": 0.2952,
      "grad_norm": 4.97578239440918,
      "learning_rate": 3.55959595959596e-05,
      "loss": 0.1672,
      "step": 2952
    },
    {
      "epoch": 0.2953,
      "grad_norm": 8.303533554077148,
      "learning_rate": 3.559090909090909e-05,
      "loss": 0.8104,
      "step": 2953
    },
    {
      "epoch": 0.2954,
      "grad_norm": 3.7338955402374268,
      "learning_rate": 3.558585858585859e-05,
      "loss": 0.1548,
      "step": 2954
    },
    {
      "epoch": 0.2955,
      "grad_norm": 8.953221321105957,
      "learning_rate": 3.558080808080808e-05,
      "loss": 0.3121,
      "step": 2955
    },
    {
      "epoch": 0.2956,
      "grad_norm": 52.18149185180664,
      "learning_rate": 3.5575757575757576e-05,
      "loss": 0.188,
      "step": 2956
    },
    {
      "epoch": 0.2957,
      "grad_norm": 3.7178497314453125,
      "learning_rate": 3.557070707070707e-05,
      "loss": 0.182,
      "step": 2957
    },
    {
      "epoch": 0.2958,
      "grad_norm": 9.932621955871582,
      "learning_rate": 3.556565656565657e-05,
      "loss": 0.2903,
      "step": 2958
    },
    {
      "epoch": 0.2959,
      "grad_norm": 8.990171432495117,
      "learning_rate": 3.5560606060606064e-05,
      "loss": 0.1518,
      "step": 2959
    },
    {
      "epoch": 0.296,
      "grad_norm": 9.744991302490234,
      "learning_rate": 3.555555555555556e-05,
      "loss": 0.0904,
      "step": 2960
    },
    {
      "epoch": 0.2961,
      "grad_norm": 12.85103702545166,
      "learning_rate": 3.555050505050505e-05,
      "loss": 0.623,
      "step": 2961
    },
    {
      "epoch": 0.2962,
      "grad_norm": 6.8779730796813965,
      "learning_rate": 3.5545454545454545e-05,
      "loss": 0.2796,
      "step": 2962
    },
    {
      "epoch": 0.2963,
      "grad_norm": 3.646409749984741,
      "learning_rate": 3.554040404040404e-05,
      "loss": 0.0892,
      "step": 2963
    },
    {
      "epoch": 0.2964,
      "grad_norm": 7.279552936553955,
      "learning_rate": 3.553535353535354e-05,
      "loss": 0.3629,
      "step": 2964
    },
    {
      "epoch": 0.2965,
      "grad_norm": 9.02840518951416,
      "learning_rate": 3.553030303030303e-05,
      "loss": 0.432,
      "step": 2965
    },
    {
      "epoch": 0.2966,
      "grad_norm": 7.386312961578369,
      "learning_rate": 3.552525252525253e-05,
      "loss": 0.602,
      "step": 2966
    },
    {
      "epoch": 0.2967,
      "grad_norm": 3.779273748397827,
      "learning_rate": 3.552020202020202e-05,
      "loss": 0.1542,
      "step": 2967
    },
    {
      "epoch": 0.2968,
      "grad_norm": 6.409442901611328,
      "learning_rate": 3.551515151515152e-05,
      "loss": 0.1344,
      "step": 2968
    },
    {
      "epoch": 0.2969,
      "grad_norm": 12.246888160705566,
      "learning_rate": 3.551010101010101e-05,
      "loss": 0.2928,
      "step": 2969
    },
    {
      "epoch": 0.297,
      "grad_norm": 5.294589519500732,
      "learning_rate": 3.5505050505050506e-05,
      "loss": 0.1745,
      "step": 2970
    },
    {
      "epoch": 0.2971,
      "grad_norm": 11.082958221435547,
      "learning_rate": 3.55e-05,
      "loss": 0.7241,
      "step": 2971
    },
    {
      "epoch": 0.2972,
      "grad_norm": 5.186214923858643,
      "learning_rate": 3.54949494949495e-05,
      "loss": 0.1419,
      "step": 2972
    },
    {
      "epoch": 0.2973,
      "grad_norm": 6.701607704162598,
      "learning_rate": 3.548989898989899e-05,
      "loss": 0.2474,
      "step": 2973
    },
    {
      "epoch": 0.2974,
      "grad_norm": 30.89055061340332,
      "learning_rate": 3.548484848484849e-05,
      "loss": 0.7544,
      "step": 2974
    },
    {
      "epoch": 0.2975,
      "grad_norm": 8.779099464416504,
      "learning_rate": 3.547979797979798e-05,
      "loss": 0.1576,
      "step": 2975
    },
    {
      "epoch": 0.2976,
      "grad_norm": 7.609508514404297,
      "learning_rate": 3.5474747474747475e-05,
      "loss": 0.3408,
      "step": 2976
    },
    {
      "epoch": 0.2977,
      "grad_norm": 1.8091555833816528,
      "learning_rate": 3.546969696969697e-05,
      "loss": 0.0372,
      "step": 2977
    },
    {
      "epoch": 0.2978,
      "grad_norm": 25.62150764465332,
      "learning_rate": 3.546464646464647e-05,
      "loss": 0.5086,
      "step": 2978
    },
    {
      "epoch": 0.2979,
      "grad_norm": 10.396839141845703,
      "learning_rate": 3.545959595959596e-05,
      "loss": 0.2269,
      "step": 2979
    },
    {
      "epoch": 0.298,
      "grad_norm": 4.7427873611450195,
      "learning_rate": 3.545454545454546e-05,
      "loss": 0.2043,
      "step": 2980
    },
    {
      "epoch": 0.2981,
      "grad_norm": 10.001832008361816,
      "learning_rate": 3.544949494949495e-05,
      "loss": 0.4254,
      "step": 2981
    },
    {
      "epoch": 0.2982,
      "grad_norm": 1.504431962966919,
      "learning_rate": 3.5444444444444445e-05,
      "loss": 0.0607,
      "step": 2982
    },
    {
      "epoch": 0.2983,
      "grad_norm": 6.037239074707031,
      "learning_rate": 3.543939393939394e-05,
      "loss": 0.2657,
      "step": 2983
    },
    {
      "epoch": 0.2984,
      "grad_norm": 5.691475868225098,
      "learning_rate": 3.543434343434344e-05,
      "loss": 0.3511,
      "step": 2984
    },
    {
      "epoch": 0.2985,
      "grad_norm": 2.5042002201080322,
      "learning_rate": 3.5429292929292926e-05,
      "loss": 0.0635,
      "step": 2985
    },
    {
      "epoch": 0.2986,
      "grad_norm": 6.702972412109375,
      "learning_rate": 3.542424242424243e-05,
      "loss": 0.3273,
      "step": 2986
    },
    {
      "epoch": 0.2987,
      "grad_norm": 3.278676986694336,
      "learning_rate": 3.541919191919192e-05,
      "loss": 0.1098,
      "step": 2987
    },
    {
      "epoch": 0.2988,
      "grad_norm": 5.269872188568115,
      "learning_rate": 3.5414141414141414e-05,
      "loss": 0.1797,
      "step": 2988
    },
    {
      "epoch": 0.2989,
      "grad_norm": 4.022830486297607,
      "learning_rate": 3.540909090909091e-05,
      "loss": 0.189,
      "step": 2989
    },
    {
      "epoch": 0.299,
      "grad_norm": 5.121367454528809,
      "learning_rate": 3.5404040404040406e-05,
      "loss": 0.3489,
      "step": 2990
    },
    {
      "epoch": 0.2991,
      "grad_norm": 7.547750949859619,
      "learning_rate": 3.5398989898989895e-05,
      "loss": 0.7497,
      "step": 2991
    },
    {
      "epoch": 0.2992,
      "grad_norm": 5.84588623046875,
      "learning_rate": 3.53939393939394e-05,
      "loss": 0.2938,
      "step": 2992
    },
    {
      "epoch": 0.2993,
      "grad_norm": 7.81294584274292,
      "learning_rate": 3.538888888888889e-05,
      "loss": 0.3333,
      "step": 2993
    },
    {
      "epoch": 0.2994,
      "grad_norm": 7.569976806640625,
      "learning_rate": 3.538383838383838e-05,
      "loss": 0.1525,
      "step": 2994
    },
    {
      "epoch": 0.2995,
      "grad_norm": 5.5156707763671875,
      "learning_rate": 3.537878787878788e-05,
      "loss": 0.4096,
      "step": 2995
    },
    {
      "epoch": 0.2996,
      "grad_norm": 7.6986083984375,
      "learning_rate": 3.5373737373737375e-05,
      "loss": 0.0192,
      "step": 2996
    },
    {
      "epoch": 0.2997,
      "grad_norm": 5.128865718841553,
      "learning_rate": 3.536868686868687e-05,
      "loss": 0.2498,
      "step": 2997
    },
    {
      "epoch": 0.2998,
      "grad_norm": 5.830610275268555,
      "learning_rate": 3.536363636363637e-05,
      "loss": 0.1943,
      "step": 2998
    },
    {
      "epoch": 0.2999,
      "grad_norm": 3.8463263511657715,
      "learning_rate": 3.5358585858585856e-05,
      "loss": 0.1447,
      "step": 2999
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.305494785308838,
      "learning_rate": 3.535353535353535e-05,
      "loss": 0.0427,
      "step": 3000
    },
    {
      "epoch": 0.3001,
      "grad_norm": 1.6886930465698242,
      "learning_rate": 3.534848484848485e-05,
      "loss": 0.0781,
      "step": 3001
    },
    {
      "epoch": 0.3002,
      "grad_norm": 3.3400025367736816,
      "learning_rate": 3.5343434343434344e-05,
      "loss": 0.1248,
      "step": 3002
    },
    {
      "epoch": 0.3003,
      "grad_norm": 5.7248663902282715,
      "learning_rate": 3.533838383838384e-05,
      "loss": 0.2385,
      "step": 3003
    },
    {
      "epoch": 0.3004,
      "grad_norm": 1.9628992080688477,
      "learning_rate": 3.5333333333333336e-05,
      "loss": 0.0519,
      "step": 3004
    },
    {
      "epoch": 0.3005,
      "grad_norm": 3.9858639240264893,
      "learning_rate": 3.532828282828283e-05,
      "loss": 0.0898,
      "step": 3005
    },
    {
      "epoch": 0.3006,
      "grad_norm": 16.61094093322754,
      "learning_rate": 3.532323232323233e-05,
      "loss": 0.1726,
      "step": 3006
    },
    {
      "epoch": 0.3007,
      "grad_norm": 5.745222091674805,
      "learning_rate": 3.5318181818181824e-05,
      "loss": 0.2993,
      "step": 3007
    },
    {
      "epoch": 0.3008,
      "grad_norm": 3.0821187496185303,
      "learning_rate": 3.531313131313131e-05,
      "loss": 0.1091,
      "step": 3008
    },
    {
      "epoch": 0.3009,
      "grad_norm": 6.733768939971924,
      "learning_rate": 3.530808080808081e-05,
      "loss": 0.1899,
      "step": 3009
    },
    {
      "epoch": 0.301,
      "grad_norm": 9.392088890075684,
      "learning_rate": 3.5303030303030305e-05,
      "loss": 0.6259,
      "step": 3010
    },
    {
      "epoch": 0.3011,
      "grad_norm": 3.1441030502319336,
      "learning_rate": 3.52979797979798e-05,
      "loss": 0.1005,
      "step": 3011
    },
    {
      "epoch": 0.3012,
      "grad_norm": 4.417050838470459,
      "learning_rate": 3.52929292929293e-05,
      "loss": 0.2688,
      "step": 3012
    },
    {
      "epoch": 0.3013,
      "grad_norm": 9.104517936706543,
      "learning_rate": 3.528787878787879e-05,
      "loss": 0.6661,
      "step": 3013
    },
    {
      "epoch": 0.3014,
      "grad_norm": 0.9232983589172363,
      "learning_rate": 3.528282828282828e-05,
      "loss": 0.0126,
      "step": 3014
    },
    {
      "epoch": 0.3015,
      "grad_norm": 5.142658710479736,
      "learning_rate": 3.527777777777778e-05,
      "loss": 0.1198,
      "step": 3015
    },
    {
      "epoch": 0.3016,
      "grad_norm": 8.579031944274902,
      "learning_rate": 3.5272727272727274e-05,
      "loss": 0.3783,
      "step": 3016
    },
    {
      "epoch": 0.3017,
      "grad_norm": 3.1973652839660645,
      "learning_rate": 3.526767676767677e-05,
      "loss": 0.0985,
      "step": 3017
    },
    {
      "epoch": 0.3018,
      "grad_norm": 5.423099040985107,
      "learning_rate": 3.5262626262626266e-05,
      "loss": 0.3059,
      "step": 3018
    },
    {
      "epoch": 0.3019,
      "grad_norm": 10.397669792175293,
      "learning_rate": 3.525757575757576e-05,
      "loss": 0.2832,
      "step": 3019
    },
    {
      "epoch": 0.302,
      "grad_norm": 3.2989895343780518,
      "learning_rate": 3.525252525252525e-05,
      "loss": 0.1501,
      "step": 3020
    },
    {
      "epoch": 0.3021,
      "grad_norm": 16.82509422302246,
      "learning_rate": 3.5247474747474754e-05,
      "loss": 0.3805,
      "step": 3021
    },
    {
      "epoch": 0.3022,
      "grad_norm": 7.914895534515381,
      "learning_rate": 3.5242424242424244e-05,
      "loss": 0.1847,
      "step": 3022
    },
    {
      "epoch": 0.3023,
      "grad_norm": 2.3598194122314453,
      "learning_rate": 3.523737373737374e-05,
      "loss": 0.0857,
      "step": 3023
    },
    {
      "epoch": 0.3024,
      "grad_norm": 7.395264625549316,
      "learning_rate": 3.5232323232323236e-05,
      "loss": 0.2936,
      "step": 3024
    },
    {
      "epoch": 0.3025,
      "grad_norm": 4.00274133682251,
      "learning_rate": 3.522727272727273e-05,
      "loss": 0.0938,
      "step": 3025
    },
    {
      "epoch": 0.3026,
      "grad_norm": 5.283915042877197,
      "learning_rate": 3.522222222222222e-05,
      "loss": 0.1472,
      "step": 3026
    },
    {
      "epoch": 0.3027,
      "grad_norm": 2.6399216651916504,
      "learning_rate": 3.5217171717171724e-05,
      "loss": 0.0814,
      "step": 3027
    },
    {
      "epoch": 0.3028,
      "grad_norm": 10.085487365722656,
      "learning_rate": 3.521212121212121e-05,
      "loss": 0.3242,
      "step": 3028
    },
    {
      "epoch": 0.3029,
      "grad_norm": 13.139913558959961,
      "learning_rate": 3.520707070707071e-05,
      "loss": 0.1728,
      "step": 3029
    },
    {
      "epoch": 0.303,
      "grad_norm": 4.077488422393799,
      "learning_rate": 3.5202020202020205e-05,
      "loss": 0.1487,
      "step": 3030
    },
    {
      "epoch": 0.3031,
      "grad_norm": 6.456545829772949,
      "learning_rate": 3.51969696969697e-05,
      "loss": 0.2435,
      "step": 3031
    },
    {
      "epoch": 0.3032,
      "grad_norm": 7.784978866577148,
      "learning_rate": 3.519191919191919e-05,
      "loss": 0.4184,
      "step": 3032
    },
    {
      "epoch": 0.3033,
      "grad_norm": 5.935754299163818,
      "learning_rate": 3.518686868686869e-05,
      "loss": 0.2167,
      "step": 3033
    },
    {
      "epoch": 0.3034,
      "grad_norm": 3.966275215148926,
      "learning_rate": 3.518181818181818e-05,
      "loss": 0.1494,
      "step": 3034
    },
    {
      "epoch": 0.3035,
      "grad_norm": 7.480403900146484,
      "learning_rate": 3.517676767676768e-05,
      "loss": 0.2054,
      "step": 3035
    },
    {
      "epoch": 0.3036,
      "grad_norm": 3.8252665996551514,
      "learning_rate": 3.5171717171717174e-05,
      "loss": 0.1348,
      "step": 3036
    },
    {
      "epoch": 0.3037,
      "grad_norm": 4.612822532653809,
      "learning_rate": 3.516666666666667e-05,
      "loss": 0.0963,
      "step": 3037
    },
    {
      "epoch": 0.3038,
      "grad_norm": 8.95353889465332,
      "learning_rate": 3.516161616161616e-05,
      "loss": 0.4701,
      "step": 3038
    },
    {
      "epoch": 0.3039,
      "grad_norm": 6.6581268310546875,
      "learning_rate": 3.515656565656566e-05,
      "loss": 0.3868,
      "step": 3039
    },
    {
      "epoch": 0.304,
      "grad_norm": 2.6752748489379883,
      "learning_rate": 3.515151515151515e-05,
      "loss": 0.0409,
      "step": 3040
    },
    {
      "epoch": 0.3041,
      "grad_norm": 14.085469245910645,
      "learning_rate": 3.514646464646465e-05,
      "loss": 0.3228,
      "step": 3041
    },
    {
      "epoch": 0.3042,
      "grad_norm": 3.9810495376586914,
      "learning_rate": 3.514141414141414e-05,
      "loss": 0.0948,
      "step": 3042
    },
    {
      "epoch": 0.3043,
      "grad_norm": 5.304328441619873,
      "learning_rate": 3.513636363636364e-05,
      "loss": 0.216,
      "step": 3043
    },
    {
      "epoch": 0.3044,
      "grad_norm": 8.363078117370605,
      "learning_rate": 3.513131313131313e-05,
      "loss": 0.2746,
      "step": 3044
    },
    {
      "epoch": 0.3045,
      "grad_norm": 3.524003505706787,
      "learning_rate": 3.512626262626263e-05,
      "loss": 0.1202,
      "step": 3045
    },
    {
      "epoch": 0.3046,
      "grad_norm": 2.6287269592285156,
      "learning_rate": 3.512121212121212e-05,
      "loss": 0.1535,
      "step": 3046
    },
    {
      "epoch": 0.3047,
      "grad_norm": 7.896958827972412,
      "learning_rate": 3.5116161616161616e-05,
      "loss": 0.4376,
      "step": 3047
    },
    {
      "epoch": 0.3048,
      "grad_norm": 4.409791469573975,
      "learning_rate": 3.511111111111111e-05,
      "loss": 0.4337,
      "step": 3048
    },
    {
      "epoch": 0.3049,
      "grad_norm": 10.020447731018066,
      "learning_rate": 3.510606060606061e-05,
      "loss": 0.058,
      "step": 3049
    },
    {
      "epoch": 0.305,
      "grad_norm": 2.4831643104553223,
      "learning_rate": 3.5101010101010104e-05,
      "loss": 0.1079,
      "step": 3050
    },
    {
      "epoch": 0.3051,
      "grad_norm": 16.317537307739258,
      "learning_rate": 3.50959595959596e-05,
      "loss": 0.2635,
      "step": 3051
    },
    {
      "epoch": 0.3052,
      "grad_norm": 1.2129571437835693,
      "learning_rate": 3.509090909090909e-05,
      "loss": 0.0375,
      "step": 3052
    },
    {
      "epoch": 0.3053,
      "grad_norm": 5.207540988922119,
      "learning_rate": 3.5085858585858585e-05,
      "loss": 0.2996,
      "step": 3053
    },
    {
      "epoch": 0.3054,
      "grad_norm": 4.632513046264648,
      "learning_rate": 3.508080808080808e-05,
      "loss": 0.0818,
      "step": 3054
    },
    {
      "epoch": 0.3055,
      "grad_norm": 4.127265453338623,
      "learning_rate": 3.507575757575758e-05,
      "loss": 0.1411,
      "step": 3055
    },
    {
      "epoch": 0.3056,
      "grad_norm": 2.9996979236602783,
      "learning_rate": 3.5070707070707073e-05,
      "loss": 0.1236,
      "step": 3056
    },
    {
      "epoch": 0.3057,
      "grad_norm": 59.19034957885742,
      "learning_rate": 3.506565656565657e-05,
      "loss": 0.8047,
      "step": 3057
    },
    {
      "epoch": 0.3058,
      "grad_norm": 7.966578960418701,
      "learning_rate": 3.506060606060606e-05,
      "loss": 0.088,
      "step": 3058
    },
    {
      "epoch": 0.3059,
      "grad_norm": 5.887051582336426,
      "learning_rate": 3.505555555555556e-05,
      "loss": 0.2338,
      "step": 3059
    },
    {
      "epoch": 0.306,
      "grad_norm": 1.9538787603378296,
      "learning_rate": 3.505050505050505e-05,
      "loss": 0.049,
      "step": 3060
    },
    {
      "epoch": 0.3061,
      "grad_norm": 3.0234291553497314,
      "learning_rate": 3.5045454545454547e-05,
      "loss": 0.0893,
      "step": 3061
    },
    {
      "epoch": 0.3062,
      "grad_norm": 3.193443775177002,
      "learning_rate": 3.504040404040404e-05,
      "loss": 0.132,
      "step": 3062
    },
    {
      "epoch": 0.3063,
      "grad_norm": 5.092253684997559,
      "learning_rate": 3.503535353535354e-05,
      "loss": 0.0646,
      "step": 3063
    },
    {
      "epoch": 0.3064,
      "grad_norm": 2.531686305999756,
      "learning_rate": 3.503030303030303e-05,
      "loss": 0.0497,
      "step": 3064
    },
    {
      "epoch": 0.3065,
      "grad_norm": 4.967421054840088,
      "learning_rate": 3.502525252525253e-05,
      "loss": 0.2615,
      "step": 3065
    },
    {
      "epoch": 0.3066,
      "grad_norm": 2.914369821548462,
      "learning_rate": 3.502020202020202e-05,
      "loss": 0.0653,
      "step": 3066
    },
    {
      "epoch": 0.3067,
      "grad_norm": 6.547751426696777,
      "learning_rate": 3.5015151515151516e-05,
      "loss": 0.2941,
      "step": 3067
    },
    {
      "epoch": 0.3068,
      "grad_norm": 11.406584739685059,
      "learning_rate": 3.501010101010101e-05,
      "loss": 0.1492,
      "step": 3068
    },
    {
      "epoch": 0.3069,
      "grad_norm": 3.767721652984619,
      "learning_rate": 3.500505050505051e-05,
      "loss": 0.1973,
      "step": 3069
    },
    {
      "epoch": 0.307,
      "grad_norm": 4.181711196899414,
      "learning_rate": 3.5e-05,
      "loss": 0.1277,
      "step": 3070
    },
    {
      "epoch": 0.3071,
      "grad_norm": 4.995031833648682,
      "learning_rate": 3.49949494949495e-05,
      "loss": 0.2626,
      "step": 3071
    },
    {
      "epoch": 0.3072,
      "grad_norm": 6.668546199798584,
      "learning_rate": 3.498989898989899e-05,
      "loss": 0.4716,
      "step": 3072
    },
    {
      "epoch": 0.3073,
      "grad_norm": 4.580221652984619,
      "learning_rate": 3.4984848484848485e-05,
      "loss": 0.2054,
      "step": 3073
    },
    {
      "epoch": 0.3074,
      "grad_norm": 9.506974220275879,
      "learning_rate": 3.497979797979798e-05,
      "loss": 0.3118,
      "step": 3074
    },
    {
      "epoch": 0.3075,
      "grad_norm": 17.05318260192871,
      "learning_rate": 3.497474747474748e-05,
      "loss": 0.4127,
      "step": 3075
    },
    {
      "epoch": 0.3076,
      "grad_norm": 5.230858325958252,
      "learning_rate": 3.4969696969696966e-05,
      "loss": 0.232,
      "step": 3076
    },
    {
      "epoch": 0.3077,
      "grad_norm": 8.34482479095459,
      "learning_rate": 3.496464646464647e-05,
      "loss": 0.2276,
      "step": 3077
    },
    {
      "epoch": 0.3078,
      "grad_norm": 3.999443292617798,
      "learning_rate": 3.495959595959596e-05,
      "loss": 0.2249,
      "step": 3078
    },
    {
      "epoch": 0.3079,
      "grad_norm": 7.356476306915283,
      "learning_rate": 3.4954545454545454e-05,
      "loss": 0.2874,
      "step": 3079
    },
    {
      "epoch": 0.308,
      "grad_norm": 10.439369201660156,
      "learning_rate": 3.494949494949495e-05,
      "loss": 0.4055,
      "step": 3080
    },
    {
      "epoch": 0.3081,
      "grad_norm": 9.224799156188965,
      "learning_rate": 3.4944444444444446e-05,
      "loss": 0.2598,
      "step": 3081
    },
    {
      "epoch": 0.3082,
      "grad_norm": 12.97057056427002,
      "learning_rate": 3.4939393939393935e-05,
      "loss": 0.1568,
      "step": 3082
    },
    {
      "epoch": 0.3083,
      "grad_norm": 18.305675506591797,
      "learning_rate": 3.493434343434344e-05,
      "loss": 0.2014,
      "step": 3083
    },
    {
      "epoch": 0.3084,
      "grad_norm": 6.514719486236572,
      "learning_rate": 3.492929292929293e-05,
      "loss": 0.2449,
      "step": 3084
    },
    {
      "epoch": 0.3085,
      "grad_norm": 4.647457599639893,
      "learning_rate": 3.492424242424242e-05,
      "loss": 0.141,
      "step": 3085
    },
    {
      "epoch": 0.3086,
      "grad_norm": 3.9358413219451904,
      "learning_rate": 3.491919191919192e-05,
      "loss": 0.3371,
      "step": 3086
    },
    {
      "epoch": 0.3087,
      "grad_norm": 8.517447471618652,
      "learning_rate": 3.4914141414141415e-05,
      "loss": 0.4277,
      "step": 3087
    },
    {
      "epoch": 0.3088,
      "grad_norm": 2.7490437030792236,
      "learning_rate": 3.490909090909091e-05,
      "loss": 0.0768,
      "step": 3088
    },
    {
      "epoch": 0.3089,
      "grad_norm": 2.9202780723571777,
      "learning_rate": 3.490404040404041e-05,
      "loss": 0.1164,
      "step": 3089
    },
    {
      "epoch": 0.309,
      "grad_norm": 7.811361789703369,
      "learning_rate": 3.48989898989899e-05,
      "loss": 0.3852,
      "step": 3090
    },
    {
      "epoch": 0.3091,
      "grad_norm": 11.955670356750488,
      "learning_rate": 3.489393939393939e-05,
      "loss": 0.396,
      "step": 3091
    },
    {
      "epoch": 0.3092,
      "grad_norm": 8.775151252746582,
      "learning_rate": 3.4888888888888895e-05,
      "loss": 0.7625,
      "step": 3092
    },
    {
      "epoch": 0.3093,
      "grad_norm": 16.417560577392578,
      "learning_rate": 3.4883838383838384e-05,
      "loss": 0.4602,
      "step": 3093
    },
    {
      "epoch": 0.3094,
      "grad_norm": 6.095056533813477,
      "learning_rate": 3.487878787878788e-05,
      "loss": 0.1325,
      "step": 3094
    },
    {
      "epoch": 0.3095,
      "grad_norm": 11.352190017700195,
      "learning_rate": 3.4873737373737376e-05,
      "loss": 0.2735,
      "step": 3095
    },
    {
      "epoch": 0.3096,
      "grad_norm": 7.554999351501465,
      "learning_rate": 3.486868686868687e-05,
      "loss": 0.2437,
      "step": 3096
    },
    {
      "epoch": 0.3097,
      "grad_norm": 7.116336822509766,
      "learning_rate": 3.486363636363637e-05,
      "loss": 0.3094,
      "step": 3097
    },
    {
      "epoch": 0.3098,
      "grad_norm": 4.7672438621521,
      "learning_rate": 3.4858585858585864e-05,
      "loss": 0.192,
      "step": 3098
    },
    {
      "epoch": 0.3099,
      "grad_norm": 9.591049194335938,
      "learning_rate": 3.4853535353535354e-05,
      "loss": 0.6765,
      "step": 3099
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.709299564361572,
      "learning_rate": 3.484848484848485e-05,
      "loss": 0.2222,
      "step": 3100
    },
    {
      "epoch": 0.3101,
      "grad_norm": 6.584843635559082,
      "learning_rate": 3.4843434343434346e-05,
      "loss": 0.1602,
      "step": 3101
    },
    {
      "epoch": 0.3102,
      "grad_norm": 4.725000381469727,
      "learning_rate": 3.483838383838384e-05,
      "loss": 0.1462,
      "step": 3102
    },
    {
      "epoch": 0.3103,
      "grad_norm": 8.999610900878906,
      "learning_rate": 3.483333333333334e-05,
      "loss": 0.3009,
      "step": 3103
    },
    {
      "epoch": 0.3104,
      "grad_norm": 8.233969688415527,
      "learning_rate": 3.4828282828282834e-05,
      "loss": 0.2959,
      "step": 3104
    },
    {
      "epoch": 0.3105,
      "grad_norm": 3.1536693572998047,
      "learning_rate": 3.482323232323232e-05,
      "loss": 0.1883,
      "step": 3105
    },
    {
      "epoch": 0.3106,
      "grad_norm": 2.831092357635498,
      "learning_rate": 3.481818181818182e-05,
      "loss": 0.1506,
      "step": 3106
    },
    {
      "epoch": 0.3107,
      "grad_norm": 9.72860050201416,
      "learning_rate": 3.4813131313131315e-05,
      "loss": 0.1314,
      "step": 3107
    },
    {
      "epoch": 0.3108,
      "grad_norm": 10.651424407958984,
      "learning_rate": 3.480808080808081e-05,
      "loss": 0.2693,
      "step": 3108
    },
    {
      "epoch": 0.3109,
      "grad_norm": 5.669564247131348,
      "learning_rate": 3.480303030303031e-05,
      "loss": 0.597,
      "step": 3109
    },
    {
      "epoch": 0.311,
      "grad_norm": 6.188735485076904,
      "learning_rate": 3.47979797979798e-05,
      "loss": 0.3041,
      "step": 3110
    },
    {
      "epoch": 0.3111,
      "grad_norm": 8.169401168823242,
      "learning_rate": 3.479292929292929e-05,
      "loss": 0.4217,
      "step": 3111
    },
    {
      "epoch": 0.3112,
      "grad_norm": 10.562681198120117,
      "learning_rate": 3.4787878787878795e-05,
      "loss": 0.3646,
      "step": 3112
    },
    {
      "epoch": 0.3113,
      "grad_norm": 6.54860258102417,
      "learning_rate": 3.4782828282828284e-05,
      "loss": 0.3717,
      "step": 3113
    },
    {
      "epoch": 0.3114,
      "grad_norm": 21.3681583404541,
      "learning_rate": 3.477777777777778e-05,
      "loss": 0.3724,
      "step": 3114
    },
    {
      "epoch": 0.3115,
      "grad_norm": 7.624261379241943,
      "learning_rate": 3.4772727272727276e-05,
      "loss": 0.442,
      "step": 3115
    },
    {
      "epoch": 0.3116,
      "grad_norm": 12.375515937805176,
      "learning_rate": 3.476767676767677e-05,
      "loss": 0.6748,
      "step": 3116
    },
    {
      "epoch": 0.3117,
      "grad_norm": 14.827579498291016,
      "learning_rate": 3.476262626262626e-05,
      "loss": 0.6955,
      "step": 3117
    },
    {
      "epoch": 0.3118,
      "grad_norm": 17.75837516784668,
      "learning_rate": 3.4757575757575764e-05,
      "loss": 0.955,
      "step": 3118
    },
    {
      "epoch": 0.3119,
      "grad_norm": 21.56935691833496,
      "learning_rate": 3.475252525252525e-05,
      "loss": 0.7274,
      "step": 3119
    },
    {
      "epoch": 0.312,
      "grad_norm": 13.306337356567383,
      "learning_rate": 3.474747474747475e-05,
      "loss": 0.6782,
      "step": 3120
    },
    {
      "epoch": 0.3121,
      "grad_norm": 11.780943870544434,
      "learning_rate": 3.4742424242424245e-05,
      "loss": 0.4687,
      "step": 3121
    },
    {
      "epoch": 0.3122,
      "grad_norm": 9.720146179199219,
      "learning_rate": 3.473737373737374e-05,
      "loss": 0.7432,
      "step": 3122
    },
    {
      "epoch": 0.3123,
      "grad_norm": 9.756132125854492,
      "learning_rate": 3.473232323232323e-05,
      "loss": 0.4002,
      "step": 3123
    },
    {
      "epoch": 0.3124,
      "grad_norm": 8.465720176696777,
      "learning_rate": 3.472727272727273e-05,
      "loss": 0.4255,
      "step": 3124
    },
    {
      "epoch": 0.3125,
      "grad_norm": 8.185579299926758,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.4175,
      "step": 3125
    },
    {
      "epoch": 0.3126,
      "grad_norm": 9.700648307800293,
      "learning_rate": 3.471717171717172e-05,
      "loss": 0.6819,
      "step": 3126
    },
    {
      "epoch": 0.3127,
      "grad_norm": 19.37156867980957,
      "learning_rate": 3.4712121212121214e-05,
      "loss": 0.5589,
      "step": 3127
    },
    {
      "epoch": 0.3128,
      "grad_norm": 4.9773359298706055,
      "learning_rate": 3.470707070707071e-05,
      "loss": 0.3612,
      "step": 3128
    },
    {
      "epoch": 0.3129,
      "grad_norm": 7.031639099121094,
      "learning_rate": 3.47020202020202e-05,
      "loss": 0.2419,
      "step": 3129
    },
    {
      "epoch": 0.313,
      "grad_norm": 106.68329620361328,
      "learning_rate": 3.46969696969697e-05,
      "loss": 0.6921,
      "step": 3130
    },
    {
      "epoch": 0.3131,
      "grad_norm": 10.108094215393066,
      "learning_rate": 3.469191919191919e-05,
      "loss": 0.4326,
      "step": 3131
    },
    {
      "epoch": 0.3132,
      "grad_norm": 8.479120254516602,
      "learning_rate": 3.468686868686869e-05,
      "loss": 0.3662,
      "step": 3132
    },
    {
      "epoch": 0.3133,
      "grad_norm": 11.23681354522705,
      "learning_rate": 3.468181818181818e-05,
      "loss": 0.4666,
      "step": 3133
    },
    {
      "epoch": 0.3134,
      "grad_norm": 14.746354103088379,
      "learning_rate": 3.467676767676768e-05,
      "loss": 0.4959,
      "step": 3134
    },
    {
      "epoch": 0.3135,
      "grad_norm": 7.273380756378174,
      "learning_rate": 3.467171717171717e-05,
      "loss": 0.4982,
      "step": 3135
    },
    {
      "epoch": 0.3136,
      "grad_norm": 12.812945365905762,
      "learning_rate": 3.466666666666667e-05,
      "loss": 0.9746,
      "step": 3136
    },
    {
      "epoch": 0.3137,
      "grad_norm": 5.726686954498291,
      "learning_rate": 3.466161616161616e-05,
      "loss": 0.3696,
      "step": 3137
    },
    {
      "epoch": 0.3138,
      "grad_norm": 35.5625114440918,
      "learning_rate": 3.4656565656565657e-05,
      "loss": 0.6487,
      "step": 3138
    },
    {
      "epoch": 0.3139,
      "grad_norm": 8.973008155822754,
      "learning_rate": 3.465151515151515e-05,
      "loss": 0.6859,
      "step": 3139
    },
    {
      "epoch": 0.314,
      "grad_norm": 10.226852416992188,
      "learning_rate": 3.464646464646465e-05,
      "loss": 0.5502,
      "step": 3140
    },
    {
      "epoch": 0.3141,
      "grad_norm": 9.919788360595703,
      "learning_rate": 3.4641414141414144e-05,
      "loss": 0.5521,
      "step": 3141
    },
    {
      "epoch": 0.3142,
      "grad_norm": 11.381484985351562,
      "learning_rate": 3.463636363636364e-05,
      "loss": 0.7348,
      "step": 3142
    },
    {
      "epoch": 0.3143,
      "grad_norm": 6.944659233093262,
      "learning_rate": 3.463131313131313e-05,
      "loss": 0.4929,
      "step": 3143
    },
    {
      "epoch": 0.3144,
      "grad_norm": 10.244672775268555,
      "learning_rate": 3.4626262626262626e-05,
      "loss": 0.7144,
      "step": 3144
    },
    {
      "epoch": 0.3145,
      "grad_norm": 10.909564018249512,
      "learning_rate": 3.462121212121212e-05,
      "loss": 0.5222,
      "step": 3145
    },
    {
      "epoch": 0.3146,
      "grad_norm": 11.33370304107666,
      "learning_rate": 3.461616161616162e-05,
      "loss": 0.6947,
      "step": 3146
    },
    {
      "epoch": 0.3147,
      "grad_norm": 10.031362533569336,
      "learning_rate": 3.4611111111111114e-05,
      "loss": 0.8911,
      "step": 3147
    },
    {
      "epoch": 0.3148,
      "grad_norm": 13.224504470825195,
      "learning_rate": 3.460606060606061e-05,
      "loss": 0.3481,
      "step": 3148
    },
    {
      "epoch": 0.3149,
      "grad_norm": 17.368221282958984,
      "learning_rate": 3.46010101010101e-05,
      "loss": 0.5562,
      "step": 3149
    },
    {
      "epoch": 0.315,
      "grad_norm": 9.862252235412598,
      "learning_rate": 3.45959595959596e-05,
      "loss": 0.5705,
      "step": 3150
    },
    {
      "epoch": 0.3151,
      "grad_norm": 11.988974571228027,
      "learning_rate": 3.459090909090909e-05,
      "loss": 0.9288,
      "step": 3151
    },
    {
      "epoch": 0.3152,
      "grad_norm": 10.282784461975098,
      "learning_rate": 3.458585858585859e-05,
      "loss": 0.3777,
      "step": 3152
    },
    {
      "epoch": 0.3153,
      "grad_norm": 37.24238586425781,
      "learning_rate": 3.458080808080808e-05,
      "loss": 0.7476,
      "step": 3153
    },
    {
      "epoch": 0.3154,
      "grad_norm": 7.601330280303955,
      "learning_rate": 3.457575757575758e-05,
      "loss": 0.4639,
      "step": 3154
    },
    {
      "epoch": 0.3155,
      "grad_norm": 13.308720588684082,
      "learning_rate": 3.457070707070707e-05,
      "loss": 0.3138,
      "step": 3155
    },
    {
      "epoch": 0.3156,
      "grad_norm": 7.592794895172119,
      "learning_rate": 3.456565656565657e-05,
      "loss": 0.4378,
      "step": 3156
    },
    {
      "epoch": 0.3157,
      "grad_norm": 5.662071704864502,
      "learning_rate": 3.456060606060606e-05,
      "loss": 0.284,
      "step": 3157
    },
    {
      "epoch": 0.3158,
      "grad_norm": 6.851987838745117,
      "learning_rate": 3.4555555555555556e-05,
      "loss": 0.2445,
      "step": 3158
    },
    {
      "epoch": 0.3159,
      "grad_norm": 5.848402976989746,
      "learning_rate": 3.455050505050505e-05,
      "loss": 0.3338,
      "step": 3159
    },
    {
      "epoch": 0.316,
      "grad_norm": 9.376887321472168,
      "learning_rate": 3.454545454545455e-05,
      "loss": 0.5098,
      "step": 3160
    },
    {
      "epoch": 0.3161,
      "grad_norm": 5.041945934295654,
      "learning_rate": 3.454040404040404e-05,
      "loss": 0.5601,
      "step": 3161
    },
    {
      "epoch": 0.3162,
      "grad_norm": 8.676281929016113,
      "learning_rate": 3.453535353535354e-05,
      "loss": 0.4435,
      "step": 3162
    },
    {
      "epoch": 0.3163,
      "grad_norm": 5.454412460327148,
      "learning_rate": 3.453030303030303e-05,
      "loss": 0.4198,
      "step": 3163
    },
    {
      "epoch": 0.3164,
      "grad_norm": 6.304916858673096,
      "learning_rate": 3.4525252525252525e-05,
      "loss": 0.4968,
      "step": 3164
    },
    {
      "epoch": 0.3165,
      "grad_norm": 5.541451454162598,
      "learning_rate": 3.452020202020202e-05,
      "loss": 0.4392,
      "step": 3165
    },
    {
      "epoch": 0.3166,
      "grad_norm": 5.769317150115967,
      "learning_rate": 3.451515151515152e-05,
      "loss": 0.307,
      "step": 3166
    },
    {
      "epoch": 0.3167,
      "grad_norm": 7.18734884262085,
      "learning_rate": 3.4510101010101006e-05,
      "loss": 0.4744,
      "step": 3167
    },
    {
      "epoch": 0.3168,
      "grad_norm": 6.935091495513916,
      "learning_rate": 3.450505050505051e-05,
      "loss": 0.3633,
      "step": 3168
    },
    {
      "epoch": 0.3169,
      "grad_norm": 9.354524612426758,
      "learning_rate": 3.45e-05,
      "loss": 0.4663,
      "step": 3169
    },
    {
      "epoch": 0.317,
      "grad_norm": 9.906572341918945,
      "learning_rate": 3.4494949494949494e-05,
      "loss": 0.4366,
      "step": 3170
    },
    {
      "epoch": 0.3171,
      "grad_norm": 11.933297157287598,
      "learning_rate": 3.448989898989899e-05,
      "loss": 0.4685,
      "step": 3171
    },
    {
      "epoch": 0.3172,
      "grad_norm": 10.808647155761719,
      "learning_rate": 3.4484848484848486e-05,
      "loss": 0.6246,
      "step": 3172
    },
    {
      "epoch": 0.3173,
      "grad_norm": 7.737061977386475,
      "learning_rate": 3.447979797979798e-05,
      "loss": 0.5493,
      "step": 3173
    },
    {
      "epoch": 0.3174,
      "grad_norm": 66.26374053955078,
      "learning_rate": 3.447474747474748e-05,
      "loss": 0.5119,
      "step": 3174
    },
    {
      "epoch": 0.3175,
      "grad_norm": 7.889978408813477,
      "learning_rate": 3.4469696969696974e-05,
      "loss": 0.7587,
      "step": 3175
    },
    {
      "epoch": 0.3176,
      "grad_norm": 8.169955253601074,
      "learning_rate": 3.4464646464646463e-05,
      "loss": 0.355,
      "step": 3176
    },
    {
      "epoch": 0.3177,
      "grad_norm": 11.625186920166016,
      "learning_rate": 3.4459595959595966e-05,
      "loss": 0.5505,
      "step": 3177
    },
    {
      "epoch": 0.3178,
      "grad_norm": 18.6889705657959,
      "learning_rate": 3.4454545454545455e-05,
      "loss": 0.7684,
      "step": 3178
    },
    {
      "epoch": 0.3179,
      "grad_norm": 7.571485996246338,
      "learning_rate": 3.444949494949495e-05,
      "loss": 0.4178,
      "step": 3179
    },
    {
      "epoch": 0.318,
      "grad_norm": 9.913938522338867,
      "learning_rate": 3.444444444444445e-05,
      "loss": 0.8426,
      "step": 3180
    },
    {
      "epoch": 0.3181,
      "grad_norm": 8.374246597290039,
      "learning_rate": 3.4439393939393943e-05,
      "loss": 0.7367,
      "step": 3181
    },
    {
      "epoch": 0.3182,
      "grad_norm": 8.70030689239502,
      "learning_rate": 3.443434343434343e-05,
      "loss": 0.9609,
      "step": 3182
    },
    {
      "epoch": 0.3183,
      "grad_norm": 6.742825508117676,
      "learning_rate": 3.4429292929292935e-05,
      "loss": 0.3047,
      "step": 3183
    },
    {
      "epoch": 0.3184,
      "grad_norm": 10.505990982055664,
      "learning_rate": 3.4424242424242425e-05,
      "loss": 0.7882,
      "step": 3184
    },
    {
      "epoch": 0.3185,
      "grad_norm": 74.44992065429688,
      "learning_rate": 3.441919191919192e-05,
      "loss": 0.6666,
      "step": 3185
    },
    {
      "epoch": 0.3186,
      "grad_norm": 15.565254211425781,
      "learning_rate": 3.441414141414142e-05,
      "loss": 0.4518,
      "step": 3186
    },
    {
      "epoch": 0.3187,
      "grad_norm": 26.565799713134766,
      "learning_rate": 3.440909090909091e-05,
      "loss": 0.5495,
      "step": 3187
    },
    {
      "epoch": 0.3188,
      "grad_norm": 11.282177925109863,
      "learning_rate": 3.44040404040404e-05,
      "loss": 0.6898,
      "step": 3188
    },
    {
      "epoch": 0.3189,
      "grad_norm": 5.6520586013793945,
      "learning_rate": 3.4398989898989905e-05,
      "loss": 0.6377,
      "step": 3189
    },
    {
      "epoch": 0.319,
      "grad_norm": 10.2531099319458,
      "learning_rate": 3.4393939393939394e-05,
      "loss": 0.7603,
      "step": 3190
    },
    {
      "epoch": 0.3191,
      "grad_norm": 10.90894889831543,
      "learning_rate": 3.438888888888889e-05,
      "loss": 1.0072,
      "step": 3191
    },
    {
      "epoch": 0.3192,
      "grad_norm": 10.823732376098633,
      "learning_rate": 3.4383838383838386e-05,
      "loss": 0.5819,
      "step": 3192
    },
    {
      "epoch": 0.3193,
      "grad_norm": 10.82473373413086,
      "learning_rate": 3.437878787878788e-05,
      "loss": 1.0482,
      "step": 3193
    },
    {
      "epoch": 0.3194,
      "grad_norm": 13.452539443969727,
      "learning_rate": 3.437373737373738e-05,
      "loss": 1.2398,
      "step": 3194
    },
    {
      "epoch": 0.3195,
      "grad_norm": 10.149060249328613,
      "learning_rate": 3.4368686868686874e-05,
      "loss": 0.511,
      "step": 3195
    },
    {
      "epoch": 0.3196,
      "grad_norm": 12.043524742126465,
      "learning_rate": 3.436363636363636e-05,
      "loss": 0.8712,
      "step": 3196
    },
    {
      "epoch": 0.3197,
      "grad_norm": 5.230590343475342,
      "learning_rate": 3.435858585858586e-05,
      "loss": 0.2848,
      "step": 3197
    },
    {
      "epoch": 0.3198,
      "grad_norm": 10.165369987487793,
      "learning_rate": 3.4353535353535355e-05,
      "loss": 0.3613,
      "step": 3198
    },
    {
      "epoch": 0.3199,
      "grad_norm": 10.382078170776367,
      "learning_rate": 3.434848484848485e-05,
      "loss": 0.5257,
      "step": 3199
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.56210994720459,
      "learning_rate": 3.434343434343435e-05,
      "loss": 0.8349,
      "step": 3200
    },
    {
      "epoch": 0.3201,
      "grad_norm": 21.061750411987305,
      "learning_rate": 3.433838383838384e-05,
      "loss": 1.4607,
      "step": 3201
    },
    {
      "epoch": 0.3202,
      "grad_norm": 6.7445244789123535,
      "learning_rate": 3.433333333333333e-05,
      "loss": 0.5151,
      "step": 3202
    },
    {
      "epoch": 0.3203,
      "grad_norm": 7.415151119232178,
      "learning_rate": 3.4328282828282835e-05,
      "loss": 0.484,
      "step": 3203
    },
    {
      "epoch": 0.3204,
      "grad_norm": 9.454521179199219,
      "learning_rate": 3.4323232323232324e-05,
      "loss": 0.4354,
      "step": 3204
    },
    {
      "epoch": 0.3205,
      "grad_norm": 7.947419166564941,
      "learning_rate": 3.431818181818182e-05,
      "loss": 0.6454,
      "step": 3205
    },
    {
      "epoch": 0.3206,
      "grad_norm": 6.396843910217285,
      "learning_rate": 3.4313131313131316e-05,
      "loss": 0.3998,
      "step": 3206
    },
    {
      "epoch": 0.3207,
      "grad_norm": 6.799158573150635,
      "learning_rate": 3.430808080808081e-05,
      "loss": 0.3353,
      "step": 3207
    },
    {
      "epoch": 0.3208,
      "grad_norm": 10.204440116882324,
      "learning_rate": 3.43030303030303e-05,
      "loss": 0.4286,
      "step": 3208
    },
    {
      "epoch": 0.3209,
      "grad_norm": 15.819999694824219,
      "learning_rate": 3.4297979797979804e-05,
      "loss": 0.6996,
      "step": 3209
    },
    {
      "epoch": 0.321,
      "grad_norm": 15.487722396850586,
      "learning_rate": 3.429292929292929e-05,
      "loss": 0.5503,
      "step": 3210
    },
    {
      "epoch": 0.3211,
      "grad_norm": 7.578764915466309,
      "learning_rate": 3.428787878787879e-05,
      "loss": 0.5321,
      "step": 3211
    },
    {
      "epoch": 0.3212,
      "grad_norm": 20.487079620361328,
      "learning_rate": 3.4282828282828285e-05,
      "loss": 0.5405,
      "step": 3212
    },
    {
      "epoch": 0.3213,
      "grad_norm": 6.928445816040039,
      "learning_rate": 3.427777777777778e-05,
      "loss": 0.3426,
      "step": 3213
    },
    {
      "epoch": 0.3214,
      "grad_norm": 19.535995483398438,
      "learning_rate": 3.427272727272727e-05,
      "loss": 0.4557,
      "step": 3214
    },
    {
      "epoch": 0.3215,
      "grad_norm": 7.40097713470459,
      "learning_rate": 3.426767676767677e-05,
      "loss": 0.3814,
      "step": 3215
    },
    {
      "epoch": 0.3216,
      "grad_norm": 40.48771286010742,
      "learning_rate": 3.426262626262626e-05,
      "loss": 0.4366,
      "step": 3216
    },
    {
      "epoch": 0.3217,
      "grad_norm": 5.4037909507751465,
      "learning_rate": 3.425757575757576e-05,
      "loss": 0.3441,
      "step": 3217
    },
    {
      "epoch": 0.3218,
      "grad_norm": 12.695777893066406,
      "learning_rate": 3.4252525252525254e-05,
      "loss": 0.565,
      "step": 3218
    },
    {
      "epoch": 0.3219,
      "grad_norm": 9.421560287475586,
      "learning_rate": 3.424747474747475e-05,
      "loss": 0.535,
      "step": 3219
    },
    {
      "epoch": 0.322,
      "grad_norm": 12.443925857543945,
      "learning_rate": 3.424242424242424e-05,
      "loss": 1.0446,
      "step": 3220
    },
    {
      "epoch": 0.3221,
      "grad_norm": 9.996309280395508,
      "learning_rate": 3.423737373737374e-05,
      "loss": 0.4448,
      "step": 3221
    },
    {
      "epoch": 0.3222,
      "grad_norm": 16.148225784301758,
      "learning_rate": 3.423232323232323e-05,
      "loss": 0.4859,
      "step": 3222
    },
    {
      "epoch": 0.3223,
      "grad_norm": 5.247128963470459,
      "learning_rate": 3.422727272727273e-05,
      "loss": 0.2097,
      "step": 3223
    },
    {
      "epoch": 0.3224,
      "grad_norm": 8.854423522949219,
      "learning_rate": 3.4222222222222224e-05,
      "loss": 0.4856,
      "step": 3224
    },
    {
      "epoch": 0.3225,
      "grad_norm": 16.1281795501709,
      "learning_rate": 3.421717171717172e-05,
      "loss": 0.8742,
      "step": 3225
    },
    {
      "epoch": 0.3226,
      "grad_norm": 6.227382183074951,
      "learning_rate": 3.421212121212121e-05,
      "loss": 0.3955,
      "step": 3226
    },
    {
      "epoch": 0.3227,
      "grad_norm": 14.047429084777832,
      "learning_rate": 3.420707070707071e-05,
      "loss": 0.5556,
      "step": 3227
    },
    {
      "epoch": 0.3228,
      "grad_norm": 14.364662170410156,
      "learning_rate": 3.42020202020202e-05,
      "loss": 0.9564,
      "step": 3228
    },
    {
      "epoch": 0.3229,
      "grad_norm": 8.17110538482666,
      "learning_rate": 3.41969696969697e-05,
      "loss": 0.5187,
      "step": 3229
    },
    {
      "epoch": 0.323,
      "grad_norm": 10.19742202758789,
      "learning_rate": 3.419191919191919e-05,
      "loss": 0.6453,
      "step": 3230
    },
    {
      "epoch": 0.3231,
      "grad_norm": 25.366657257080078,
      "learning_rate": 3.418686868686869e-05,
      "loss": 0.707,
      "step": 3231
    },
    {
      "epoch": 0.3232,
      "grad_norm": 6.395910739898682,
      "learning_rate": 3.4181818181818185e-05,
      "loss": 0.354,
      "step": 3232
    },
    {
      "epoch": 0.3233,
      "grad_norm": 12.455163955688477,
      "learning_rate": 3.417676767676768e-05,
      "loss": 0.5333,
      "step": 3233
    },
    {
      "epoch": 0.3234,
      "grad_norm": 32.61097717285156,
      "learning_rate": 3.417171717171717e-05,
      "loss": 0.5418,
      "step": 3234
    },
    {
      "epoch": 0.3235,
      "grad_norm": 8.479198455810547,
      "learning_rate": 3.4166666666666666e-05,
      "loss": 0.4506,
      "step": 3235
    },
    {
      "epoch": 0.3236,
      "grad_norm": 12.048752784729004,
      "learning_rate": 3.416161616161616e-05,
      "loss": 0.3822,
      "step": 3236
    },
    {
      "epoch": 0.3237,
      "grad_norm": 9.503148078918457,
      "learning_rate": 3.415656565656566e-05,
      "loss": 0.4079,
      "step": 3237
    },
    {
      "epoch": 0.3238,
      "grad_norm": 6.308927059173584,
      "learning_rate": 3.4151515151515154e-05,
      "loss": 0.5002,
      "step": 3238
    },
    {
      "epoch": 0.3239,
      "grad_norm": 13.483419418334961,
      "learning_rate": 3.414646464646465e-05,
      "loss": 0.5113,
      "step": 3239
    },
    {
      "epoch": 0.324,
      "grad_norm": 6.304064750671387,
      "learning_rate": 3.414141414141414e-05,
      "loss": 0.3362,
      "step": 3240
    },
    {
      "epoch": 0.3241,
      "grad_norm": 11.622244834899902,
      "learning_rate": 3.413636363636364e-05,
      "loss": 0.3559,
      "step": 3241
    },
    {
      "epoch": 0.3242,
      "grad_norm": 8.017684936523438,
      "learning_rate": 3.413131313131313e-05,
      "loss": 0.546,
      "step": 3242
    },
    {
      "epoch": 0.3243,
      "grad_norm": 7.59069299697876,
      "learning_rate": 3.412626262626263e-05,
      "loss": 0.4495,
      "step": 3243
    },
    {
      "epoch": 0.3244,
      "grad_norm": 8.14551067352295,
      "learning_rate": 3.412121212121212e-05,
      "loss": 0.4066,
      "step": 3244
    },
    {
      "epoch": 0.3245,
      "grad_norm": 8.886354446411133,
      "learning_rate": 3.411616161616162e-05,
      "loss": 0.5176,
      "step": 3245
    },
    {
      "epoch": 0.3246,
      "grad_norm": 10.847501754760742,
      "learning_rate": 3.411111111111111e-05,
      "loss": 0.271,
      "step": 3246
    },
    {
      "epoch": 0.3247,
      "grad_norm": 9.48243236541748,
      "learning_rate": 3.410606060606061e-05,
      "loss": 0.5806,
      "step": 3247
    },
    {
      "epoch": 0.3248,
      "grad_norm": 11.784943580627441,
      "learning_rate": 3.41010101010101e-05,
      "loss": 0.7023,
      "step": 3248
    },
    {
      "epoch": 0.3249,
      "grad_norm": 6.068408966064453,
      "learning_rate": 3.4095959595959596e-05,
      "loss": 0.3787,
      "step": 3249
    },
    {
      "epoch": 0.325,
      "grad_norm": 9.223394393920898,
      "learning_rate": 3.409090909090909e-05,
      "loss": 0.4204,
      "step": 3250
    },
    {
      "epoch": 0.3251,
      "grad_norm": 11.689226150512695,
      "learning_rate": 3.408585858585859e-05,
      "loss": 0.4244,
      "step": 3251
    },
    {
      "epoch": 0.3252,
      "grad_norm": 20.091707229614258,
      "learning_rate": 3.408080808080808e-05,
      "loss": 0.4155,
      "step": 3252
    },
    {
      "epoch": 0.3253,
      "grad_norm": 37.167510986328125,
      "learning_rate": 3.407575757575758e-05,
      "loss": 0.9241,
      "step": 3253
    },
    {
      "epoch": 0.3254,
      "grad_norm": 8.224632263183594,
      "learning_rate": 3.407070707070707e-05,
      "loss": 0.5726,
      "step": 3254
    },
    {
      "epoch": 0.3255,
      "grad_norm": 6.630545139312744,
      "learning_rate": 3.4065656565656565e-05,
      "loss": 0.3951,
      "step": 3255
    },
    {
      "epoch": 0.3256,
      "grad_norm": 7.790348052978516,
      "learning_rate": 3.406060606060606e-05,
      "loss": 0.3932,
      "step": 3256
    },
    {
      "epoch": 0.3257,
      "grad_norm": 15.693055152893066,
      "learning_rate": 3.405555555555556e-05,
      "loss": 0.7963,
      "step": 3257
    },
    {
      "epoch": 0.3258,
      "grad_norm": 7.561847686767578,
      "learning_rate": 3.405050505050505e-05,
      "loss": 0.4654,
      "step": 3258
    },
    {
      "epoch": 0.3259,
      "grad_norm": 6.775233268737793,
      "learning_rate": 3.404545454545455e-05,
      "loss": 0.3509,
      "step": 3259
    },
    {
      "epoch": 0.326,
      "grad_norm": 7.362460136413574,
      "learning_rate": 3.4040404040404045e-05,
      "loss": 0.5765,
      "step": 3260
    },
    {
      "epoch": 0.3261,
      "grad_norm": 9.303271293640137,
      "learning_rate": 3.4035353535353535e-05,
      "loss": 0.676,
      "step": 3261
    },
    {
      "epoch": 0.3262,
      "grad_norm": 6.460914611816406,
      "learning_rate": 3.403030303030304e-05,
      "loss": 0.4896,
      "step": 3262
    },
    {
      "epoch": 0.3263,
      "grad_norm": 18.60186195373535,
      "learning_rate": 3.4025252525252527e-05,
      "loss": 1.1894,
      "step": 3263
    },
    {
      "epoch": 0.3264,
      "grad_norm": 9.62051773071289,
      "learning_rate": 3.402020202020202e-05,
      "loss": 0.6626,
      "step": 3264
    },
    {
      "epoch": 0.3265,
      "grad_norm": 8.979338645935059,
      "learning_rate": 3.401515151515152e-05,
      "loss": 0.5824,
      "step": 3265
    },
    {
      "epoch": 0.3266,
      "grad_norm": 14.275004386901855,
      "learning_rate": 3.4010101010101015e-05,
      "loss": 0.3642,
      "step": 3266
    },
    {
      "epoch": 0.3267,
      "grad_norm": 10.674919128417969,
      "learning_rate": 3.4005050505050504e-05,
      "loss": 0.491,
      "step": 3267
    },
    {
      "epoch": 0.3268,
      "grad_norm": 13.42551040649414,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.4141,
      "step": 3268
    },
    {
      "epoch": 0.3269,
      "grad_norm": 4.890893936157227,
      "learning_rate": 3.3994949494949496e-05,
      "loss": 0.3505,
      "step": 3269
    },
    {
      "epoch": 0.327,
      "grad_norm": 5.261952877044678,
      "learning_rate": 3.398989898989899e-05,
      "loss": 0.3651,
      "step": 3270
    },
    {
      "epoch": 0.3271,
      "grad_norm": 10.725546836853027,
      "learning_rate": 3.398484848484849e-05,
      "loss": 0.4134,
      "step": 3271
    },
    {
      "epoch": 0.3272,
      "grad_norm": 11.494209289550781,
      "learning_rate": 3.3979797979797984e-05,
      "loss": 0.453,
      "step": 3272
    },
    {
      "epoch": 0.3273,
      "grad_norm": 8.574488639831543,
      "learning_rate": 3.397474747474747e-05,
      "loss": 0.3822,
      "step": 3273
    },
    {
      "epoch": 0.3274,
      "grad_norm": 17.752971649169922,
      "learning_rate": 3.3969696969696976e-05,
      "loss": 0.5747,
      "step": 3274
    },
    {
      "epoch": 0.3275,
      "grad_norm": 5.166340351104736,
      "learning_rate": 3.3964646464646465e-05,
      "loss": 0.3306,
      "step": 3275
    },
    {
      "epoch": 0.3276,
      "grad_norm": 14.182815551757812,
      "learning_rate": 3.395959595959596e-05,
      "loss": 0.6223,
      "step": 3276
    },
    {
      "epoch": 0.3277,
      "grad_norm": 7.668318748474121,
      "learning_rate": 3.395454545454546e-05,
      "loss": 0.4832,
      "step": 3277
    },
    {
      "epoch": 0.3278,
      "grad_norm": 14.602546691894531,
      "learning_rate": 3.394949494949495e-05,
      "loss": 0.3258,
      "step": 3278
    },
    {
      "epoch": 0.3279,
      "grad_norm": 9.44419002532959,
      "learning_rate": 3.394444444444444e-05,
      "loss": 0.4639,
      "step": 3279
    },
    {
      "epoch": 0.328,
      "grad_norm": 6.743853569030762,
      "learning_rate": 3.3939393939393945e-05,
      "loss": 0.4207,
      "step": 3280
    },
    {
      "epoch": 0.3281,
      "grad_norm": 13.38227653503418,
      "learning_rate": 3.3934343434343434e-05,
      "loss": 0.5259,
      "step": 3281
    },
    {
      "epoch": 0.3282,
      "grad_norm": 7.493679046630859,
      "learning_rate": 3.392929292929293e-05,
      "loss": 0.7123,
      "step": 3282
    },
    {
      "epoch": 0.3283,
      "grad_norm": 5.550490856170654,
      "learning_rate": 3.3924242424242426e-05,
      "loss": 0.4402,
      "step": 3283
    },
    {
      "epoch": 0.3284,
      "grad_norm": 12.268549919128418,
      "learning_rate": 3.391919191919192e-05,
      "loss": 0.4309,
      "step": 3284
    },
    {
      "epoch": 0.3285,
      "grad_norm": 14.05813217163086,
      "learning_rate": 3.391414141414142e-05,
      "loss": 0.4401,
      "step": 3285
    },
    {
      "epoch": 0.3286,
      "grad_norm": 6.608469486236572,
      "learning_rate": 3.3909090909090914e-05,
      "loss": 0.4795,
      "step": 3286
    },
    {
      "epoch": 0.3287,
      "grad_norm": 6.38838005065918,
      "learning_rate": 3.39040404040404e-05,
      "loss": 0.3914,
      "step": 3287
    },
    {
      "epoch": 0.3288,
      "grad_norm": 12.378384590148926,
      "learning_rate": 3.38989898989899e-05,
      "loss": 0.7921,
      "step": 3288
    },
    {
      "epoch": 0.3289,
      "grad_norm": 11.857975959777832,
      "learning_rate": 3.3893939393939395e-05,
      "loss": 0.5681,
      "step": 3289
    },
    {
      "epoch": 0.329,
      "grad_norm": 9.49673843383789,
      "learning_rate": 3.388888888888889e-05,
      "loss": 0.5861,
      "step": 3290
    },
    {
      "epoch": 0.3291,
      "grad_norm": 13.035931587219238,
      "learning_rate": 3.388383838383839e-05,
      "loss": 0.7138,
      "step": 3291
    },
    {
      "epoch": 0.3292,
      "grad_norm": 11.139151573181152,
      "learning_rate": 3.387878787878788e-05,
      "loss": 0.7856,
      "step": 3292
    },
    {
      "epoch": 0.3293,
      "grad_norm": 28.152225494384766,
      "learning_rate": 3.387373737373737e-05,
      "loss": 0.4642,
      "step": 3293
    },
    {
      "epoch": 0.3294,
      "grad_norm": 11.817635536193848,
      "learning_rate": 3.3868686868686875e-05,
      "loss": 0.3867,
      "step": 3294
    },
    {
      "epoch": 0.3295,
      "grad_norm": 12.738205909729004,
      "learning_rate": 3.3863636363636364e-05,
      "loss": 0.3002,
      "step": 3295
    },
    {
      "epoch": 0.3296,
      "grad_norm": 10.273354530334473,
      "learning_rate": 3.385858585858586e-05,
      "loss": 0.5942,
      "step": 3296
    },
    {
      "epoch": 0.3297,
      "grad_norm": 14.170232772827148,
      "learning_rate": 3.3853535353535356e-05,
      "loss": 0.5469,
      "step": 3297
    },
    {
      "epoch": 0.3298,
      "grad_norm": 9.336678504943848,
      "learning_rate": 3.384848484848485e-05,
      "loss": 0.4343,
      "step": 3298
    },
    {
      "epoch": 0.3299,
      "grad_norm": 16.83331298828125,
      "learning_rate": 3.384343434343434e-05,
      "loss": 0.2665,
      "step": 3299
    },
    {
      "epoch": 0.33,
      "grad_norm": 35.40017318725586,
      "learning_rate": 3.3838383838383844e-05,
      "loss": 0.5271,
      "step": 3300
    },
    {
      "epoch": 0.3301,
      "grad_norm": 7.5755720138549805,
      "learning_rate": 3.3833333333333334e-05,
      "loss": 0.3479,
      "step": 3301
    },
    {
      "epoch": 0.3302,
      "grad_norm": 6.116621017456055,
      "learning_rate": 3.382828282828283e-05,
      "loss": 0.311,
      "step": 3302
    },
    {
      "epoch": 0.3303,
      "grad_norm": 14.51291561126709,
      "learning_rate": 3.3823232323232326e-05,
      "loss": 0.7704,
      "step": 3303
    },
    {
      "epoch": 0.3304,
      "grad_norm": 7.521892547607422,
      "learning_rate": 3.381818181818182e-05,
      "loss": 0.4705,
      "step": 3304
    },
    {
      "epoch": 0.3305,
      "grad_norm": 24.65784454345703,
      "learning_rate": 3.381313131313131e-05,
      "loss": 1.2021,
      "step": 3305
    },
    {
      "epoch": 0.3306,
      "grad_norm": 16.232389450073242,
      "learning_rate": 3.3808080808080813e-05,
      "loss": 0.7399,
      "step": 3306
    },
    {
      "epoch": 0.3307,
      "grad_norm": 41.998531341552734,
      "learning_rate": 3.38030303030303e-05,
      "loss": 0.8508,
      "step": 3307
    },
    {
      "epoch": 0.3308,
      "grad_norm": 13.012611389160156,
      "learning_rate": 3.37979797979798e-05,
      "loss": 0.8598,
      "step": 3308
    },
    {
      "epoch": 0.3309,
      "grad_norm": 7.63029670715332,
      "learning_rate": 3.3792929292929295e-05,
      "loss": 0.4438,
      "step": 3309
    },
    {
      "epoch": 0.331,
      "grad_norm": 8.368406295776367,
      "learning_rate": 3.378787878787879e-05,
      "loss": 0.2953,
      "step": 3310
    },
    {
      "epoch": 0.3311,
      "grad_norm": 8.278997421264648,
      "learning_rate": 3.378282828282828e-05,
      "loss": 0.5088,
      "step": 3311
    },
    {
      "epoch": 0.3312,
      "grad_norm": 18.881837844848633,
      "learning_rate": 3.377777777777778e-05,
      "loss": 0.6079,
      "step": 3312
    },
    {
      "epoch": 0.3313,
      "grad_norm": 24.1560115814209,
      "learning_rate": 3.377272727272727e-05,
      "loss": 0.4137,
      "step": 3313
    },
    {
      "epoch": 0.3314,
      "grad_norm": 39.13422393798828,
      "learning_rate": 3.376767676767677e-05,
      "loss": 1.6993,
      "step": 3314
    },
    {
      "epoch": 0.3315,
      "grad_norm": 16.93741226196289,
      "learning_rate": 3.3762626262626264e-05,
      "loss": 0.5018,
      "step": 3315
    },
    {
      "epoch": 0.3316,
      "grad_norm": 20.418926239013672,
      "learning_rate": 3.375757575757576e-05,
      "loss": 0.9873,
      "step": 3316
    },
    {
      "epoch": 0.3317,
      "grad_norm": 11.829527854919434,
      "learning_rate": 3.375252525252525e-05,
      "loss": 0.8268,
      "step": 3317
    },
    {
      "epoch": 0.3318,
      "grad_norm": 8.311075210571289,
      "learning_rate": 3.374747474747475e-05,
      "loss": 0.3968,
      "step": 3318
    },
    {
      "epoch": 0.3319,
      "grad_norm": 8.528266906738281,
      "learning_rate": 3.374242424242424e-05,
      "loss": 0.4843,
      "step": 3319
    },
    {
      "epoch": 0.332,
      "grad_norm": 7.502557277679443,
      "learning_rate": 3.373737373737374e-05,
      "loss": 0.465,
      "step": 3320
    },
    {
      "epoch": 0.3321,
      "grad_norm": 7.270773887634277,
      "learning_rate": 3.373232323232323e-05,
      "loss": 0.4792,
      "step": 3321
    },
    {
      "epoch": 0.3322,
      "grad_norm": 8.061473846435547,
      "learning_rate": 3.372727272727273e-05,
      "loss": 0.3665,
      "step": 3322
    },
    {
      "epoch": 0.3323,
      "grad_norm": 7.081747531890869,
      "learning_rate": 3.3722222222222225e-05,
      "loss": 0.4584,
      "step": 3323
    },
    {
      "epoch": 0.3324,
      "grad_norm": 13.870759963989258,
      "learning_rate": 3.371717171717172e-05,
      "loss": 0.5053,
      "step": 3324
    },
    {
      "epoch": 0.3325,
      "grad_norm": 7.583732604980469,
      "learning_rate": 3.371212121212121e-05,
      "loss": 0.4507,
      "step": 3325
    },
    {
      "epoch": 0.3326,
      "grad_norm": 7.7244062423706055,
      "learning_rate": 3.3707070707070706e-05,
      "loss": 0.3027,
      "step": 3326
    },
    {
      "epoch": 0.3327,
      "grad_norm": 26.313756942749023,
      "learning_rate": 3.37020202020202e-05,
      "loss": 0.4243,
      "step": 3327
    },
    {
      "epoch": 0.3328,
      "grad_norm": 8.175816535949707,
      "learning_rate": 3.36969696969697e-05,
      "loss": 0.2006,
      "step": 3328
    },
    {
      "epoch": 0.3329,
      "grad_norm": 7.819025993347168,
      "learning_rate": 3.3691919191919194e-05,
      "loss": 0.3756,
      "step": 3329
    },
    {
      "epoch": 0.333,
      "grad_norm": 15.25704288482666,
      "learning_rate": 3.368686868686869e-05,
      "loss": 0.7393,
      "step": 3330
    },
    {
      "epoch": 0.3331,
      "grad_norm": 9.564476013183594,
      "learning_rate": 3.368181818181818e-05,
      "loss": 0.5656,
      "step": 3331
    },
    {
      "epoch": 0.3332,
      "grad_norm": 9.044797897338867,
      "learning_rate": 3.367676767676768e-05,
      "loss": 0.3576,
      "step": 3332
    },
    {
      "epoch": 0.3333,
      "grad_norm": 13.161434173583984,
      "learning_rate": 3.367171717171717e-05,
      "loss": 1.0233,
      "step": 3333
    },
    {
      "epoch": 0.3334,
      "grad_norm": 15.551797866821289,
      "learning_rate": 3.366666666666667e-05,
      "loss": 0.4726,
      "step": 3334
    },
    {
      "epoch": 0.3335,
      "grad_norm": 19.006423950195312,
      "learning_rate": 3.366161616161616e-05,
      "loss": 0.4159,
      "step": 3335
    },
    {
      "epoch": 0.3336,
      "grad_norm": 10.727816581726074,
      "learning_rate": 3.365656565656566e-05,
      "loss": 0.4826,
      "step": 3336
    },
    {
      "epoch": 0.3337,
      "grad_norm": 12.494144439697266,
      "learning_rate": 3.365151515151515e-05,
      "loss": 0.5903,
      "step": 3337
    },
    {
      "epoch": 0.3338,
      "grad_norm": 50.88247299194336,
      "learning_rate": 3.364646464646465e-05,
      "loss": 0.4044,
      "step": 3338
    },
    {
      "epoch": 0.3339,
      "grad_norm": 6.742476463317871,
      "learning_rate": 3.364141414141414e-05,
      "loss": 0.5133,
      "step": 3339
    },
    {
      "epoch": 0.334,
      "grad_norm": 6.648828506469727,
      "learning_rate": 3.3636363636363636e-05,
      "loss": 0.3302,
      "step": 3340
    },
    {
      "epoch": 0.3341,
      "grad_norm": 4.997283458709717,
      "learning_rate": 3.363131313131313e-05,
      "loss": 0.291,
      "step": 3341
    },
    {
      "epoch": 0.3342,
      "grad_norm": 6.482469081878662,
      "learning_rate": 3.362626262626263e-05,
      "loss": 0.3002,
      "step": 3342
    },
    {
      "epoch": 0.3343,
      "grad_norm": 21.467926025390625,
      "learning_rate": 3.362121212121212e-05,
      "loss": 1.7834,
      "step": 3343
    },
    {
      "epoch": 0.3344,
      "grad_norm": 11.28796672821045,
      "learning_rate": 3.361616161616162e-05,
      "loss": 1.0072,
      "step": 3344
    },
    {
      "epoch": 0.3345,
      "grad_norm": 11.513529777526855,
      "learning_rate": 3.3611111111111116e-05,
      "loss": 0.567,
      "step": 3345
    },
    {
      "epoch": 0.3346,
      "grad_norm": 6.406225681304932,
      "learning_rate": 3.3606060606060606e-05,
      "loss": 0.42,
      "step": 3346
    },
    {
      "epoch": 0.3347,
      "grad_norm": 6.341663360595703,
      "learning_rate": 3.360101010101011e-05,
      "loss": 0.2129,
      "step": 3347
    },
    {
      "epoch": 0.3348,
      "grad_norm": 11.414776802062988,
      "learning_rate": 3.35959595959596e-05,
      "loss": 0.255,
      "step": 3348
    },
    {
      "epoch": 0.3349,
      "grad_norm": 8.64917278289795,
      "learning_rate": 3.3590909090909094e-05,
      "loss": 0.9055,
      "step": 3349
    },
    {
      "epoch": 0.335,
      "grad_norm": 5.873537540435791,
      "learning_rate": 3.358585858585859e-05,
      "loss": 0.3487,
      "step": 3350
    },
    {
      "epoch": 0.3351,
      "grad_norm": 13.156485557556152,
      "learning_rate": 3.3580808080808086e-05,
      "loss": 0.2876,
      "step": 3351
    },
    {
      "epoch": 0.3352,
      "grad_norm": 13.649948120117188,
      "learning_rate": 3.3575757575757575e-05,
      "loss": 0.6071,
      "step": 3352
    },
    {
      "epoch": 0.3353,
      "grad_norm": 14.274819374084473,
      "learning_rate": 3.357070707070708e-05,
      "loss": 0.2971,
      "step": 3353
    },
    {
      "epoch": 0.3354,
      "grad_norm": 6.820731163024902,
      "learning_rate": 3.356565656565657e-05,
      "loss": 0.3976,
      "step": 3354
    },
    {
      "epoch": 0.3355,
      "grad_norm": 9.366166114807129,
      "learning_rate": 3.356060606060606e-05,
      "loss": 0.7288,
      "step": 3355
    },
    {
      "epoch": 0.3356,
      "grad_norm": 7.071563243865967,
      "learning_rate": 3.355555555555556e-05,
      "loss": 0.2283,
      "step": 3356
    },
    {
      "epoch": 0.3357,
      "grad_norm": 4.864643096923828,
      "learning_rate": 3.3550505050505055e-05,
      "loss": 0.1943,
      "step": 3357
    },
    {
      "epoch": 0.3358,
      "grad_norm": 13.82902717590332,
      "learning_rate": 3.3545454545454544e-05,
      "loss": 1.7039,
      "step": 3358
    },
    {
      "epoch": 0.3359,
      "grad_norm": 11.385222434997559,
      "learning_rate": 3.354040404040405e-05,
      "loss": 0.622,
      "step": 3359
    },
    {
      "epoch": 0.336,
      "grad_norm": 10.793828964233398,
      "learning_rate": 3.3535353535353536e-05,
      "loss": 0.6428,
      "step": 3360
    },
    {
      "epoch": 0.3361,
      "grad_norm": 12.813886642456055,
      "learning_rate": 3.353030303030303e-05,
      "loss": 1.0799,
      "step": 3361
    },
    {
      "epoch": 0.3362,
      "grad_norm": 9.690165519714355,
      "learning_rate": 3.352525252525253e-05,
      "loss": 0.2523,
      "step": 3362
    },
    {
      "epoch": 0.3363,
      "grad_norm": 6.352994918823242,
      "learning_rate": 3.3520202020202024e-05,
      "loss": 0.2404,
      "step": 3363
    },
    {
      "epoch": 0.3364,
      "grad_norm": 8.724591255187988,
      "learning_rate": 3.351515151515151e-05,
      "loss": 0.3811,
      "step": 3364
    },
    {
      "epoch": 0.3365,
      "grad_norm": 4.703886032104492,
      "learning_rate": 3.3510101010101016e-05,
      "loss": 0.2452,
      "step": 3365
    },
    {
      "epoch": 0.3366,
      "grad_norm": 8.700347900390625,
      "learning_rate": 3.3505050505050505e-05,
      "loss": 0.4024,
      "step": 3366
    },
    {
      "epoch": 0.3367,
      "grad_norm": 8.132723808288574,
      "learning_rate": 3.35e-05,
      "loss": 0.3348,
      "step": 3367
    },
    {
      "epoch": 0.3368,
      "grad_norm": 5.291264057159424,
      "learning_rate": 3.34949494949495e-05,
      "loss": 0.3061,
      "step": 3368
    },
    {
      "epoch": 0.3369,
      "grad_norm": 17.586957931518555,
      "learning_rate": 3.348989898989899e-05,
      "loss": 1.0784,
      "step": 3369
    },
    {
      "epoch": 0.337,
      "grad_norm": 10.276276588439941,
      "learning_rate": 3.348484848484848e-05,
      "loss": 0.3978,
      "step": 3370
    },
    {
      "epoch": 0.3371,
      "grad_norm": 6.30996036529541,
      "learning_rate": 3.3479797979797985e-05,
      "loss": 0.4001,
      "step": 3371
    },
    {
      "epoch": 0.3372,
      "grad_norm": 12.593470573425293,
      "learning_rate": 3.3474747474747474e-05,
      "loss": 0.4226,
      "step": 3372
    },
    {
      "epoch": 0.3373,
      "grad_norm": 18.69746208190918,
      "learning_rate": 3.346969696969697e-05,
      "loss": 0.2921,
      "step": 3373
    },
    {
      "epoch": 0.3374,
      "grad_norm": 7.822885990142822,
      "learning_rate": 3.3464646464646466e-05,
      "loss": 0.393,
      "step": 3374
    },
    {
      "epoch": 0.3375,
      "grad_norm": 13.406366348266602,
      "learning_rate": 3.345959595959596e-05,
      "loss": 1.6057,
      "step": 3375
    },
    {
      "epoch": 0.3376,
      "grad_norm": 5.037789821624756,
      "learning_rate": 3.345454545454546e-05,
      "loss": 0.2816,
      "step": 3376
    },
    {
      "epoch": 0.3377,
      "grad_norm": 4.287835597991943,
      "learning_rate": 3.3449494949494954e-05,
      "loss": 0.2004,
      "step": 3377
    },
    {
      "epoch": 0.3378,
      "grad_norm": 6.152866363525391,
      "learning_rate": 3.3444444444444443e-05,
      "loss": 0.4411,
      "step": 3378
    },
    {
      "epoch": 0.3379,
      "grad_norm": 9.041057586669922,
      "learning_rate": 3.343939393939394e-05,
      "loss": 0.4661,
      "step": 3379
    },
    {
      "epoch": 0.338,
      "grad_norm": 7.007879257202148,
      "learning_rate": 3.3434343434343435e-05,
      "loss": 0.5053,
      "step": 3380
    },
    {
      "epoch": 0.3381,
      "grad_norm": 10.302860260009766,
      "learning_rate": 3.342929292929293e-05,
      "loss": 0.4369,
      "step": 3381
    },
    {
      "epoch": 0.3382,
      "grad_norm": 14.627808570861816,
      "learning_rate": 3.342424242424243e-05,
      "loss": 0.51,
      "step": 3382
    },
    {
      "epoch": 0.3383,
      "grad_norm": 10.958639144897461,
      "learning_rate": 3.3419191919191923e-05,
      "loss": 0.3492,
      "step": 3383
    },
    {
      "epoch": 0.3384,
      "grad_norm": 6.0565900802612305,
      "learning_rate": 3.341414141414141e-05,
      "loss": 0.1847,
      "step": 3384
    },
    {
      "epoch": 0.3385,
      "grad_norm": 5.846011638641357,
      "learning_rate": 3.3409090909090915e-05,
      "loss": 0.2377,
      "step": 3385
    },
    {
      "epoch": 0.3386,
      "grad_norm": 5.141702175140381,
      "learning_rate": 3.3404040404040405e-05,
      "loss": 0.324,
      "step": 3386
    },
    {
      "epoch": 0.3387,
      "grad_norm": 7.127209663391113,
      "learning_rate": 3.33989898989899e-05,
      "loss": 0.4218,
      "step": 3387
    },
    {
      "epoch": 0.3388,
      "grad_norm": 6.035386085510254,
      "learning_rate": 3.3393939393939397e-05,
      "loss": 0.4729,
      "step": 3388
    },
    {
      "epoch": 0.3389,
      "grad_norm": 5.541815280914307,
      "learning_rate": 3.338888888888889e-05,
      "loss": 0.3569,
      "step": 3389
    },
    {
      "epoch": 0.339,
      "grad_norm": 10.378235816955566,
      "learning_rate": 3.338383838383838e-05,
      "loss": 0.4066,
      "step": 3390
    },
    {
      "epoch": 0.3391,
      "grad_norm": 23.053049087524414,
      "learning_rate": 3.3378787878787885e-05,
      "loss": 0.2087,
      "step": 3391
    },
    {
      "epoch": 0.3392,
      "grad_norm": 14.374070167541504,
      "learning_rate": 3.3373737373737374e-05,
      "loss": 0.3109,
      "step": 3392
    },
    {
      "epoch": 0.3393,
      "grad_norm": 9.109834671020508,
      "learning_rate": 3.336868686868687e-05,
      "loss": 0.2656,
      "step": 3393
    },
    {
      "epoch": 0.3394,
      "grad_norm": 8.41564655303955,
      "learning_rate": 3.3363636363636366e-05,
      "loss": 0.1876,
      "step": 3394
    },
    {
      "epoch": 0.3395,
      "grad_norm": 8.279855728149414,
      "learning_rate": 3.335858585858586e-05,
      "loss": 0.3934,
      "step": 3395
    },
    {
      "epoch": 0.3396,
      "grad_norm": 9.91726303100586,
      "learning_rate": 3.335353535353535e-05,
      "loss": 0.373,
      "step": 3396
    },
    {
      "epoch": 0.3397,
      "grad_norm": 9.744751930236816,
      "learning_rate": 3.3348484848484854e-05,
      "loss": 0.6868,
      "step": 3397
    },
    {
      "epoch": 0.3398,
      "grad_norm": 10.128570556640625,
      "learning_rate": 3.334343434343434e-05,
      "loss": 0.3745,
      "step": 3398
    },
    {
      "epoch": 0.3399,
      "grad_norm": 12.379450798034668,
      "learning_rate": 3.333838383838384e-05,
      "loss": 0.655,
      "step": 3399
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.2608060836792,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.7886,
      "step": 3400
    },
    {
      "epoch": 0.3401,
      "grad_norm": 5.553603172302246,
      "learning_rate": 3.332828282828283e-05,
      "loss": 0.2932,
      "step": 3401
    },
    {
      "epoch": 0.3402,
      "grad_norm": 11.883697509765625,
      "learning_rate": 3.332323232323232e-05,
      "loss": 0.3926,
      "step": 3402
    },
    {
      "epoch": 0.3403,
      "grad_norm": 11.574810981750488,
      "learning_rate": 3.331818181818182e-05,
      "loss": 0.2747,
      "step": 3403
    },
    {
      "epoch": 0.3404,
      "grad_norm": 19.75455093383789,
      "learning_rate": 3.331313131313131e-05,
      "loss": 0.7579,
      "step": 3404
    },
    {
      "epoch": 0.3405,
      "grad_norm": 12.81447982788086,
      "learning_rate": 3.330808080808081e-05,
      "loss": 0.7273,
      "step": 3405
    },
    {
      "epoch": 0.3406,
      "grad_norm": 4.971012115478516,
      "learning_rate": 3.3303030303030304e-05,
      "loss": 0.1934,
      "step": 3406
    },
    {
      "epoch": 0.3407,
      "grad_norm": 12.313985824584961,
      "learning_rate": 3.32979797979798e-05,
      "loss": 0.5135,
      "step": 3407
    },
    {
      "epoch": 0.3408,
      "grad_norm": 5.3570027351379395,
      "learning_rate": 3.329292929292929e-05,
      "loss": 0.2192,
      "step": 3408
    },
    {
      "epoch": 0.3409,
      "grad_norm": 10.58215045928955,
      "learning_rate": 3.328787878787879e-05,
      "loss": 0.8696,
      "step": 3409
    },
    {
      "epoch": 0.341,
      "grad_norm": 40.75031280517578,
      "learning_rate": 3.328282828282828e-05,
      "loss": 0.3732,
      "step": 3410
    },
    {
      "epoch": 0.3411,
      "grad_norm": 6.300884246826172,
      "learning_rate": 3.327777777777778e-05,
      "loss": 0.2814,
      "step": 3411
    },
    {
      "epoch": 0.3412,
      "grad_norm": 4.441686153411865,
      "learning_rate": 3.327272727272727e-05,
      "loss": 0.2549,
      "step": 3412
    },
    {
      "epoch": 0.3413,
      "grad_norm": 15.565003395080566,
      "learning_rate": 3.326767676767677e-05,
      "loss": 0.7257,
      "step": 3413
    },
    {
      "epoch": 0.3414,
      "grad_norm": 6.09019660949707,
      "learning_rate": 3.3262626262626265e-05,
      "loss": 0.2921,
      "step": 3414
    },
    {
      "epoch": 0.3415,
      "grad_norm": 25.619848251342773,
      "learning_rate": 3.325757575757576e-05,
      "loss": 0.388,
      "step": 3415
    },
    {
      "epoch": 0.3416,
      "grad_norm": 8.017135620117188,
      "learning_rate": 3.325252525252525e-05,
      "loss": 0.3667,
      "step": 3416
    },
    {
      "epoch": 0.3417,
      "grad_norm": 11.762948036193848,
      "learning_rate": 3.3247474747474746e-05,
      "loss": 0.9396,
      "step": 3417
    },
    {
      "epoch": 0.3418,
      "grad_norm": 6.956809043884277,
      "learning_rate": 3.324242424242424e-05,
      "loss": 0.2575,
      "step": 3418
    },
    {
      "epoch": 0.3419,
      "grad_norm": 9.747032165527344,
      "learning_rate": 3.323737373737374e-05,
      "loss": 0.3856,
      "step": 3419
    },
    {
      "epoch": 0.342,
      "grad_norm": 9.379077911376953,
      "learning_rate": 3.3232323232323234e-05,
      "loss": 0.5364,
      "step": 3420
    },
    {
      "epoch": 0.3421,
      "grad_norm": 18.817096710205078,
      "learning_rate": 3.322727272727273e-05,
      "loss": 0.5073,
      "step": 3421
    },
    {
      "epoch": 0.3422,
      "grad_norm": 9.290568351745605,
      "learning_rate": 3.322222222222222e-05,
      "loss": 0.5139,
      "step": 3422
    },
    {
      "epoch": 0.3423,
      "grad_norm": 16.05621337890625,
      "learning_rate": 3.3217171717171716e-05,
      "loss": 0.1939,
      "step": 3423
    },
    {
      "epoch": 0.3424,
      "grad_norm": 11.268848419189453,
      "learning_rate": 3.321212121212121e-05,
      "loss": 0.5548,
      "step": 3424
    },
    {
      "epoch": 0.3425,
      "grad_norm": 11.245975494384766,
      "learning_rate": 3.320707070707071e-05,
      "loss": 0.9624,
      "step": 3425
    },
    {
      "epoch": 0.3426,
      "grad_norm": 10.5596923828125,
      "learning_rate": 3.3202020202020204e-05,
      "loss": 0.73,
      "step": 3426
    },
    {
      "epoch": 0.3427,
      "grad_norm": 14.125576972961426,
      "learning_rate": 3.31969696969697e-05,
      "loss": 0.598,
      "step": 3427
    },
    {
      "epoch": 0.3428,
      "grad_norm": 6.577502250671387,
      "learning_rate": 3.319191919191919e-05,
      "loss": 0.441,
      "step": 3428
    },
    {
      "epoch": 0.3429,
      "grad_norm": 10.251008033752441,
      "learning_rate": 3.318686868686869e-05,
      "loss": 0.5061,
      "step": 3429
    },
    {
      "epoch": 0.343,
      "grad_norm": 4.33845853805542,
      "learning_rate": 3.318181818181819e-05,
      "loss": 0.3156,
      "step": 3430
    },
    {
      "epoch": 0.3431,
      "grad_norm": 7.887010097503662,
      "learning_rate": 3.317676767676768e-05,
      "loss": 0.2621,
      "step": 3431
    },
    {
      "epoch": 0.3432,
      "grad_norm": 6.613101959228516,
      "learning_rate": 3.317171717171717e-05,
      "loss": 0.2548,
      "step": 3432
    },
    {
      "epoch": 0.3433,
      "grad_norm": 4.763879299163818,
      "learning_rate": 3.316666666666667e-05,
      "loss": 0.2358,
      "step": 3433
    },
    {
      "epoch": 0.3434,
      "grad_norm": 10.310782432556152,
      "learning_rate": 3.3161616161616165e-05,
      "loss": 0.7348,
      "step": 3434
    },
    {
      "epoch": 0.3435,
      "grad_norm": 12.322126388549805,
      "learning_rate": 3.315656565656566e-05,
      "loss": 0.7223,
      "step": 3435
    },
    {
      "epoch": 0.3436,
      "grad_norm": 6.329802513122559,
      "learning_rate": 3.315151515151516e-05,
      "loss": 0.3276,
      "step": 3436
    },
    {
      "epoch": 0.3437,
      "grad_norm": 7.317410945892334,
      "learning_rate": 3.3146464646464646e-05,
      "loss": 0.3881,
      "step": 3437
    },
    {
      "epoch": 0.3438,
      "grad_norm": 12.403636932373047,
      "learning_rate": 3.314141414141415e-05,
      "loss": 0.7456,
      "step": 3438
    },
    {
      "epoch": 0.3439,
      "grad_norm": 11.583650588989258,
      "learning_rate": 3.313636363636364e-05,
      "loss": 1.1284,
      "step": 3439
    },
    {
      "epoch": 0.344,
      "grad_norm": 9.933998107910156,
      "learning_rate": 3.3131313131313134e-05,
      "loss": 0.2333,
      "step": 3440
    },
    {
      "epoch": 0.3441,
      "grad_norm": 8.3192777633667,
      "learning_rate": 3.312626262626263e-05,
      "loss": 0.426,
      "step": 3441
    },
    {
      "epoch": 0.3442,
      "grad_norm": 7.160661697387695,
      "learning_rate": 3.3121212121212126e-05,
      "loss": 0.4193,
      "step": 3442
    },
    {
      "epoch": 0.3443,
      "grad_norm": 9.04338264465332,
      "learning_rate": 3.3116161616161615e-05,
      "loss": 0.3786,
      "step": 3443
    },
    {
      "epoch": 0.3444,
      "grad_norm": 6.9415130615234375,
      "learning_rate": 3.311111111111112e-05,
      "loss": 0.4066,
      "step": 3444
    },
    {
      "epoch": 0.3445,
      "grad_norm": 10.285576820373535,
      "learning_rate": 3.310606060606061e-05,
      "loss": 0.4102,
      "step": 3445
    },
    {
      "epoch": 0.3446,
      "grad_norm": 7.195133686065674,
      "learning_rate": 3.31010101010101e-05,
      "loss": 0.2277,
      "step": 3446
    },
    {
      "epoch": 0.3447,
      "grad_norm": 8.357559204101562,
      "learning_rate": 3.30959595959596e-05,
      "loss": 0.5727,
      "step": 3447
    },
    {
      "epoch": 0.3448,
      "grad_norm": 47.81364822387695,
      "learning_rate": 3.3090909090909095e-05,
      "loss": 0.2553,
      "step": 3448
    },
    {
      "epoch": 0.3449,
      "grad_norm": 11.995875358581543,
      "learning_rate": 3.3085858585858584e-05,
      "loss": 0.5512,
      "step": 3449
    },
    {
      "epoch": 0.345,
      "grad_norm": 9.880575180053711,
      "learning_rate": 3.308080808080809e-05,
      "loss": 0.6705,
      "step": 3450
    },
    {
      "epoch": 0.3451,
      "grad_norm": 22.068737030029297,
      "learning_rate": 3.3075757575757576e-05,
      "loss": 0.6142,
      "step": 3451
    },
    {
      "epoch": 0.3452,
      "grad_norm": 11.228776931762695,
      "learning_rate": 3.307070707070707e-05,
      "loss": 0.3679,
      "step": 3452
    },
    {
      "epoch": 0.3453,
      "grad_norm": 10.544238090515137,
      "learning_rate": 3.306565656565657e-05,
      "loss": 0.3315,
      "step": 3453
    },
    {
      "epoch": 0.3454,
      "grad_norm": 11.531426429748535,
      "learning_rate": 3.3060606060606064e-05,
      "loss": 0.3431,
      "step": 3454
    },
    {
      "epoch": 0.3455,
      "grad_norm": 7.555500507354736,
      "learning_rate": 3.3055555555555553e-05,
      "loss": 0.5185,
      "step": 3455
    },
    {
      "epoch": 0.3456,
      "grad_norm": 13.147711753845215,
      "learning_rate": 3.3050505050505056e-05,
      "loss": 0.5218,
      "step": 3456
    },
    {
      "epoch": 0.3457,
      "grad_norm": 11.949089050292969,
      "learning_rate": 3.3045454545454545e-05,
      "loss": 0.7531,
      "step": 3457
    },
    {
      "epoch": 0.3458,
      "grad_norm": 24.406644821166992,
      "learning_rate": 3.304040404040404e-05,
      "loss": 0.9177,
      "step": 3458
    },
    {
      "epoch": 0.3459,
      "grad_norm": 7.281851291656494,
      "learning_rate": 3.303535353535354e-05,
      "loss": 0.3662,
      "step": 3459
    },
    {
      "epoch": 0.346,
      "grad_norm": 13.295408248901367,
      "learning_rate": 3.303030303030303e-05,
      "loss": 0.5061,
      "step": 3460
    },
    {
      "epoch": 0.3461,
      "grad_norm": 11.674854278564453,
      "learning_rate": 3.302525252525252e-05,
      "loss": 0.4549,
      "step": 3461
    },
    {
      "epoch": 0.3462,
      "grad_norm": 4.731212139129639,
      "learning_rate": 3.3020202020202025e-05,
      "loss": 0.2788,
      "step": 3462
    },
    {
      "epoch": 0.3463,
      "grad_norm": 3.7147693634033203,
      "learning_rate": 3.3015151515151515e-05,
      "loss": 0.1715,
      "step": 3463
    },
    {
      "epoch": 0.3464,
      "grad_norm": 6.622838497161865,
      "learning_rate": 3.301010101010101e-05,
      "loss": 0.1756,
      "step": 3464
    },
    {
      "epoch": 0.3465,
      "grad_norm": 6.065534591674805,
      "learning_rate": 3.3005050505050507e-05,
      "loss": 0.2781,
      "step": 3465
    },
    {
      "epoch": 0.3466,
      "grad_norm": 5.562045097351074,
      "learning_rate": 3.3e-05,
      "loss": 0.217,
      "step": 3466
    },
    {
      "epoch": 0.3467,
      "grad_norm": 16.840784072875977,
      "learning_rate": 3.29949494949495e-05,
      "loss": 0.5469,
      "step": 3467
    },
    {
      "epoch": 0.3468,
      "grad_norm": 9.091641426086426,
      "learning_rate": 3.2989898989898995e-05,
      "loss": 0.587,
      "step": 3468
    },
    {
      "epoch": 0.3469,
      "grad_norm": 20.124740600585938,
      "learning_rate": 3.2984848484848484e-05,
      "loss": 0.2915,
      "step": 3469
    },
    {
      "epoch": 0.347,
      "grad_norm": 6.989038944244385,
      "learning_rate": 3.297979797979798e-05,
      "loss": 0.2029,
      "step": 3470
    },
    {
      "epoch": 0.3471,
      "grad_norm": 3.893014430999756,
      "learning_rate": 3.2974747474747476e-05,
      "loss": 0.1263,
      "step": 3471
    },
    {
      "epoch": 0.3472,
      "grad_norm": 4.068609714508057,
      "learning_rate": 3.296969696969697e-05,
      "loss": 0.1866,
      "step": 3472
    },
    {
      "epoch": 0.3473,
      "grad_norm": 7.805154323577881,
      "learning_rate": 3.296464646464647e-05,
      "loss": 0.2427,
      "step": 3473
    },
    {
      "epoch": 0.3474,
      "grad_norm": 9.56243896484375,
      "learning_rate": 3.2959595959595964e-05,
      "loss": 0.2842,
      "step": 3474
    },
    {
      "epoch": 0.3475,
      "grad_norm": 8.433920860290527,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.4471,
      "step": 3475
    },
    {
      "epoch": 0.3476,
      "grad_norm": 9.392358779907227,
      "learning_rate": 3.2949494949494956e-05,
      "loss": 0.4219,
      "step": 3476
    },
    {
      "epoch": 0.3477,
      "grad_norm": 3.374540090560913,
      "learning_rate": 3.2944444444444445e-05,
      "loss": 0.1971,
      "step": 3477
    },
    {
      "epoch": 0.3478,
      "grad_norm": 45.81094741821289,
      "learning_rate": 3.293939393939394e-05,
      "loss": 0.2301,
      "step": 3478
    },
    {
      "epoch": 0.3479,
      "grad_norm": 7.917216777801514,
      "learning_rate": 3.293434343434344e-05,
      "loss": 0.3789,
      "step": 3479
    },
    {
      "epoch": 0.348,
      "grad_norm": 8.881930351257324,
      "learning_rate": 3.292929292929293e-05,
      "loss": 0.3423,
      "step": 3480
    },
    {
      "epoch": 0.3481,
      "grad_norm": 5.645875453948975,
      "learning_rate": 3.292424242424242e-05,
      "loss": 0.2374,
      "step": 3481
    },
    {
      "epoch": 0.3482,
      "grad_norm": 6.062918186187744,
      "learning_rate": 3.2919191919191925e-05,
      "loss": 0.4231,
      "step": 3482
    },
    {
      "epoch": 0.3483,
      "grad_norm": 9.291180610656738,
      "learning_rate": 3.2914141414141414e-05,
      "loss": 0.3394,
      "step": 3483
    },
    {
      "epoch": 0.3484,
      "grad_norm": 9.224453926086426,
      "learning_rate": 3.290909090909091e-05,
      "loss": 0.6024,
      "step": 3484
    },
    {
      "epoch": 0.3485,
      "grad_norm": 12.767841339111328,
      "learning_rate": 3.2904040404040406e-05,
      "loss": 0.3935,
      "step": 3485
    },
    {
      "epoch": 0.3486,
      "grad_norm": 7.192336082458496,
      "learning_rate": 3.28989898989899e-05,
      "loss": 0.3845,
      "step": 3486
    },
    {
      "epoch": 0.3487,
      "grad_norm": 5.554691314697266,
      "learning_rate": 3.289393939393939e-05,
      "loss": 0.3211,
      "step": 3487
    },
    {
      "epoch": 0.3488,
      "grad_norm": 6.912847518920898,
      "learning_rate": 3.2888888888888894e-05,
      "loss": 0.2601,
      "step": 3488
    },
    {
      "epoch": 0.3489,
      "grad_norm": 8.505126953125,
      "learning_rate": 3.288383838383838e-05,
      "loss": 0.5355,
      "step": 3489
    },
    {
      "epoch": 0.349,
      "grad_norm": 5.280939102172852,
      "learning_rate": 3.287878787878788e-05,
      "loss": 0.2831,
      "step": 3490
    },
    {
      "epoch": 0.3491,
      "grad_norm": 5.887515544891357,
      "learning_rate": 3.2873737373737375e-05,
      "loss": 0.2711,
      "step": 3491
    },
    {
      "epoch": 0.3492,
      "grad_norm": 11.228851318359375,
      "learning_rate": 3.286868686868687e-05,
      "loss": 0.9917,
      "step": 3492
    },
    {
      "epoch": 0.3493,
      "grad_norm": 6.535587787628174,
      "learning_rate": 3.286363636363636e-05,
      "loss": 0.2817,
      "step": 3493
    },
    {
      "epoch": 0.3494,
      "grad_norm": 6.024842739105225,
      "learning_rate": 3.285858585858586e-05,
      "loss": 0.1947,
      "step": 3494
    },
    {
      "epoch": 0.3495,
      "grad_norm": 5.438609600067139,
      "learning_rate": 3.285353535353535e-05,
      "loss": 0.3159,
      "step": 3495
    },
    {
      "epoch": 0.3496,
      "grad_norm": 8.021937370300293,
      "learning_rate": 3.284848484848485e-05,
      "loss": 0.4316,
      "step": 3496
    },
    {
      "epoch": 0.3497,
      "grad_norm": 6.461270332336426,
      "learning_rate": 3.2843434343434344e-05,
      "loss": 0.4111,
      "step": 3497
    },
    {
      "epoch": 0.3498,
      "grad_norm": 12.634438514709473,
      "learning_rate": 3.283838383838384e-05,
      "loss": 0.7213,
      "step": 3498
    },
    {
      "epoch": 0.3499,
      "grad_norm": 9.399948120117188,
      "learning_rate": 3.283333333333333e-05,
      "loss": 0.2698,
      "step": 3499
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.119591236114502,
      "learning_rate": 3.282828282828283e-05,
      "loss": 0.2932,
      "step": 3500
    },
    {
      "epoch": 0.3501,
      "grad_norm": 8.762221336364746,
      "learning_rate": 3.282323232323232e-05,
      "loss": 0.4345,
      "step": 3501
    },
    {
      "epoch": 0.3502,
      "grad_norm": 8.68287467956543,
      "learning_rate": 3.281818181818182e-05,
      "loss": 0.1867,
      "step": 3502
    },
    {
      "epoch": 0.3503,
      "grad_norm": 24.314455032348633,
      "learning_rate": 3.2813131313131314e-05,
      "loss": 0.5838,
      "step": 3503
    },
    {
      "epoch": 0.3504,
      "grad_norm": 5.324301719665527,
      "learning_rate": 3.280808080808081e-05,
      "loss": 0.3469,
      "step": 3504
    },
    {
      "epoch": 0.3505,
      "grad_norm": 8.807413101196289,
      "learning_rate": 3.2803030303030305e-05,
      "loss": 0.2825,
      "step": 3505
    },
    {
      "epoch": 0.3506,
      "grad_norm": 9.133069038391113,
      "learning_rate": 3.27979797979798e-05,
      "loss": 0.3398,
      "step": 3506
    },
    {
      "epoch": 0.3507,
      "grad_norm": 9.489143371582031,
      "learning_rate": 3.279292929292929e-05,
      "loss": 0.3194,
      "step": 3507
    },
    {
      "epoch": 0.3508,
      "grad_norm": 5.439024448394775,
      "learning_rate": 3.278787878787879e-05,
      "loss": 0.3129,
      "step": 3508
    },
    {
      "epoch": 0.3509,
      "grad_norm": 5.873137474060059,
      "learning_rate": 3.278282828282828e-05,
      "loss": 0.2353,
      "step": 3509
    },
    {
      "epoch": 0.351,
      "grad_norm": 6.945653915405273,
      "learning_rate": 3.277777777777778e-05,
      "loss": 0.1545,
      "step": 3510
    },
    {
      "epoch": 0.3511,
      "grad_norm": 3.570629358291626,
      "learning_rate": 3.2772727272727275e-05,
      "loss": 0.1924,
      "step": 3511
    },
    {
      "epoch": 0.3512,
      "grad_norm": 6.525230884552002,
      "learning_rate": 3.276767676767677e-05,
      "loss": 0.2707,
      "step": 3512
    },
    {
      "epoch": 0.3513,
      "grad_norm": 10.396090507507324,
      "learning_rate": 3.276262626262626e-05,
      "loss": 0.4665,
      "step": 3513
    },
    {
      "epoch": 0.3514,
      "grad_norm": 24.545915603637695,
      "learning_rate": 3.2757575757575756e-05,
      "loss": 0.4835,
      "step": 3514
    },
    {
      "epoch": 0.3515,
      "grad_norm": 13.859746932983398,
      "learning_rate": 3.275252525252526e-05,
      "loss": 0.9085,
      "step": 3515
    },
    {
      "epoch": 0.3516,
      "grad_norm": 7.199886322021484,
      "learning_rate": 3.274747474747475e-05,
      "loss": 0.239,
      "step": 3516
    },
    {
      "epoch": 0.3517,
      "grad_norm": 11.748091697692871,
      "learning_rate": 3.2742424242424244e-05,
      "loss": 0.3888,
      "step": 3517
    },
    {
      "epoch": 0.3518,
      "grad_norm": 4.13361120223999,
      "learning_rate": 3.273737373737374e-05,
      "loss": 0.1868,
      "step": 3518
    },
    {
      "epoch": 0.3519,
      "grad_norm": 10.276573181152344,
      "learning_rate": 3.2732323232323236e-05,
      "loss": 0.3459,
      "step": 3519
    },
    {
      "epoch": 0.352,
      "grad_norm": 5.79747200012207,
      "learning_rate": 3.272727272727273e-05,
      "loss": 0.2952,
      "step": 3520
    },
    {
      "epoch": 0.3521,
      "grad_norm": 5.689869403839111,
      "learning_rate": 3.272222222222223e-05,
      "loss": 0.3311,
      "step": 3521
    },
    {
      "epoch": 0.3522,
      "grad_norm": 69.75318145751953,
      "learning_rate": 3.271717171717172e-05,
      "loss": 0.2579,
      "step": 3522
    },
    {
      "epoch": 0.3523,
      "grad_norm": 6.842641353607178,
      "learning_rate": 3.271212121212121e-05,
      "loss": 0.4314,
      "step": 3523
    },
    {
      "epoch": 0.3524,
      "grad_norm": 9.777610778808594,
      "learning_rate": 3.270707070707071e-05,
      "loss": 0.7316,
      "step": 3524
    },
    {
      "epoch": 0.3525,
      "grad_norm": 4.68624210357666,
      "learning_rate": 3.2702020202020205e-05,
      "loss": 0.2496,
      "step": 3525
    },
    {
      "epoch": 0.3526,
      "grad_norm": 6.869032382965088,
      "learning_rate": 3.26969696969697e-05,
      "loss": 0.3173,
      "step": 3526
    },
    {
      "epoch": 0.3527,
      "grad_norm": 5.9927449226379395,
      "learning_rate": 3.26919191919192e-05,
      "loss": 0.2442,
      "step": 3527
    },
    {
      "epoch": 0.3528,
      "grad_norm": 6.10164213180542,
      "learning_rate": 3.2686868686868686e-05,
      "loss": 0.3348,
      "step": 3528
    },
    {
      "epoch": 0.3529,
      "grad_norm": 7.009406566619873,
      "learning_rate": 3.268181818181819e-05,
      "loss": 0.3075,
      "step": 3529
    },
    {
      "epoch": 0.353,
      "grad_norm": 21.434419631958008,
      "learning_rate": 3.267676767676768e-05,
      "loss": 0.852,
      "step": 3530
    },
    {
      "epoch": 0.3531,
      "grad_norm": 15.160115242004395,
      "learning_rate": 3.2671717171717174e-05,
      "loss": 0.7047,
      "step": 3531
    },
    {
      "epoch": 0.3532,
      "grad_norm": 5.2100725173950195,
      "learning_rate": 3.266666666666667e-05,
      "loss": 0.3143,
      "step": 3532
    },
    {
      "epoch": 0.3533,
      "grad_norm": 6.302238941192627,
      "learning_rate": 3.2661616161616166e-05,
      "loss": 0.2395,
      "step": 3533
    },
    {
      "epoch": 0.3534,
      "grad_norm": 12.329451560974121,
      "learning_rate": 3.2656565656565655e-05,
      "loss": 0.3906,
      "step": 3534
    },
    {
      "epoch": 0.3535,
      "grad_norm": 20.560916900634766,
      "learning_rate": 3.265151515151516e-05,
      "loss": 0.3437,
      "step": 3535
    },
    {
      "epoch": 0.3536,
      "grad_norm": 10.862360954284668,
      "learning_rate": 3.264646464646465e-05,
      "loss": 0.4563,
      "step": 3536
    },
    {
      "epoch": 0.3537,
      "grad_norm": 11.921815872192383,
      "learning_rate": 3.264141414141414e-05,
      "loss": 0.5244,
      "step": 3537
    },
    {
      "epoch": 0.3538,
      "grad_norm": 7.67103910446167,
      "learning_rate": 3.263636363636364e-05,
      "loss": 0.363,
      "step": 3538
    },
    {
      "epoch": 0.3539,
      "grad_norm": 11.97107219696045,
      "learning_rate": 3.2631313131313135e-05,
      "loss": 0.4326,
      "step": 3539
    },
    {
      "epoch": 0.354,
      "grad_norm": 11.358877182006836,
      "learning_rate": 3.2626262626262624e-05,
      "loss": 0.3868,
      "step": 3540
    },
    {
      "epoch": 0.3541,
      "grad_norm": 18.851232528686523,
      "learning_rate": 3.262121212121213e-05,
      "loss": 0.5953,
      "step": 3541
    },
    {
      "epoch": 0.3542,
      "grad_norm": 6.475265979766846,
      "learning_rate": 3.2616161616161616e-05,
      "loss": 0.3456,
      "step": 3542
    },
    {
      "epoch": 0.3543,
      "grad_norm": 10.833802223205566,
      "learning_rate": 3.261111111111111e-05,
      "loss": 0.1585,
      "step": 3543
    },
    {
      "epoch": 0.3544,
      "grad_norm": 5.348949909210205,
      "learning_rate": 3.260606060606061e-05,
      "loss": 0.2633,
      "step": 3544
    },
    {
      "epoch": 0.3545,
      "grad_norm": 5.327052593231201,
      "learning_rate": 3.2601010101010104e-05,
      "loss": 0.3077,
      "step": 3545
    },
    {
      "epoch": 0.3546,
      "grad_norm": 9.638204574584961,
      "learning_rate": 3.2595959595959594e-05,
      "loss": 0.3,
      "step": 3546
    },
    {
      "epoch": 0.3547,
      "grad_norm": 15.40208625793457,
      "learning_rate": 3.2590909090909096e-05,
      "loss": 0.4824,
      "step": 3547
    },
    {
      "epoch": 0.3548,
      "grad_norm": 6.033774375915527,
      "learning_rate": 3.2585858585858586e-05,
      "loss": 0.2113,
      "step": 3548
    },
    {
      "epoch": 0.3549,
      "grad_norm": 14.04631519317627,
      "learning_rate": 3.258080808080808e-05,
      "loss": 0.2178,
      "step": 3549
    },
    {
      "epoch": 0.355,
      "grad_norm": 15.686373710632324,
      "learning_rate": 3.257575757575758e-05,
      "loss": 0.2835,
      "step": 3550
    },
    {
      "epoch": 0.3551,
      "grad_norm": 7.196442127227783,
      "learning_rate": 3.2570707070707074e-05,
      "loss": 0.4337,
      "step": 3551
    },
    {
      "epoch": 0.3552,
      "grad_norm": 6.433614253997803,
      "learning_rate": 3.256565656565656e-05,
      "loss": 0.201,
      "step": 3552
    },
    {
      "epoch": 0.3553,
      "grad_norm": 7.2628703117370605,
      "learning_rate": 3.2560606060606066e-05,
      "loss": 0.368,
      "step": 3553
    },
    {
      "epoch": 0.3554,
      "grad_norm": 5.379080772399902,
      "learning_rate": 3.2555555555555555e-05,
      "loss": 0.254,
      "step": 3554
    },
    {
      "epoch": 0.3555,
      "grad_norm": 11.146819114685059,
      "learning_rate": 3.255050505050505e-05,
      "loss": 0.4253,
      "step": 3555
    },
    {
      "epoch": 0.3556,
      "grad_norm": 9.27754020690918,
      "learning_rate": 3.254545454545455e-05,
      "loss": 0.3199,
      "step": 3556
    },
    {
      "epoch": 0.3557,
      "grad_norm": 8.756689071655273,
      "learning_rate": 3.254040404040404e-05,
      "loss": 0.3445,
      "step": 3557
    },
    {
      "epoch": 0.3558,
      "grad_norm": 7.644088268280029,
      "learning_rate": 3.253535353535354e-05,
      "loss": 0.2939,
      "step": 3558
    },
    {
      "epoch": 0.3559,
      "grad_norm": 22.366262435913086,
      "learning_rate": 3.2530303030303035e-05,
      "loss": 0.3751,
      "step": 3559
    },
    {
      "epoch": 0.356,
      "grad_norm": 4.409122467041016,
      "learning_rate": 3.2525252525252524e-05,
      "loss": 0.2559,
      "step": 3560
    },
    {
      "epoch": 0.3561,
      "grad_norm": 9.719507217407227,
      "learning_rate": 3.252020202020202e-05,
      "loss": 0.4201,
      "step": 3561
    },
    {
      "epoch": 0.3562,
      "grad_norm": 19.541728973388672,
      "learning_rate": 3.2515151515151516e-05,
      "loss": 0.4721,
      "step": 3562
    },
    {
      "epoch": 0.3563,
      "grad_norm": 19.052942276000977,
      "learning_rate": 3.251010101010101e-05,
      "loss": 0.3279,
      "step": 3563
    },
    {
      "epoch": 0.3564,
      "grad_norm": 6.407190322875977,
      "learning_rate": 3.250505050505051e-05,
      "loss": 0.1523,
      "step": 3564
    },
    {
      "epoch": 0.3565,
      "grad_norm": 4.426276683807373,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.2057,
      "step": 3565
    },
    {
      "epoch": 0.3566,
      "grad_norm": 10.768468856811523,
      "learning_rate": 3.249494949494949e-05,
      "loss": 0.4224,
      "step": 3566
    },
    {
      "epoch": 0.3567,
      "grad_norm": 8.945962905883789,
      "learning_rate": 3.248989898989899e-05,
      "loss": 0.4314,
      "step": 3567
    },
    {
      "epoch": 0.3568,
      "grad_norm": 16.455806732177734,
      "learning_rate": 3.2484848484848485e-05,
      "loss": 0.3386,
      "step": 3568
    },
    {
      "epoch": 0.3569,
      "grad_norm": 7.884345531463623,
      "learning_rate": 3.247979797979798e-05,
      "loss": 0.4053,
      "step": 3569
    },
    {
      "epoch": 0.357,
      "grad_norm": 8.637998580932617,
      "learning_rate": 3.247474747474748e-05,
      "loss": 0.5463,
      "step": 3570
    },
    {
      "epoch": 0.3571,
      "grad_norm": 13.469834327697754,
      "learning_rate": 3.246969696969697e-05,
      "loss": 0.6633,
      "step": 3571
    },
    {
      "epoch": 0.3572,
      "grad_norm": 11.51959228515625,
      "learning_rate": 3.246464646464646e-05,
      "loss": 0.5247,
      "step": 3572
    },
    {
      "epoch": 0.3573,
      "grad_norm": 9.419876098632812,
      "learning_rate": 3.2459595959595965e-05,
      "loss": 0.3557,
      "step": 3573
    },
    {
      "epoch": 0.3574,
      "grad_norm": 16.57023048400879,
      "learning_rate": 3.2454545454545454e-05,
      "loss": 0.1893,
      "step": 3574
    },
    {
      "epoch": 0.3575,
      "grad_norm": 8.808079719543457,
      "learning_rate": 3.244949494949495e-05,
      "loss": 0.2904,
      "step": 3575
    },
    {
      "epoch": 0.3576,
      "grad_norm": 6.445556163787842,
      "learning_rate": 3.2444444444444446e-05,
      "loss": 0.2833,
      "step": 3576
    },
    {
      "epoch": 0.3577,
      "grad_norm": 4.896984100341797,
      "learning_rate": 3.243939393939394e-05,
      "loss": 0.2302,
      "step": 3577
    },
    {
      "epoch": 0.3578,
      "grad_norm": 4.945743083953857,
      "learning_rate": 3.243434343434343e-05,
      "loss": 0.2125,
      "step": 3578
    },
    {
      "epoch": 0.3579,
      "grad_norm": 9.054647445678711,
      "learning_rate": 3.2429292929292934e-05,
      "loss": 0.2919,
      "step": 3579
    },
    {
      "epoch": 0.358,
      "grad_norm": 8.592548370361328,
      "learning_rate": 3.2424242424242423e-05,
      "loss": 0.3232,
      "step": 3580
    },
    {
      "epoch": 0.3581,
      "grad_norm": 5.094370365142822,
      "learning_rate": 3.241919191919192e-05,
      "loss": 0.2531,
      "step": 3581
    },
    {
      "epoch": 0.3582,
      "grad_norm": 5.315758228302002,
      "learning_rate": 3.2414141414141415e-05,
      "loss": 0.1553,
      "step": 3582
    },
    {
      "epoch": 0.3583,
      "grad_norm": 8.435925483703613,
      "learning_rate": 3.240909090909091e-05,
      "loss": 0.2201,
      "step": 3583
    },
    {
      "epoch": 0.3584,
      "grad_norm": 5.6575751304626465,
      "learning_rate": 3.24040404040404e-05,
      "loss": 0.2151,
      "step": 3584
    },
    {
      "epoch": 0.3585,
      "grad_norm": 4.224269866943359,
      "learning_rate": 3.2398989898989903e-05,
      "loss": 0.2221,
      "step": 3585
    },
    {
      "epoch": 0.3586,
      "grad_norm": 5.459326267242432,
      "learning_rate": 3.239393939393939e-05,
      "loss": 0.2076,
      "step": 3586
    },
    {
      "epoch": 0.3587,
      "grad_norm": 7.233055114746094,
      "learning_rate": 3.238888888888889e-05,
      "loss": 0.1713,
      "step": 3587
    },
    {
      "epoch": 0.3588,
      "grad_norm": 4.845981121063232,
      "learning_rate": 3.2383838383838385e-05,
      "loss": 0.2535,
      "step": 3588
    },
    {
      "epoch": 0.3589,
      "grad_norm": 8.419696807861328,
      "learning_rate": 3.237878787878788e-05,
      "loss": 0.316,
      "step": 3589
    },
    {
      "epoch": 0.359,
      "grad_norm": 9.71530818939209,
      "learning_rate": 3.237373737373737e-05,
      "loss": 0.2316,
      "step": 3590
    },
    {
      "epoch": 0.3591,
      "grad_norm": 4.37883996963501,
      "learning_rate": 3.236868686868687e-05,
      "loss": 0.2492,
      "step": 3591
    },
    {
      "epoch": 0.3592,
      "grad_norm": 3.8667333126068115,
      "learning_rate": 3.236363636363636e-05,
      "loss": 0.2466,
      "step": 3592
    },
    {
      "epoch": 0.3593,
      "grad_norm": 10.293268203735352,
      "learning_rate": 3.235858585858586e-05,
      "loss": 0.4927,
      "step": 3593
    },
    {
      "epoch": 0.3594,
      "grad_norm": 7.353858947753906,
      "learning_rate": 3.2353535353535354e-05,
      "loss": 0.3879,
      "step": 3594
    },
    {
      "epoch": 0.3595,
      "grad_norm": 6.833389759063721,
      "learning_rate": 3.234848484848485e-05,
      "loss": 0.2961,
      "step": 3595
    },
    {
      "epoch": 0.3596,
      "grad_norm": 8.473453521728516,
      "learning_rate": 3.234343434343434e-05,
      "loss": 0.3682,
      "step": 3596
    },
    {
      "epoch": 0.3597,
      "grad_norm": 8.985250473022461,
      "learning_rate": 3.233838383838384e-05,
      "loss": 0.4515,
      "step": 3597
    },
    {
      "epoch": 0.3598,
      "grad_norm": 11.717499732971191,
      "learning_rate": 3.233333333333333e-05,
      "loss": 0.3091,
      "step": 3598
    },
    {
      "epoch": 0.3599,
      "grad_norm": 16.168987274169922,
      "learning_rate": 3.232828282828283e-05,
      "loss": 0.3228,
      "step": 3599
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.006065368652344,
      "learning_rate": 3.232323232323233e-05,
      "loss": 0.2665,
      "step": 3600
    },
    {
      "epoch": 0.3601,
      "grad_norm": 6.179235458374023,
      "learning_rate": 3.231818181818182e-05,
      "loss": 0.3539,
      "step": 3601
    },
    {
      "epoch": 0.3602,
      "grad_norm": 7.853567600250244,
      "learning_rate": 3.2313131313131315e-05,
      "loss": 0.5002,
      "step": 3602
    },
    {
      "epoch": 0.3603,
      "grad_norm": 9.82901382446289,
      "learning_rate": 3.230808080808081e-05,
      "loss": 0.3408,
      "step": 3603
    },
    {
      "epoch": 0.3604,
      "grad_norm": 4.88955020904541,
      "learning_rate": 3.230303030303031e-05,
      "loss": 0.2089,
      "step": 3604
    },
    {
      "epoch": 0.3605,
      "grad_norm": 6.467957019805908,
      "learning_rate": 3.2297979797979796e-05,
      "loss": 0.5122,
      "step": 3605
    },
    {
      "epoch": 0.3606,
      "grad_norm": 4.839207649230957,
      "learning_rate": 3.22929292929293e-05,
      "loss": 0.3369,
      "step": 3606
    },
    {
      "epoch": 0.3607,
      "grad_norm": 5.150304794311523,
      "learning_rate": 3.228787878787879e-05,
      "loss": 0.3152,
      "step": 3607
    },
    {
      "epoch": 0.3608,
      "grad_norm": 3.992298126220703,
      "learning_rate": 3.2282828282828284e-05,
      "loss": 0.213,
      "step": 3608
    },
    {
      "epoch": 0.3609,
      "grad_norm": 3.6785831451416016,
      "learning_rate": 3.227777777777778e-05,
      "loss": 0.2668,
      "step": 3609
    },
    {
      "epoch": 0.361,
      "grad_norm": 17.68304443359375,
      "learning_rate": 3.2272727272727276e-05,
      "loss": 0.1979,
      "step": 3610
    },
    {
      "epoch": 0.3611,
      "grad_norm": 3.767380952835083,
      "learning_rate": 3.226767676767677e-05,
      "loss": 0.1208,
      "step": 3611
    },
    {
      "epoch": 0.3612,
      "grad_norm": 9.040901184082031,
      "learning_rate": 3.226262626262627e-05,
      "loss": 0.2566,
      "step": 3612
    },
    {
      "epoch": 0.3613,
      "grad_norm": 8.327820777893066,
      "learning_rate": 3.225757575757576e-05,
      "loss": 0.2261,
      "step": 3613
    },
    {
      "epoch": 0.3614,
      "grad_norm": 8.158108711242676,
      "learning_rate": 3.225252525252525e-05,
      "loss": 0.2305,
      "step": 3614
    },
    {
      "epoch": 0.3615,
      "grad_norm": 19.10566520690918,
      "learning_rate": 3.224747474747475e-05,
      "loss": 0.2978,
      "step": 3615
    },
    {
      "epoch": 0.3616,
      "grad_norm": 6.178238868713379,
      "learning_rate": 3.2242424242424245e-05,
      "loss": 0.2755,
      "step": 3616
    },
    {
      "epoch": 0.3617,
      "grad_norm": 9.650116920471191,
      "learning_rate": 3.223737373737374e-05,
      "loss": 0.3279,
      "step": 3617
    },
    {
      "epoch": 0.3618,
      "grad_norm": 7.644715785980225,
      "learning_rate": 3.223232323232324e-05,
      "loss": 0.2956,
      "step": 3618
    },
    {
      "epoch": 0.3619,
      "grad_norm": 6.400201320648193,
      "learning_rate": 3.2227272727272726e-05,
      "loss": 0.3172,
      "step": 3619
    },
    {
      "epoch": 0.362,
      "grad_norm": 7.554383754730225,
      "learning_rate": 3.222222222222223e-05,
      "loss": 0.2875,
      "step": 3620
    },
    {
      "epoch": 0.3621,
      "grad_norm": 6.053802490234375,
      "learning_rate": 3.221717171717172e-05,
      "loss": 0.2888,
      "step": 3621
    },
    {
      "epoch": 0.3622,
      "grad_norm": 7.208088397979736,
      "learning_rate": 3.2212121212121214e-05,
      "loss": 0.3282,
      "step": 3622
    },
    {
      "epoch": 0.3623,
      "grad_norm": 5.005105018615723,
      "learning_rate": 3.220707070707071e-05,
      "loss": 0.2316,
      "step": 3623
    },
    {
      "epoch": 0.3624,
      "grad_norm": 9.341145515441895,
      "learning_rate": 3.2202020202020206e-05,
      "loss": 0.2776,
      "step": 3624
    },
    {
      "epoch": 0.3625,
      "grad_norm": 8.482481002807617,
      "learning_rate": 3.2196969696969696e-05,
      "loss": 0.2689,
      "step": 3625
    },
    {
      "epoch": 0.3626,
      "grad_norm": 6.663022518157959,
      "learning_rate": 3.21919191919192e-05,
      "loss": 0.2876,
      "step": 3626
    },
    {
      "epoch": 0.3627,
      "grad_norm": 16.021154403686523,
      "learning_rate": 3.218686868686869e-05,
      "loss": 0.2929,
      "step": 3627
    },
    {
      "epoch": 0.3628,
      "grad_norm": 6.424506187438965,
      "learning_rate": 3.2181818181818184e-05,
      "loss": 0.3612,
      "step": 3628
    },
    {
      "epoch": 0.3629,
      "grad_norm": 6.290036201477051,
      "learning_rate": 3.217676767676768e-05,
      "loss": 0.2419,
      "step": 3629
    },
    {
      "epoch": 0.363,
      "grad_norm": 5.759331703186035,
      "learning_rate": 3.2171717171717176e-05,
      "loss": 0.2122,
      "step": 3630
    },
    {
      "epoch": 0.3631,
      "grad_norm": 5.318015098571777,
      "learning_rate": 3.2166666666666665e-05,
      "loss": 0.2817,
      "step": 3631
    },
    {
      "epoch": 0.3632,
      "grad_norm": 10.156153678894043,
      "learning_rate": 3.216161616161617e-05,
      "loss": 0.4378,
      "step": 3632
    },
    {
      "epoch": 0.3633,
      "grad_norm": 52.87786865234375,
      "learning_rate": 3.215656565656566e-05,
      "loss": 0.2382,
      "step": 3633
    },
    {
      "epoch": 0.3634,
      "grad_norm": 15.25259780883789,
      "learning_rate": 3.215151515151515e-05,
      "loss": 0.3619,
      "step": 3634
    },
    {
      "epoch": 0.3635,
      "grad_norm": 14.668888092041016,
      "learning_rate": 3.214646464646465e-05,
      "loss": 0.4734,
      "step": 3635
    },
    {
      "epoch": 0.3636,
      "grad_norm": 5.72560453414917,
      "learning_rate": 3.2141414141414145e-05,
      "loss": 0.2549,
      "step": 3636
    },
    {
      "epoch": 0.3637,
      "grad_norm": 9.549607276916504,
      "learning_rate": 3.2136363636363634e-05,
      "loss": 0.3121,
      "step": 3637
    },
    {
      "epoch": 0.3638,
      "grad_norm": 19.792192459106445,
      "learning_rate": 3.213131313131314e-05,
      "loss": 0.7896,
      "step": 3638
    },
    {
      "epoch": 0.3639,
      "grad_norm": 8.6248779296875,
      "learning_rate": 3.2126262626262626e-05,
      "loss": 0.604,
      "step": 3639
    },
    {
      "epoch": 0.364,
      "grad_norm": 20.829710006713867,
      "learning_rate": 3.212121212121212e-05,
      "loss": 0.4374,
      "step": 3640
    },
    {
      "epoch": 0.3641,
      "grad_norm": 19.585166931152344,
      "learning_rate": 3.211616161616162e-05,
      "loss": 0.4402,
      "step": 3641
    },
    {
      "epoch": 0.3642,
      "grad_norm": 7.477211952209473,
      "learning_rate": 3.2111111111111114e-05,
      "loss": 0.3192,
      "step": 3642
    },
    {
      "epoch": 0.3643,
      "grad_norm": 14.036511421203613,
      "learning_rate": 3.21060606060606e-05,
      "loss": 0.3004,
      "step": 3643
    },
    {
      "epoch": 0.3644,
      "grad_norm": 5.941062927246094,
      "learning_rate": 3.2101010101010106e-05,
      "loss": 0.2772,
      "step": 3644
    },
    {
      "epoch": 0.3645,
      "grad_norm": 5.327350616455078,
      "learning_rate": 3.2095959595959595e-05,
      "loss": 0.3681,
      "step": 3645
    },
    {
      "epoch": 0.3646,
      "grad_norm": 8.54353141784668,
      "learning_rate": 3.209090909090909e-05,
      "loss": 0.3013,
      "step": 3646
    },
    {
      "epoch": 0.3647,
      "grad_norm": 6.942908763885498,
      "learning_rate": 3.208585858585859e-05,
      "loss": 0.2489,
      "step": 3647
    },
    {
      "epoch": 0.3648,
      "grad_norm": 5.576939582824707,
      "learning_rate": 3.208080808080808e-05,
      "loss": 0.3081,
      "step": 3648
    },
    {
      "epoch": 0.3649,
      "grad_norm": 6.8575873374938965,
      "learning_rate": 3.207575757575758e-05,
      "loss": 0.3115,
      "step": 3649
    },
    {
      "epoch": 0.365,
      "grad_norm": 7.1283159255981445,
      "learning_rate": 3.2070707070707075e-05,
      "loss": 0.3257,
      "step": 3650
    },
    {
      "epoch": 0.3651,
      "grad_norm": 13.150546073913574,
      "learning_rate": 3.2065656565656564e-05,
      "loss": 0.2812,
      "step": 3651
    },
    {
      "epoch": 0.3652,
      "grad_norm": 6.142471790313721,
      "learning_rate": 3.206060606060606e-05,
      "loss": 0.338,
      "step": 3652
    },
    {
      "epoch": 0.3653,
      "grad_norm": 9.727821350097656,
      "learning_rate": 3.2055555555555556e-05,
      "loss": 0.324,
      "step": 3653
    },
    {
      "epoch": 0.3654,
      "grad_norm": 6.325886249542236,
      "learning_rate": 3.205050505050505e-05,
      "loss": 0.2127,
      "step": 3654
    },
    {
      "epoch": 0.3655,
      "grad_norm": 11.271162033081055,
      "learning_rate": 3.204545454545455e-05,
      "loss": 0.3549,
      "step": 3655
    },
    {
      "epoch": 0.3656,
      "grad_norm": 20.631624221801758,
      "learning_rate": 3.2040404040404044e-05,
      "loss": 0.2942,
      "step": 3656
    },
    {
      "epoch": 0.3657,
      "grad_norm": 6.748211860656738,
      "learning_rate": 3.2035353535353533e-05,
      "loss": 0.3421,
      "step": 3657
    },
    {
      "epoch": 0.3658,
      "grad_norm": 5.722880840301514,
      "learning_rate": 3.203030303030303e-05,
      "loss": 0.3931,
      "step": 3658
    },
    {
      "epoch": 0.3659,
      "grad_norm": 16.514915466308594,
      "learning_rate": 3.2025252525252525e-05,
      "loss": 0.404,
      "step": 3659
    },
    {
      "epoch": 0.366,
      "grad_norm": 5.053277492523193,
      "learning_rate": 3.202020202020202e-05,
      "loss": 0.256,
      "step": 3660
    },
    {
      "epoch": 0.3661,
      "grad_norm": 7.194996356964111,
      "learning_rate": 3.201515151515152e-05,
      "loss": 0.4369,
      "step": 3661
    },
    {
      "epoch": 0.3662,
      "grad_norm": 5.80418586730957,
      "learning_rate": 3.201010101010101e-05,
      "loss": 0.2933,
      "step": 3662
    },
    {
      "epoch": 0.3663,
      "grad_norm": 5.247701644897461,
      "learning_rate": 3.20050505050505e-05,
      "loss": 0.42,
      "step": 3663
    },
    {
      "epoch": 0.3664,
      "grad_norm": 4.9518351554870605,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.3305,
      "step": 3664
    },
    {
      "epoch": 0.3665,
      "grad_norm": 7.3300886154174805,
      "learning_rate": 3.1994949494949495e-05,
      "loss": 0.3048,
      "step": 3665
    },
    {
      "epoch": 0.3666,
      "grad_norm": 5.662550926208496,
      "learning_rate": 3.198989898989899e-05,
      "loss": 0.5079,
      "step": 3666
    },
    {
      "epoch": 0.3667,
      "grad_norm": 9.213573455810547,
      "learning_rate": 3.1984848484848487e-05,
      "loss": 0.3973,
      "step": 3667
    },
    {
      "epoch": 0.3668,
      "grad_norm": 7.285808086395264,
      "learning_rate": 3.197979797979798e-05,
      "loss": 0.4196,
      "step": 3668
    },
    {
      "epoch": 0.3669,
      "grad_norm": 4.903191566467285,
      "learning_rate": 3.197474747474747e-05,
      "loss": 0.5337,
      "step": 3669
    },
    {
      "epoch": 0.367,
      "grad_norm": 6.977812767028809,
      "learning_rate": 3.1969696969696974e-05,
      "loss": 0.4158,
      "step": 3670
    },
    {
      "epoch": 0.3671,
      "grad_norm": 6.705687046051025,
      "learning_rate": 3.1964646464646464e-05,
      "loss": 0.3983,
      "step": 3671
    },
    {
      "epoch": 0.3672,
      "grad_norm": 10.14280891418457,
      "learning_rate": 3.195959595959596e-05,
      "loss": 0.4174,
      "step": 3672
    },
    {
      "epoch": 0.3673,
      "grad_norm": 15.340149879455566,
      "learning_rate": 3.1954545454545456e-05,
      "loss": 0.4074,
      "step": 3673
    },
    {
      "epoch": 0.3674,
      "grad_norm": 13.192736625671387,
      "learning_rate": 3.194949494949495e-05,
      "loss": 0.4116,
      "step": 3674
    },
    {
      "epoch": 0.3675,
      "grad_norm": 15.862801551818848,
      "learning_rate": 3.194444444444444e-05,
      "loss": 0.4322,
      "step": 3675
    },
    {
      "epoch": 0.3676,
      "grad_norm": 11.078970909118652,
      "learning_rate": 3.1939393939393944e-05,
      "loss": 0.4626,
      "step": 3676
    },
    {
      "epoch": 0.3677,
      "grad_norm": 5.5501580238342285,
      "learning_rate": 3.193434343434343e-05,
      "loss": 0.4422,
      "step": 3677
    },
    {
      "epoch": 0.3678,
      "grad_norm": 8.079594612121582,
      "learning_rate": 3.192929292929293e-05,
      "loss": 0.321,
      "step": 3678
    },
    {
      "epoch": 0.3679,
      "grad_norm": 6.075235843658447,
      "learning_rate": 3.1924242424242425e-05,
      "loss": 0.5395,
      "step": 3679
    },
    {
      "epoch": 0.368,
      "grad_norm": 7.934453964233398,
      "learning_rate": 3.191919191919192e-05,
      "loss": 0.3883,
      "step": 3680
    },
    {
      "epoch": 0.3681,
      "grad_norm": 12.407731056213379,
      "learning_rate": 3.191414141414141e-05,
      "loss": 0.4876,
      "step": 3681
    },
    {
      "epoch": 0.3682,
      "grad_norm": 8.984400749206543,
      "learning_rate": 3.190909090909091e-05,
      "loss": 0.3762,
      "step": 3682
    },
    {
      "epoch": 0.3683,
      "grad_norm": 10.04879379272461,
      "learning_rate": 3.19040404040404e-05,
      "loss": 0.358,
      "step": 3683
    },
    {
      "epoch": 0.3684,
      "grad_norm": 6.306590557098389,
      "learning_rate": 3.18989898989899e-05,
      "loss": 0.3663,
      "step": 3684
    },
    {
      "epoch": 0.3685,
      "grad_norm": 15.1588773727417,
      "learning_rate": 3.18939393939394e-05,
      "loss": 0.3478,
      "step": 3685
    },
    {
      "epoch": 0.3686,
      "grad_norm": 10.997772216796875,
      "learning_rate": 3.188888888888889e-05,
      "loss": 0.4055,
      "step": 3686
    },
    {
      "epoch": 0.3687,
      "grad_norm": 14.879104614257812,
      "learning_rate": 3.1883838383838386e-05,
      "loss": 0.3329,
      "step": 3687
    },
    {
      "epoch": 0.3688,
      "grad_norm": 12.031668663024902,
      "learning_rate": 3.187878787878788e-05,
      "loss": 0.5891,
      "step": 3688
    },
    {
      "epoch": 0.3689,
      "grad_norm": 13.185200691223145,
      "learning_rate": 3.187373737373738e-05,
      "loss": 0.5548,
      "step": 3689
    },
    {
      "epoch": 0.369,
      "grad_norm": 7.682370185852051,
      "learning_rate": 3.186868686868687e-05,
      "loss": 0.5039,
      "step": 3690
    },
    {
      "epoch": 0.3691,
      "grad_norm": 11.609082221984863,
      "learning_rate": 3.186363636363637e-05,
      "loss": 0.4683,
      "step": 3691
    },
    {
      "epoch": 0.3692,
      "grad_norm": 8.012970924377441,
      "learning_rate": 3.185858585858586e-05,
      "loss": 0.4938,
      "step": 3692
    },
    {
      "epoch": 0.3693,
      "grad_norm": 15.919112205505371,
      "learning_rate": 3.1853535353535355e-05,
      "loss": 0.4624,
      "step": 3693
    },
    {
      "epoch": 0.3694,
      "grad_norm": 12.30611515045166,
      "learning_rate": 3.184848484848485e-05,
      "loss": 0.4297,
      "step": 3694
    },
    {
      "epoch": 0.3695,
      "grad_norm": 10.391206741333008,
      "learning_rate": 3.184343434343435e-05,
      "loss": 0.404,
      "step": 3695
    },
    {
      "epoch": 0.3696,
      "grad_norm": 23.732242584228516,
      "learning_rate": 3.1838383838383836e-05,
      "loss": 0.5311,
      "step": 3696
    },
    {
      "epoch": 0.3697,
      "grad_norm": 17.7181396484375,
      "learning_rate": 3.183333333333334e-05,
      "loss": 0.4804,
      "step": 3697
    },
    {
      "epoch": 0.3698,
      "grad_norm": 21.90742301940918,
      "learning_rate": 3.182828282828283e-05,
      "loss": 0.4675,
      "step": 3698
    },
    {
      "epoch": 0.3699,
      "grad_norm": 11.62314224243164,
      "learning_rate": 3.1823232323232324e-05,
      "loss": 0.5024,
      "step": 3699
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.34152603149414,
      "learning_rate": 3.181818181818182e-05,
      "loss": 0.5339,
      "step": 3700
    },
    {
      "epoch": 0.3701,
      "grad_norm": 14.680389404296875,
      "learning_rate": 3.1813131313131316e-05,
      "loss": 0.4433,
      "step": 3701
    },
    {
      "epoch": 0.3702,
      "grad_norm": 24.471879959106445,
      "learning_rate": 3.180808080808081e-05,
      "loss": 0.568,
      "step": 3702
    },
    {
      "epoch": 0.3703,
      "grad_norm": 11.772561073303223,
      "learning_rate": 3.180303030303031e-05,
      "loss": 0.4996,
      "step": 3703
    },
    {
      "epoch": 0.3704,
      "grad_norm": 16.458837509155273,
      "learning_rate": 3.17979797979798e-05,
      "loss": 0.4595,
      "step": 3704
    },
    {
      "epoch": 0.3705,
      "grad_norm": 15.428691864013672,
      "learning_rate": 3.1792929292929293e-05,
      "loss": 0.4093,
      "step": 3705
    },
    {
      "epoch": 0.3706,
      "grad_norm": 24.192472457885742,
      "learning_rate": 3.178787878787879e-05,
      "loss": 0.4876,
      "step": 3706
    },
    {
      "epoch": 0.3707,
      "grad_norm": 16.233844757080078,
      "learning_rate": 3.1782828282828285e-05,
      "loss": 0.403,
      "step": 3707
    },
    {
      "epoch": 0.3708,
      "grad_norm": 9.97594165802002,
      "learning_rate": 3.177777777777778e-05,
      "loss": 0.4897,
      "step": 3708
    },
    {
      "epoch": 0.3709,
      "grad_norm": 8.81302261352539,
      "learning_rate": 3.177272727272728e-05,
      "loss": 0.4008,
      "step": 3709
    },
    {
      "epoch": 0.371,
      "grad_norm": 10.96662425994873,
      "learning_rate": 3.176767676767677e-05,
      "loss": 0.4937,
      "step": 3710
    },
    {
      "epoch": 0.3711,
      "grad_norm": 9.686018943786621,
      "learning_rate": 3.176262626262627e-05,
      "loss": 0.4925,
      "step": 3711
    },
    {
      "epoch": 0.3712,
      "grad_norm": 10.052613258361816,
      "learning_rate": 3.175757575757576e-05,
      "loss": 0.4641,
      "step": 3712
    },
    {
      "epoch": 0.3713,
      "grad_norm": 24.405099868774414,
      "learning_rate": 3.1752525252525255e-05,
      "loss": 0.4529,
      "step": 3713
    },
    {
      "epoch": 0.3714,
      "grad_norm": 110.64179229736328,
      "learning_rate": 3.174747474747475e-05,
      "loss": 0.4836,
      "step": 3714
    },
    {
      "epoch": 0.3715,
      "grad_norm": 30.095518112182617,
      "learning_rate": 3.174242424242425e-05,
      "loss": 0.571,
      "step": 3715
    },
    {
      "epoch": 0.3716,
      "grad_norm": 19.94891357421875,
      "learning_rate": 3.1737373737373736e-05,
      "loss": 0.3997,
      "step": 3716
    },
    {
      "epoch": 1.0001,
      "grad_norm": 9.229707717895508,
      "learning_rate": 3.173232323232324e-05,
      "loss": 0.6888,
      "step": 3717
    },
    {
      "epoch": 1.0002,
      "grad_norm": 12.592443466186523,
      "learning_rate": 3.172727272727273e-05,
      "loss": 0.4436,
      "step": 3718
    },
    {
      "epoch": 1.0003,
      "grad_norm": 8.962774276733398,
      "learning_rate": 3.1722222222222224e-05,
      "loss": 0.7366,
      "step": 3719
    },
    {
      "epoch": 1.0004,
      "grad_norm": 12.67691421508789,
      "learning_rate": 3.171717171717172e-05,
      "loss": 0.6676,
      "step": 3720
    },
    {
      "epoch": 1.0005,
      "grad_norm": 10.661205291748047,
      "learning_rate": 3.1712121212121216e-05,
      "loss": 0.84,
      "step": 3721
    },
    {
      "epoch": 1.0006,
      "grad_norm": 26.559293746948242,
      "learning_rate": 3.1707070707070705e-05,
      "loss": 0.4587,
      "step": 3722
    },
    {
      "epoch": 1.0007,
      "grad_norm": 29.54787826538086,
      "learning_rate": 3.170202020202021e-05,
      "loss": 0.7316,
      "step": 3723
    },
    {
      "epoch": 1.0008,
      "grad_norm": 14.642072677612305,
      "learning_rate": 3.16969696969697e-05,
      "loss": 1.1048,
      "step": 3724
    },
    {
      "epoch": 1.0009,
      "grad_norm": 6.419451713562012,
      "learning_rate": 3.169191919191919e-05,
      "loss": 0.6115,
      "step": 3725
    },
    {
      "epoch": 1.001,
      "grad_norm": 7.3321404457092285,
      "learning_rate": 3.168686868686869e-05,
      "loss": 0.9123,
      "step": 3726
    },
    {
      "epoch": 1.0011,
      "grad_norm": 9.325461387634277,
      "learning_rate": 3.1681818181818185e-05,
      "loss": 0.701,
      "step": 3727
    },
    {
      "epoch": 1.0012,
      "grad_norm": 8.126152038574219,
      "learning_rate": 3.1676767676767674e-05,
      "loss": 0.4765,
      "step": 3728
    },
    {
      "epoch": 1.0013,
      "grad_norm": 10.33903694152832,
      "learning_rate": 3.167171717171718e-05,
      "loss": 0.7174,
      "step": 3729
    },
    {
      "epoch": 1.0014,
      "grad_norm": 6.72744607925415,
      "learning_rate": 3.1666666666666666e-05,
      "loss": 0.6997,
      "step": 3730
    },
    {
      "epoch": 1.0015,
      "grad_norm": 9.718070983886719,
      "learning_rate": 3.166161616161616e-05,
      "loss": 0.4699,
      "step": 3731
    },
    {
      "epoch": 1.0016,
      "grad_norm": 10.425951957702637,
      "learning_rate": 3.165656565656566e-05,
      "loss": 1.3096,
      "step": 3732
    },
    {
      "epoch": 1.0017,
      "grad_norm": 7.114542007446289,
      "learning_rate": 3.1651515151515154e-05,
      "loss": 0.91,
      "step": 3733
    },
    {
      "epoch": 1.0018,
      "grad_norm": 8.975776672363281,
      "learning_rate": 3.164646464646464e-05,
      "loss": 1.175,
      "step": 3734
    },
    {
      "epoch": 1.0019,
      "grad_norm": 17.607547760009766,
      "learning_rate": 3.1641414141414146e-05,
      "loss": 0.3274,
      "step": 3735
    },
    {
      "epoch": 1.002,
      "grad_norm": 7.519625186920166,
      "learning_rate": 3.1636363636363635e-05,
      "loss": 1.045,
      "step": 3736
    },
    {
      "epoch": 1.0021,
      "grad_norm": 12.444254875183105,
      "learning_rate": 3.163131313131313e-05,
      "loss": 1.0873,
      "step": 3737
    },
    {
      "epoch": 1.0022,
      "grad_norm": 6.009703636169434,
      "learning_rate": 3.162626262626263e-05,
      "loss": 0.3961,
      "step": 3738
    },
    {
      "epoch": 1.0023,
      "grad_norm": 4.221015453338623,
      "learning_rate": 3.162121212121212e-05,
      "loss": 0.2496,
      "step": 3739
    },
    {
      "epoch": 1.0024,
      "grad_norm": 5.600688457489014,
      "learning_rate": 3.161616161616161e-05,
      "loss": 0.6426,
      "step": 3740
    },
    {
      "epoch": 1.0025,
      "grad_norm": 34.44460678100586,
      "learning_rate": 3.1611111111111115e-05,
      "loss": 0.5004,
      "step": 3741
    },
    {
      "epoch": 1.0026,
      "grad_norm": 6.389487266540527,
      "learning_rate": 3.1606060606060604e-05,
      "loss": 0.5773,
      "step": 3742
    },
    {
      "epoch": 1.0027,
      "grad_norm": 10.636252403259277,
      "learning_rate": 3.16010101010101e-05,
      "loss": 0.8514,
      "step": 3743
    },
    {
      "epoch": 1.0028,
      "grad_norm": 5.454176425933838,
      "learning_rate": 3.1595959595959596e-05,
      "loss": 0.4654,
      "step": 3744
    },
    {
      "epoch": 1.0029,
      "grad_norm": 3.8465828895568848,
      "learning_rate": 3.159090909090909e-05,
      "loss": 0.7881,
      "step": 3745
    },
    {
      "epoch": 1.003,
      "grad_norm": 5.298094272613525,
      "learning_rate": 3.158585858585859e-05,
      "loss": 0.7209,
      "step": 3746
    },
    {
      "epoch": 1.0031,
      "grad_norm": 7.095415115356445,
      "learning_rate": 3.1580808080808084e-05,
      "loss": 0.5192,
      "step": 3747
    },
    {
      "epoch": 1.0032,
      "grad_norm": 6.914699077606201,
      "learning_rate": 3.1575757575757574e-05,
      "loss": 0.5223,
      "step": 3748
    },
    {
      "epoch": 1.0033,
      "grad_norm": 7.335485458374023,
      "learning_rate": 3.157070707070707e-05,
      "loss": 1.2034,
      "step": 3749
    },
    {
      "epoch": 1.0034,
      "grad_norm": 12.783926010131836,
      "learning_rate": 3.1565656565656566e-05,
      "loss": 0.859,
      "step": 3750
    },
    {
      "epoch": 1.0035,
      "grad_norm": 6.236701965332031,
      "learning_rate": 3.156060606060606e-05,
      "loss": 0.5078,
      "step": 3751
    },
    {
      "epoch": 1.0036,
      "grad_norm": 7.600392818450928,
      "learning_rate": 3.155555555555556e-05,
      "loss": 0.9978,
      "step": 3752
    },
    {
      "epoch": 1.0037,
      "grad_norm": 6.011531829833984,
      "learning_rate": 3.1550505050505054e-05,
      "loss": 1.2171,
      "step": 3753
    },
    {
      "epoch": 1.0038,
      "grad_norm": 8.019801139831543,
      "learning_rate": 3.154545454545454e-05,
      "loss": 0.7048,
      "step": 3754
    },
    {
      "epoch": 1.0039,
      "grad_norm": 8.95766830444336,
      "learning_rate": 3.1540404040404046e-05,
      "loss": 0.3396,
      "step": 3755
    },
    {
      "epoch": 1.004,
      "grad_norm": 5.638806343078613,
      "learning_rate": 3.1535353535353535e-05,
      "loss": 0.3603,
      "step": 3756
    },
    {
      "epoch": 1.0041,
      "grad_norm": 8.545376777648926,
      "learning_rate": 3.153030303030303e-05,
      "loss": 0.5877,
      "step": 3757
    },
    {
      "epoch": 1.0042,
      "grad_norm": 8.92998218536377,
      "learning_rate": 3.152525252525253e-05,
      "loss": 0.7515,
      "step": 3758
    },
    {
      "epoch": 1.0043,
      "grad_norm": 11.854452133178711,
      "learning_rate": 3.152020202020202e-05,
      "loss": 1.2093,
      "step": 3759
    },
    {
      "epoch": 1.0044,
      "grad_norm": 8.83872127532959,
      "learning_rate": 3.151515151515151e-05,
      "loss": 0.6934,
      "step": 3760
    },
    {
      "epoch": 1.0045,
      "grad_norm": 14.881867408752441,
      "learning_rate": 3.1510101010101015e-05,
      "loss": 0.7582,
      "step": 3761
    },
    {
      "epoch": 1.0046,
      "grad_norm": 9.479458808898926,
      "learning_rate": 3.1505050505050504e-05,
      "loss": 0.6224,
      "step": 3762
    },
    {
      "epoch": 1.0047,
      "grad_norm": 12.425968170166016,
      "learning_rate": 3.15e-05,
      "loss": 0.3448,
      "step": 3763
    },
    {
      "epoch": 1.0048,
      "grad_norm": 8.825929641723633,
      "learning_rate": 3.1494949494949496e-05,
      "loss": 0.5676,
      "step": 3764
    },
    {
      "epoch": 1.0049,
      "grad_norm": 5.15040922164917,
      "learning_rate": 3.148989898989899e-05,
      "loss": 0.3827,
      "step": 3765
    },
    {
      "epoch": 1.005,
      "grad_norm": 5.145324230194092,
      "learning_rate": 3.148484848484848e-05,
      "loss": 0.2997,
      "step": 3766
    },
    {
      "epoch": 1.0051,
      "grad_norm": 3.289232015609741,
      "learning_rate": 3.1479797979797984e-05,
      "loss": 0.6329,
      "step": 3767
    },
    {
      "epoch": 1.0052,
      "grad_norm": 5.69789457321167,
      "learning_rate": 3.147474747474747e-05,
      "loss": 1.1006,
      "step": 3768
    },
    {
      "epoch": 1.0053,
      "grad_norm": 8.715641975402832,
      "learning_rate": 3.146969696969697e-05,
      "loss": 1.0411,
      "step": 3769
    },
    {
      "epoch": 1.0054,
      "grad_norm": 5.597524166107178,
      "learning_rate": 3.146464646464647e-05,
      "loss": 0.7051,
      "step": 3770
    },
    {
      "epoch": 1.0055,
      "grad_norm": 6.368851184844971,
      "learning_rate": 3.145959595959596e-05,
      "loss": 0.7696,
      "step": 3771
    },
    {
      "epoch": 1.0056,
      "grad_norm": 19.441234588623047,
      "learning_rate": 3.145454545454546e-05,
      "loss": 0.4806,
      "step": 3772
    },
    {
      "epoch": 1.0057,
      "grad_norm": 18.692996978759766,
      "learning_rate": 3.144949494949495e-05,
      "loss": 1.2124,
      "step": 3773
    },
    {
      "epoch": 1.0058,
      "grad_norm": 8.945595741271973,
      "learning_rate": 3.144444444444445e-05,
      "loss": 0.4939,
      "step": 3774
    },
    {
      "epoch": 1.0059,
      "grad_norm": 14.051698684692383,
      "learning_rate": 3.143939393939394e-05,
      "loss": 0.7159,
      "step": 3775
    },
    {
      "epoch": 1.006,
      "grad_norm": 5.622730255126953,
      "learning_rate": 3.143434343434344e-05,
      "loss": 0.4603,
      "step": 3776
    },
    {
      "epoch": 1.0061,
      "grad_norm": 7.256845951080322,
      "learning_rate": 3.142929292929293e-05,
      "loss": 0.8265,
      "step": 3777
    },
    {
      "epoch": 1.0062,
      "grad_norm": 8.032363891601562,
      "learning_rate": 3.1424242424242426e-05,
      "loss": 0.6787,
      "step": 3778
    },
    {
      "epoch": 1.0063,
      "grad_norm": 6.639805316925049,
      "learning_rate": 3.141919191919192e-05,
      "loss": 0.4084,
      "step": 3779
    },
    {
      "epoch": 1.0064,
      "grad_norm": 7.537051677703857,
      "learning_rate": 3.141414141414142e-05,
      "loss": 1.3738,
      "step": 3780
    },
    {
      "epoch": 1.0065,
      "grad_norm": 5.764016151428223,
      "learning_rate": 3.140909090909091e-05,
      "loss": 0.5219,
      "step": 3781
    },
    {
      "epoch": 1.0066,
      "grad_norm": 7.512638092041016,
      "learning_rate": 3.140404040404041e-05,
      "loss": 0.2529,
      "step": 3782
    },
    {
      "epoch": 1.0067,
      "grad_norm": 7.469882965087891,
      "learning_rate": 3.13989898989899e-05,
      "loss": 0.5327,
      "step": 3783
    },
    {
      "epoch": 1.0068,
      "grad_norm": 4.863465785980225,
      "learning_rate": 3.1393939393939395e-05,
      "loss": 0.2627,
      "step": 3784
    },
    {
      "epoch": 1.0069,
      "grad_norm": 10.208033561706543,
      "learning_rate": 3.138888888888889e-05,
      "loss": 1.0079,
      "step": 3785
    },
    {
      "epoch": 1.007,
      "grad_norm": 17.564565658569336,
      "learning_rate": 3.138383838383839e-05,
      "loss": 0.997,
      "step": 3786
    },
    {
      "epoch": 1.0071,
      "grad_norm": 6.41581916809082,
      "learning_rate": 3.137878787878788e-05,
      "loss": 0.9106,
      "step": 3787
    },
    {
      "epoch": 1.0072,
      "grad_norm": 13.291561126708984,
      "learning_rate": 3.137373737373738e-05,
      "loss": 0.6391,
      "step": 3788
    },
    {
      "epoch": 1.0073,
      "grad_norm": 12.402172088623047,
      "learning_rate": 3.136868686868687e-05,
      "loss": 0.9572,
      "step": 3789
    },
    {
      "epoch": 1.0074,
      "grad_norm": 8.590553283691406,
      "learning_rate": 3.1363636363636365e-05,
      "loss": 0.2717,
      "step": 3790
    },
    {
      "epoch": 1.0075,
      "grad_norm": 6.6155805587768555,
      "learning_rate": 3.135858585858586e-05,
      "loss": 0.5003,
      "step": 3791
    },
    {
      "epoch": 1.0076,
      "grad_norm": 13.413069725036621,
      "learning_rate": 3.1353535353535357e-05,
      "loss": 1.015,
      "step": 3792
    },
    {
      "epoch": 1.0077,
      "grad_norm": 7.00457763671875,
      "learning_rate": 3.134848484848485e-05,
      "loss": 0.3311,
      "step": 3793
    },
    {
      "epoch": 1.0078,
      "grad_norm": 6.999417781829834,
      "learning_rate": 3.134343434343435e-05,
      "loss": 0.8222,
      "step": 3794
    },
    {
      "epoch": 1.0079,
      "grad_norm": 5.177791118621826,
      "learning_rate": 3.133838383838384e-05,
      "loss": 0.8213,
      "step": 3795
    },
    {
      "epoch": 1.008,
      "grad_norm": 5.79742956161499,
      "learning_rate": 3.1333333333333334e-05,
      "loss": 1.3622,
      "step": 3796
    },
    {
      "epoch": 1.0081,
      "grad_norm": 6.299932479858398,
      "learning_rate": 3.132828282828283e-05,
      "loss": 0.8263,
      "step": 3797
    },
    {
      "epoch": 1.0082,
      "grad_norm": 5.000255107879639,
      "learning_rate": 3.1323232323232326e-05,
      "loss": 0.2862,
      "step": 3798
    },
    {
      "epoch": 1.0083,
      "grad_norm": 8.319257736206055,
      "learning_rate": 3.131818181818182e-05,
      "loss": 0.473,
      "step": 3799
    },
    {
      "epoch": 1.0084,
      "grad_norm": 7.4471540451049805,
      "learning_rate": 3.131313131313132e-05,
      "loss": 0.4007,
      "step": 3800
    },
    {
      "epoch": 1.0085,
      "grad_norm": 5.831930637359619,
      "learning_rate": 3.130808080808081e-05,
      "loss": 0.5508,
      "step": 3801
    },
    {
      "epoch": 1.0086,
      "grad_norm": 7.166514873504639,
      "learning_rate": 3.13030303030303e-05,
      "loss": 0.9688,
      "step": 3802
    },
    {
      "epoch": 1.0087,
      "grad_norm": 9.302580833435059,
      "learning_rate": 3.12979797979798e-05,
      "loss": 0.3807,
      "step": 3803
    },
    {
      "epoch": 1.0088,
      "grad_norm": 4.533400058746338,
      "learning_rate": 3.1292929292929295e-05,
      "loss": 0.3929,
      "step": 3804
    },
    {
      "epoch": 1.0089,
      "grad_norm": 8.27977180480957,
      "learning_rate": 3.128787878787879e-05,
      "loss": 0.6055,
      "step": 3805
    },
    {
      "epoch": 1.009,
      "grad_norm": 6.8624467849731445,
      "learning_rate": 3.128282828282829e-05,
      "loss": 0.8837,
      "step": 3806
    },
    {
      "epoch": 1.0091,
      "grad_norm": 6.599456787109375,
      "learning_rate": 3.1277777777777776e-05,
      "loss": 1.6628,
      "step": 3807
    },
    {
      "epoch": 1.0092,
      "grad_norm": 7.681637287139893,
      "learning_rate": 3.127272727272728e-05,
      "loss": 1.316,
      "step": 3808
    },
    {
      "epoch": 1.0093,
      "grad_norm": 104.82353210449219,
      "learning_rate": 3.126767676767677e-05,
      "loss": 0.8437,
      "step": 3809
    },
    {
      "epoch": 1.0094,
      "grad_norm": 5.585676193237305,
      "learning_rate": 3.1262626262626264e-05,
      "loss": 0.6114,
      "step": 3810
    },
    {
      "epoch": 1.0095,
      "grad_norm": 9.741803169250488,
      "learning_rate": 3.125757575757576e-05,
      "loss": 0.8167,
      "step": 3811
    },
    {
      "epoch": 1.0096,
      "grad_norm": 7.312097072601318,
      "learning_rate": 3.1252525252525256e-05,
      "loss": 0.6835,
      "step": 3812
    },
    {
      "epoch": 1.0097,
      "grad_norm": 4.19377326965332,
      "learning_rate": 3.1247474747474745e-05,
      "loss": 0.7552,
      "step": 3813
    },
    {
      "epoch": 1.0098,
      "grad_norm": 4.284092426300049,
      "learning_rate": 3.124242424242425e-05,
      "loss": 0.3334,
      "step": 3814
    },
    {
      "epoch": 1.0099,
      "grad_norm": 6.684450626373291,
      "learning_rate": 3.123737373737374e-05,
      "loss": 0.6327,
      "step": 3815
    },
    {
      "epoch": 1.01,
      "grad_norm": 4.749835014343262,
      "learning_rate": 3.123232323232323e-05,
      "loss": 0.799,
      "step": 3816
    },
    {
      "epoch": 1.0101,
      "grad_norm": 11.100340843200684,
      "learning_rate": 3.122727272727273e-05,
      "loss": 0.5352,
      "step": 3817
    },
    {
      "epoch": 1.0102,
      "grad_norm": 8.759513854980469,
      "learning_rate": 3.1222222222222225e-05,
      "loss": 0.4253,
      "step": 3818
    },
    {
      "epoch": 1.0103,
      "grad_norm": 5.9173102378845215,
      "learning_rate": 3.1217171717171714e-05,
      "loss": 0.5502,
      "step": 3819
    },
    {
      "epoch": 1.0104,
      "grad_norm": 9.929692268371582,
      "learning_rate": 3.121212121212122e-05,
      "loss": 0.3609,
      "step": 3820
    },
    {
      "epoch": 1.0105,
      "grad_norm": 7.13292932510376,
      "learning_rate": 3.1207070707070706e-05,
      "loss": 0.4846,
      "step": 3821
    },
    {
      "epoch": 1.0106,
      "grad_norm": 21.723295211791992,
      "learning_rate": 3.12020202020202e-05,
      "loss": 0.6014,
      "step": 3822
    },
    {
      "epoch": 1.0107,
      "grad_norm": 9.913912773132324,
      "learning_rate": 3.11969696969697e-05,
      "loss": 0.6945,
      "step": 3823
    },
    {
      "epoch": 1.0108,
      "grad_norm": 5.362124443054199,
      "learning_rate": 3.1191919191919194e-05,
      "loss": 0.4431,
      "step": 3824
    },
    {
      "epoch": 1.0109,
      "grad_norm": 5.458669185638428,
      "learning_rate": 3.1186868686868684e-05,
      "loss": 1.1999,
      "step": 3825
    },
    {
      "epoch": 1.011,
      "grad_norm": 5.276517391204834,
      "learning_rate": 3.1181818181818186e-05,
      "loss": 0.3652,
      "step": 3826
    },
    {
      "epoch": 1.0111,
      "grad_norm": 17.298259735107422,
      "learning_rate": 3.1176767676767676e-05,
      "loss": 0.4848,
      "step": 3827
    },
    {
      "epoch": 1.0112,
      "grad_norm": 5.544802188873291,
      "learning_rate": 3.117171717171717e-05,
      "loss": 0.6856,
      "step": 3828
    },
    {
      "epoch": 1.0113,
      "grad_norm": 40.81041717529297,
      "learning_rate": 3.116666666666667e-05,
      "loss": 0.8333,
      "step": 3829
    },
    {
      "epoch": 1.0114,
      "grad_norm": 15.82512092590332,
      "learning_rate": 3.1161616161616164e-05,
      "loss": 0.7124,
      "step": 3830
    },
    {
      "epoch": 1.0115,
      "grad_norm": 4.705517768859863,
      "learning_rate": 3.115656565656565e-05,
      "loss": 0.8154,
      "step": 3831
    },
    {
      "epoch": 1.0116,
      "grad_norm": 8.47294807434082,
      "learning_rate": 3.1151515151515156e-05,
      "loss": 0.5337,
      "step": 3832
    },
    {
      "epoch": 1.0117,
      "grad_norm": 54.48566818237305,
      "learning_rate": 3.1146464646464645e-05,
      "loss": 0.5968,
      "step": 3833
    },
    {
      "epoch": 1.0118,
      "grad_norm": 3.9088826179504395,
      "learning_rate": 3.114141414141414e-05,
      "loss": 0.5684,
      "step": 3834
    },
    {
      "epoch": 1.0119,
      "grad_norm": 4.778694152832031,
      "learning_rate": 3.113636363636364e-05,
      "loss": 0.3341,
      "step": 3835
    },
    {
      "epoch": 1.012,
      "grad_norm": 7.260455131530762,
      "learning_rate": 3.113131313131313e-05,
      "loss": 0.4094,
      "step": 3836
    },
    {
      "epoch": 1.0121,
      "grad_norm": 6.389208793640137,
      "learning_rate": 3.112626262626263e-05,
      "loss": 0.351,
      "step": 3837
    },
    {
      "epoch": 1.0122,
      "grad_norm": 5.439794063568115,
      "learning_rate": 3.1121212121212125e-05,
      "loss": 0.5815,
      "step": 3838
    },
    {
      "epoch": 1.0123,
      "grad_norm": 9.018392562866211,
      "learning_rate": 3.1116161616161614e-05,
      "loss": 0.611,
      "step": 3839
    },
    {
      "epoch": 1.0124,
      "grad_norm": 6.214982032775879,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.6342,
      "step": 3840
    },
    {
      "epoch": 1.0125,
      "grad_norm": 6.77665376663208,
      "learning_rate": 3.1106060606060606e-05,
      "loss": 0.4486,
      "step": 3841
    },
    {
      "epoch": 1.0126,
      "grad_norm": 5.697810173034668,
      "learning_rate": 3.11010101010101e-05,
      "loss": 0.5485,
      "step": 3842
    },
    {
      "epoch": 1.0127,
      "grad_norm": 12.372769355773926,
      "learning_rate": 3.10959595959596e-05,
      "loss": 0.5192,
      "step": 3843
    },
    {
      "epoch": 1.0128,
      "grad_norm": 7.541605472564697,
      "learning_rate": 3.1090909090909094e-05,
      "loss": 0.5655,
      "step": 3844
    },
    {
      "epoch": 1.0129,
      "grad_norm": 12.050522804260254,
      "learning_rate": 3.108585858585858e-05,
      "loss": 0.534,
      "step": 3845
    },
    {
      "epoch": 1.013,
      "grad_norm": 4.648467540740967,
      "learning_rate": 3.1080808080808086e-05,
      "loss": 0.3506,
      "step": 3846
    },
    {
      "epoch": 1.0131000000000001,
      "grad_norm": 17.882753372192383,
      "learning_rate": 3.1075757575757575e-05,
      "loss": 0.8578,
      "step": 3847
    },
    {
      "epoch": 1.0132,
      "grad_norm": 8.238088607788086,
      "learning_rate": 3.107070707070707e-05,
      "loss": 0.3777,
      "step": 3848
    },
    {
      "epoch": 1.0133,
      "grad_norm": 7.755612373352051,
      "learning_rate": 3.106565656565657e-05,
      "loss": 0.5424,
      "step": 3849
    },
    {
      "epoch": 1.0134,
      "grad_norm": 4.720749378204346,
      "learning_rate": 3.106060606060606e-05,
      "loss": 0.2643,
      "step": 3850
    },
    {
      "epoch": 1.0135,
      "grad_norm": 4.549808025360107,
      "learning_rate": 3.105555555555555e-05,
      "loss": 0.3068,
      "step": 3851
    },
    {
      "epoch": 1.0136,
      "grad_norm": 4.113548755645752,
      "learning_rate": 3.1050505050505055e-05,
      "loss": 0.3187,
      "step": 3852
    },
    {
      "epoch": 1.0137,
      "grad_norm": 6.293875694274902,
      "learning_rate": 3.1045454545454544e-05,
      "loss": 0.4924,
      "step": 3853
    },
    {
      "epoch": 1.0138,
      "grad_norm": 5.108659744262695,
      "learning_rate": 3.104040404040404e-05,
      "loss": 0.4168,
      "step": 3854
    },
    {
      "epoch": 1.0139,
      "grad_norm": 8.294623374938965,
      "learning_rate": 3.1035353535353536e-05,
      "loss": 0.7398,
      "step": 3855
    },
    {
      "epoch": 1.014,
      "grad_norm": 5.7486395835876465,
      "learning_rate": 3.103030303030303e-05,
      "loss": 0.7198,
      "step": 3856
    },
    {
      "epoch": 1.0141,
      "grad_norm": 6.939154148101807,
      "learning_rate": 3.102525252525253e-05,
      "loss": 0.989,
      "step": 3857
    },
    {
      "epoch": 1.0142,
      "grad_norm": 5.5463361740112305,
      "learning_rate": 3.1020202020202024e-05,
      "loss": 0.7144,
      "step": 3858
    },
    {
      "epoch": 1.0143,
      "grad_norm": 5.799988269805908,
      "learning_rate": 3.101515151515152e-05,
      "loss": 0.6763,
      "step": 3859
    },
    {
      "epoch": 1.0144,
      "grad_norm": 5.622649669647217,
      "learning_rate": 3.101010101010101e-05,
      "loss": 0.5348,
      "step": 3860
    },
    {
      "epoch": 1.0145,
      "grad_norm": 64.56425476074219,
      "learning_rate": 3.100505050505051e-05,
      "loss": 1.2543,
      "step": 3861
    },
    {
      "epoch": 1.0146,
      "grad_norm": 6.112376689910889,
      "learning_rate": 3.1e-05,
      "loss": 0.6538,
      "step": 3862
    },
    {
      "epoch": 1.0147,
      "grad_norm": 6.560323238372803,
      "learning_rate": 3.09949494949495e-05,
      "loss": 0.8659,
      "step": 3863
    },
    {
      "epoch": 1.0148,
      "grad_norm": 6.388541221618652,
      "learning_rate": 3.098989898989899e-05,
      "loss": 0.685,
      "step": 3864
    },
    {
      "epoch": 1.0149,
      "grad_norm": 4.48335075378418,
      "learning_rate": 3.098484848484849e-05,
      "loss": 0.2977,
      "step": 3865
    },
    {
      "epoch": 1.015,
      "grad_norm": 10.774438858032227,
      "learning_rate": 3.097979797979798e-05,
      "loss": 0.2978,
      "step": 3866
    },
    {
      "epoch": 1.0151,
      "grad_norm": 7.974110126495361,
      "learning_rate": 3.097474747474748e-05,
      "loss": 0.7695,
      "step": 3867
    },
    {
      "epoch": 1.0152,
      "grad_norm": 9.033313751220703,
      "learning_rate": 3.096969696969697e-05,
      "loss": 0.4322,
      "step": 3868
    },
    {
      "epoch": 1.0153,
      "grad_norm": 10.356146812438965,
      "learning_rate": 3.0964646464646467e-05,
      "loss": 0.5518,
      "step": 3869
    },
    {
      "epoch": 1.0154,
      "grad_norm": 7.4730730056762695,
      "learning_rate": 3.095959595959596e-05,
      "loss": 0.704,
      "step": 3870
    },
    {
      "epoch": 1.0155,
      "grad_norm": 11.878110885620117,
      "learning_rate": 3.095454545454546e-05,
      "loss": 1.0269,
      "step": 3871
    },
    {
      "epoch": 1.0156,
      "grad_norm": 12.649433135986328,
      "learning_rate": 3.094949494949495e-05,
      "loss": 0.5698,
      "step": 3872
    },
    {
      "epoch": 1.0157,
      "grad_norm": 7.670411586761475,
      "learning_rate": 3.094444444444445e-05,
      "loss": 1.4707,
      "step": 3873
    },
    {
      "epoch": 1.0158,
      "grad_norm": 6.817094326019287,
      "learning_rate": 3.093939393939394e-05,
      "loss": 0.6434,
      "step": 3874
    },
    {
      "epoch": 1.0159,
      "grad_norm": 9.960357666015625,
      "learning_rate": 3.0934343434343436e-05,
      "loss": 0.7096,
      "step": 3875
    },
    {
      "epoch": 1.016,
      "grad_norm": 6.941593170166016,
      "learning_rate": 3.092929292929293e-05,
      "loss": 0.8407,
      "step": 3876
    },
    {
      "epoch": 1.0161,
      "grad_norm": 7.271470069885254,
      "learning_rate": 3.092424242424243e-05,
      "loss": 0.4129,
      "step": 3877
    },
    {
      "epoch": 1.0162,
      "grad_norm": 3.9579946994781494,
      "learning_rate": 3.091919191919192e-05,
      "loss": 0.2754,
      "step": 3878
    },
    {
      "epoch": 1.0163,
      "grad_norm": 15.7833833694458,
      "learning_rate": 3.091414141414142e-05,
      "loss": 0.7369,
      "step": 3879
    },
    {
      "epoch": 1.0164,
      "grad_norm": 4.959418296813965,
      "learning_rate": 3.090909090909091e-05,
      "loss": 0.4639,
      "step": 3880
    },
    {
      "epoch": 1.0165,
      "grad_norm": 5.454843521118164,
      "learning_rate": 3.0904040404040405e-05,
      "loss": 0.5236,
      "step": 3881
    },
    {
      "epoch": 1.0166,
      "grad_norm": 6.056828498840332,
      "learning_rate": 3.08989898989899e-05,
      "loss": 0.4271,
      "step": 3882
    },
    {
      "epoch": 1.0167,
      "grad_norm": 6.1670145988464355,
      "learning_rate": 3.08939393939394e-05,
      "loss": 0.4156,
      "step": 3883
    },
    {
      "epoch": 1.0168,
      "grad_norm": 7.94018030166626,
      "learning_rate": 3.088888888888889e-05,
      "loss": 0.9483,
      "step": 3884
    },
    {
      "epoch": 1.0169,
      "grad_norm": 7.727536201477051,
      "learning_rate": 3.088383838383839e-05,
      "loss": 0.3679,
      "step": 3885
    },
    {
      "epoch": 1.017,
      "grad_norm": 7.592849254608154,
      "learning_rate": 3.087878787878788e-05,
      "loss": 0.5504,
      "step": 3886
    },
    {
      "epoch": 1.0171,
      "grad_norm": 10.460702896118164,
      "learning_rate": 3.0873737373737374e-05,
      "loss": 0.8044,
      "step": 3887
    },
    {
      "epoch": 1.0172,
      "grad_norm": 8.651568412780762,
      "learning_rate": 3.086868686868687e-05,
      "loss": 1.1721,
      "step": 3888
    },
    {
      "epoch": 1.0173,
      "grad_norm": 4.143702983856201,
      "learning_rate": 3.0863636363636366e-05,
      "loss": 0.4614,
      "step": 3889
    },
    {
      "epoch": 1.0174,
      "grad_norm": 6.025279998779297,
      "learning_rate": 3.085858585858586e-05,
      "loss": 0.7123,
      "step": 3890
    },
    {
      "epoch": 1.0175,
      "grad_norm": 7.142994403839111,
      "learning_rate": 3.085353535353536e-05,
      "loss": 0.4471,
      "step": 3891
    },
    {
      "epoch": 1.0176,
      "grad_norm": 10.91099739074707,
      "learning_rate": 3.084848484848485e-05,
      "loss": 0.6927,
      "step": 3892
    },
    {
      "epoch": 1.0177,
      "grad_norm": 3.840233325958252,
      "learning_rate": 3.084343434343434e-05,
      "loss": 0.2744,
      "step": 3893
    },
    {
      "epoch": 1.0178,
      "grad_norm": 4.277308940887451,
      "learning_rate": 3.083838383838384e-05,
      "loss": 0.3298,
      "step": 3894
    },
    {
      "epoch": 1.0179,
      "grad_norm": 5.952761173248291,
      "learning_rate": 3.0833333333333335e-05,
      "loss": 1.159,
      "step": 3895
    },
    {
      "epoch": 1.018,
      "grad_norm": 4.321505546569824,
      "learning_rate": 3.082828282828283e-05,
      "loss": 0.545,
      "step": 3896
    },
    {
      "epoch": 1.0181,
      "grad_norm": 19.8804874420166,
      "learning_rate": 3.082323232323233e-05,
      "loss": 1.0031,
      "step": 3897
    },
    {
      "epoch": 1.0182,
      "grad_norm": 12.801300048828125,
      "learning_rate": 3.0818181818181816e-05,
      "loss": 0.7195,
      "step": 3898
    },
    {
      "epoch": 1.0183,
      "grad_norm": 11.681839942932129,
      "learning_rate": 3.081313131313132e-05,
      "loss": 0.3623,
      "step": 3899
    },
    {
      "epoch": 1.0184,
      "grad_norm": 8.626736640930176,
      "learning_rate": 3.080808080808081e-05,
      "loss": 0.4752,
      "step": 3900
    },
    {
      "epoch": 1.0185,
      "grad_norm": 7.696104526519775,
      "learning_rate": 3.0803030303030304e-05,
      "loss": 1.0085,
      "step": 3901
    },
    {
      "epoch": 1.0186,
      "grad_norm": 6.84243106842041,
      "learning_rate": 3.07979797979798e-05,
      "loss": 1.0217,
      "step": 3902
    },
    {
      "epoch": 1.0187,
      "grad_norm": 8.889747619628906,
      "learning_rate": 3.0792929292929296e-05,
      "loss": 0.6794,
      "step": 3903
    },
    {
      "epoch": 1.0188,
      "grad_norm": 3.5609030723571777,
      "learning_rate": 3.0787878787878786e-05,
      "loss": 0.2717,
      "step": 3904
    },
    {
      "epoch": 1.0189,
      "grad_norm": 7.105630397796631,
      "learning_rate": 3.078282828282829e-05,
      "loss": 0.6165,
      "step": 3905
    },
    {
      "epoch": 1.019,
      "grad_norm": 8.01180362701416,
      "learning_rate": 3.077777777777778e-05,
      "loss": 0.7855,
      "step": 3906
    },
    {
      "epoch": 1.0191,
      "grad_norm": 6.128049850463867,
      "learning_rate": 3.0772727272727273e-05,
      "loss": 0.4764,
      "step": 3907
    },
    {
      "epoch": 1.0192,
      "grad_norm": 10.821503639221191,
      "learning_rate": 3.076767676767677e-05,
      "loss": 0.8496,
      "step": 3908
    },
    {
      "epoch": 1.0193,
      "grad_norm": 6.939219951629639,
      "learning_rate": 3.0762626262626265e-05,
      "loss": 0.3921,
      "step": 3909
    },
    {
      "epoch": 1.0194,
      "grad_norm": 5.357702732086182,
      "learning_rate": 3.0757575757575755e-05,
      "loss": 0.3886,
      "step": 3910
    },
    {
      "epoch": 1.0195,
      "grad_norm": 6.21358060836792,
      "learning_rate": 3.075252525252526e-05,
      "loss": 0.4007,
      "step": 3911
    },
    {
      "epoch": 1.0196,
      "grad_norm": 5.810003757476807,
      "learning_rate": 3.074747474747475e-05,
      "loss": 0.6406,
      "step": 3912
    },
    {
      "epoch": 1.0197,
      "grad_norm": 17.165014266967773,
      "learning_rate": 3.074242424242424e-05,
      "loss": 0.8838,
      "step": 3913
    },
    {
      "epoch": 1.0198,
      "grad_norm": 3.793957233428955,
      "learning_rate": 3.073737373737374e-05,
      "loss": 0.1392,
      "step": 3914
    },
    {
      "epoch": 1.0199,
      "grad_norm": 7.455753803253174,
      "learning_rate": 3.0732323232323235e-05,
      "loss": 0.6413,
      "step": 3915
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.144548416137695,
      "learning_rate": 3.0727272727272724e-05,
      "loss": 0.1925,
      "step": 3916
    },
    {
      "epoch": 1.0201,
      "grad_norm": 7.987794876098633,
      "learning_rate": 3.0722222222222227e-05,
      "loss": 0.361,
      "step": 3917
    },
    {
      "epoch": 1.0202,
      "grad_norm": 28.139467239379883,
      "learning_rate": 3.0717171717171716e-05,
      "loss": 0.3208,
      "step": 3918
    },
    {
      "epoch": 1.0203,
      "grad_norm": 9.572235107421875,
      "learning_rate": 3.071212121212121e-05,
      "loss": 1.0874,
      "step": 3919
    },
    {
      "epoch": 1.0204,
      "grad_norm": 90.64630126953125,
      "learning_rate": 3.070707070707071e-05,
      "loss": 0.611,
      "step": 3920
    },
    {
      "epoch": 1.0205,
      "grad_norm": 8.863326072692871,
      "learning_rate": 3.0702020202020204e-05,
      "loss": 0.5581,
      "step": 3921
    },
    {
      "epoch": 1.0206,
      "grad_norm": 12.000218391418457,
      "learning_rate": 3.069696969696969e-05,
      "loss": 0.748,
      "step": 3922
    },
    {
      "epoch": 1.0207,
      "grad_norm": 5.341367244720459,
      "learning_rate": 3.0691919191919196e-05,
      "loss": 0.2378,
      "step": 3923
    },
    {
      "epoch": 1.0208,
      "grad_norm": 13.14790153503418,
      "learning_rate": 3.0686868686868685e-05,
      "loss": 0.6132,
      "step": 3924
    },
    {
      "epoch": 1.0209,
      "grad_norm": 10.42823314666748,
      "learning_rate": 3.068181818181818e-05,
      "loss": 0.4774,
      "step": 3925
    },
    {
      "epoch": 1.021,
      "grad_norm": 4.232323169708252,
      "learning_rate": 3.067676767676768e-05,
      "loss": 0.2217,
      "step": 3926
    },
    {
      "epoch": 1.0211,
      "grad_norm": 7.426917552947998,
      "learning_rate": 3.067171717171717e-05,
      "loss": 0.5267,
      "step": 3927
    },
    {
      "epoch": 1.0212,
      "grad_norm": 6.04539155960083,
      "learning_rate": 3.066666666666667e-05,
      "loss": 0.335,
      "step": 3928
    },
    {
      "epoch": 1.0213,
      "grad_norm": 8.100221633911133,
      "learning_rate": 3.0661616161616165e-05,
      "loss": 0.35,
      "step": 3929
    },
    {
      "epoch": 1.0214,
      "grad_norm": 5.951888561248779,
      "learning_rate": 3.0656565656565654e-05,
      "loss": 0.5914,
      "step": 3930
    },
    {
      "epoch": 1.0215,
      "grad_norm": 5.916187286376953,
      "learning_rate": 3.065151515151515e-05,
      "loss": 0.3661,
      "step": 3931
    },
    {
      "epoch": 1.0216,
      "grad_norm": 9.248092651367188,
      "learning_rate": 3.0646464646464646e-05,
      "loss": 0.6926,
      "step": 3932
    },
    {
      "epoch": 1.0217,
      "grad_norm": 4.331905364990234,
      "learning_rate": 3.064141414141414e-05,
      "loss": 0.4604,
      "step": 3933
    },
    {
      "epoch": 1.0218,
      "grad_norm": 16.932048797607422,
      "learning_rate": 3.063636363636364e-05,
      "loss": 0.5831,
      "step": 3934
    },
    {
      "epoch": 1.0219,
      "grad_norm": 7.211185455322266,
      "learning_rate": 3.0631313131313134e-05,
      "loss": 0.6217,
      "step": 3935
    },
    {
      "epoch": 1.022,
      "grad_norm": 4.4129509925842285,
      "learning_rate": 3.062626262626262e-05,
      "loss": 0.318,
      "step": 3936
    },
    {
      "epoch": 1.0221,
      "grad_norm": 10.808361053466797,
      "learning_rate": 3.0621212121212126e-05,
      "loss": 1.9496,
      "step": 3937
    },
    {
      "epoch": 1.0222,
      "grad_norm": 7.500816345214844,
      "learning_rate": 3.0616161616161615e-05,
      "loss": 0.3961,
      "step": 3938
    },
    {
      "epoch": 1.0223,
      "grad_norm": 6.000560283660889,
      "learning_rate": 3.061111111111111e-05,
      "loss": 0.2788,
      "step": 3939
    },
    {
      "epoch": 1.0224,
      "grad_norm": 5.452802658081055,
      "learning_rate": 3.060606060606061e-05,
      "loss": 0.8303,
      "step": 3940
    },
    {
      "epoch": 1.0225,
      "grad_norm": 18.939939498901367,
      "learning_rate": 3.06010101010101e-05,
      "loss": 0.6443,
      "step": 3941
    },
    {
      "epoch": 1.0226,
      "grad_norm": 13.59001350402832,
      "learning_rate": 3.05959595959596e-05,
      "loss": 1.0917,
      "step": 3942
    },
    {
      "epoch": 1.0227,
      "grad_norm": 5.2266926765441895,
      "learning_rate": 3.0590909090909095e-05,
      "loss": 0.6978,
      "step": 3943
    },
    {
      "epoch": 1.0228,
      "grad_norm": 7.9468560218811035,
      "learning_rate": 3.058585858585859e-05,
      "loss": 0.7209,
      "step": 3944
    },
    {
      "epoch": 1.0229,
      "grad_norm": 10.422649383544922,
      "learning_rate": 3.058080808080808e-05,
      "loss": 1.1843,
      "step": 3945
    },
    {
      "epoch": 1.023,
      "grad_norm": 6.875418186187744,
      "learning_rate": 3.0575757575757576e-05,
      "loss": 0.6149,
      "step": 3946
    },
    {
      "epoch": 1.0231,
      "grad_norm": 10.160399436950684,
      "learning_rate": 3.057070707070707e-05,
      "loss": 0.3111,
      "step": 3947
    },
    {
      "epoch": 1.0232,
      "grad_norm": 7.451622486114502,
      "learning_rate": 3.056565656565657e-05,
      "loss": 0.3381,
      "step": 3948
    },
    {
      "epoch": 1.0233,
      "grad_norm": 3.566934108734131,
      "learning_rate": 3.0560606060606064e-05,
      "loss": 0.2425,
      "step": 3949
    },
    {
      "epoch": 1.0234,
      "grad_norm": 7.25230073928833,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.5827,
      "step": 3950
    },
    {
      "epoch": 1.0235,
      "grad_norm": 4.780713081359863,
      "learning_rate": 3.055050505050505e-05,
      "loss": 0.338,
      "step": 3951
    },
    {
      "epoch": 1.0236,
      "grad_norm": 8.309432029724121,
      "learning_rate": 3.054545454545455e-05,
      "loss": 0.4558,
      "step": 3952
    },
    {
      "epoch": 1.0237,
      "grad_norm": 3.948124647140503,
      "learning_rate": 3.054040404040404e-05,
      "loss": 0.3101,
      "step": 3953
    },
    {
      "epoch": 1.0238,
      "grad_norm": 8.213281631469727,
      "learning_rate": 3.053535353535354e-05,
      "loss": 0.8535,
      "step": 3954
    },
    {
      "epoch": 1.0239,
      "grad_norm": 4.801246643066406,
      "learning_rate": 3.0530303030303034e-05,
      "loss": 0.8216,
      "step": 3955
    },
    {
      "epoch": 1.024,
      "grad_norm": 11.194289207458496,
      "learning_rate": 3.052525252525253e-05,
      "loss": 0.4427,
      "step": 3956
    },
    {
      "epoch": 1.0241,
      "grad_norm": 11.583490371704102,
      "learning_rate": 3.052020202020202e-05,
      "loss": 0.8831,
      "step": 3957
    },
    {
      "epoch": 1.0242,
      "grad_norm": 15.51414966583252,
      "learning_rate": 3.0515151515151518e-05,
      "loss": 0.9486,
      "step": 3958
    },
    {
      "epoch": 1.0243,
      "grad_norm": 6.408544063568115,
      "learning_rate": 3.051010101010101e-05,
      "loss": 0.7851,
      "step": 3959
    },
    {
      "epoch": 1.0244,
      "grad_norm": 4.489502429962158,
      "learning_rate": 3.050505050505051e-05,
      "loss": 0.308,
      "step": 3960
    },
    {
      "epoch": 1.0245,
      "grad_norm": 10.953300476074219,
      "learning_rate": 3.05e-05,
      "loss": 0.8238,
      "step": 3961
    },
    {
      "epoch": 1.0246,
      "grad_norm": 7.677333354949951,
      "learning_rate": 3.04949494949495e-05,
      "loss": 0.988,
      "step": 3962
    },
    {
      "epoch": 1.0247,
      "grad_norm": 7.797346115112305,
      "learning_rate": 3.048989898989899e-05,
      "loss": 0.4972,
      "step": 3963
    },
    {
      "epoch": 1.0248,
      "grad_norm": 3.8630332946777344,
      "learning_rate": 3.0484848484848487e-05,
      "loss": 0.3019,
      "step": 3964
    },
    {
      "epoch": 1.0249,
      "grad_norm": 6.6454548835754395,
      "learning_rate": 3.047979797979798e-05,
      "loss": 0.5632,
      "step": 3965
    },
    {
      "epoch": 1.025,
      "grad_norm": 9.498903274536133,
      "learning_rate": 3.047474747474748e-05,
      "loss": 0.9868,
      "step": 3966
    },
    {
      "epoch": 1.0251,
      "grad_norm": 8.382486343383789,
      "learning_rate": 3.0469696969696972e-05,
      "loss": 0.8531,
      "step": 3967
    },
    {
      "epoch": 1.0252,
      "grad_norm": 9.85451889038086,
      "learning_rate": 3.0464646464646468e-05,
      "loss": 0.5317,
      "step": 3968
    },
    {
      "epoch": 1.0253,
      "grad_norm": 9.48768424987793,
      "learning_rate": 3.045959595959596e-05,
      "loss": 0.7729,
      "step": 3969
    },
    {
      "epoch": 1.0254,
      "grad_norm": 6.957034587860107,
      "learning_rate": 3.0454545454545456e-05,
      "loss": 0.3708,
      "step": 3970
    },
    {
      "epoch": 1.0255,
      "grad_norm": 4.243366718292236,
      "learning_rate": 3.044949494949495e-05,
      "loss": 0.7705,
      "step": 3971
    },
    {
      "epoch": 1.0256,
      "grad_norm": 8.578618049621582,
      "learning_rate": 3.044444444444445e-05,
      "loss": 0.6741,
      "step": 3972
    },
    {
      "epoch": 1.0257,
      "grad_norm": 21.83595848083496,
      "learning_rate": 3.043939393939394e-05,
      "loss": 0.3733,
      "step": 3973
    },
    {
      "epoch": 1.0258,
      "grad_norm": 17.04964828491211,
      "learning_rate": 3.0434343434343437e-05,
      "loss": 0.8476,
      "step": 3974
    },
    {
      "epoch": 1.0259,
      "grad_norm": 3.727980375289917,
      "learning_rate": 3.042929292929293e-05,
      "loss": 0.1918,
      "step": 3975
    },
    {
      "epoch": 1.026,
      "grad_norm": 5.944136619567871,
      "learning_rate": 3.0424242424242426e-05,
      "loss": 0.2305,
      "step": 3976
    },
    {
      "epoch": 1.0261,
      "grad_norm": 28.351055145263672,
      "learning_rate": 3.0419191919191918e-05,
      "loss": 0.5579,
      "step": 3977
    },
    {
      "epoch": 1.0262,
      "grad_norm": 5.594482898712158,
      "learning_rate": 3.0414141414141418e-05,
      "loss": 0.4916,
      "step": 3978
    },
    {
      "epoch": 1.0263,
      "grad_norm": 5.677460670471191,
      "learning_rate": 3.040909090909091e-05,
      "loss": 0.7297,
      "step": 3979
    },
    {
      "epoch": 1.0264,
      "grad_norm": 10.472756385803223,
      "learning_rate": 3.0404040404040406e-05,
      "loss": 0.3752,
      "step": 3980
    },
    {
      "epoch": 1.0265,
      "grad_norm": 5.605809688568115,
      "learning_rate": 3.03989898989899e-05,
      "loss": 0.8759,
      "step": 3981
    },
    {
      "epoch": 1.0266,
      "grad_norm": 5.915964603424072,
      "learning_rate": 3.0393939393939398e-05,
      "loss": 0.2024,
      "step": 3982
    },
    {
      "epoch": 1.0267,
      "grad_norm": 7.1327805519104,
      "learning_rate": 3.0388888888888887e-05,
      "loss": 0.6515,
      "step": 3983
    },
    {
      "epoch": 1.0268,
      "grad_norm": 6.210817337036133,
      "learning_rate": 3.0383838383838387e-05,
      "loss": 0.3601,
      "step": 3984
    },
    {
      "epoch": 1.0269,
      "grad_norm": 6.294418811798096,
      "learning_rate": 3.037878787878788e-05,
      "loss": 1.141,
      "step": 3985
    },
    {
      "epoch": 1.027,
      "grad_norm": 5.956554889678955,
      "learning_rate": 3.0373737373737375e-05,
      "loss": 0.7598,
      "step": 3986
    },
    {
      "epoch": 1.0271,
      "grad_norm": 4.563986778259277,
      "learning_rate": 3.0368686868686868e-05,
      "loss": 0.1669,
      "step": 3987
    },
    {
      "epoch": 1.0272,
      "grad_norm": 5.970609188079834,
      "learning_rate": 3.0363636363636367e-05,
      "loss": 0.3805,
      "step": 3988
    },
    {
      "epoch": 1.0273,
      "grad_norm": 4.5003838539123535,
      "learning_rate": 3.035858585858586e-05,
      "loss": 0.2046,
      "step": 3989
    },
    {
      "epoch": 1.0274,
      "grad_norm": 9.080307006835938,
      "learning_rate": 3.0353535353535356e-05,
      "loss": 0.7411,
      "step": 3990
    },
    {
      "epoch": 1.0275,
      "grad_norm": 15.180051803588867,
      "learning_rate": 3.034848484848485e-05,
      "loss": 1.0523,
      "step": 3991
    },
    {
      "epoch": 1.0276,
      "grad_norm": 10.275618553161621,
      "learning_rate": 3.0343434343434345e-05,
      "loss": 0.2865,
      "step": 3992
    },
    {
      "epoch": 1.0277,
      "grad_norm": 6.294971942901611,
      "learning_rate": 3.0338383838383837e-05,
      "loss": 0.943,
      "step": 3993
    },
    {
      "epoch": 1.0278,
      "grad_norm": 3.849810838699341,
      "learning_rate": 3.0333333333333337e-05,
      "loss": 0.634,
      "step": 3994
    },
    {
      "epoch": 1.0279,
      "grad_norm": 6.992409706115723,
      "learning_rate": 3.032828282828283e-05,
      "loss": 1.1342,
      "step": 3995
    },
    {
      "epoch": 1.028,
      "grad_norm": 17.559682846069336,
      "learning_rate": 3.0323232323232325e-05,
      "loss": 0.5074,
      "step": 3996
    },
    {
      "epoch": 1.0281,
      "grad_norm": 14.259182929992676,
      "learning_rate": 3.0318181818181818e-05,
      "loss": 0.4624,
      "step": 3997
    },
    {
      "epoch": 1.0282,
      "grad_norm": 6.438292503356934,
      "learning_rate": 3.0313131313131317e-05,
      "loss": 0.4887,
      "step": 3998
    },
    {
      "epoch": 1.0283,
      "grad_norm": 4.289889812469482,
      "learning_rate": 3.0308080808080806e-05,
      "loss": 0.6345,
      "step": 3999
    },
    {
      "epoch": 1.0284,
      "grad_norm": 10.375534057617188,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.679,
      "step": 4000
    },
    {
      "epoch": 1.0285,
      "grad_norm": 6.938895225524902,
      "learning_rate": 3.02979797979798e-05,
      "loss": 0.6554,
      "step": 4001
    },
    {
      "epoch": 1.0286,
      "grad_norm": 15.274396896362305,
      "learning_rate": 3.0292929292929294e-05,
      "loss": 0.8143,
      "step": 4002
    },
    {
      "epoch": 1.0287,
      "grad_norm": 7.5656609535217285,
      "learning_rate": 3.0287878787878787e-05,
      "loss": 0.5219,
      "step": 4003
    },
    {
      "epoch": 1.0288,
      "grad_norm": 8.28139877319336,
      "learning_rate": 3.0282828282828286e-05,
      "loss": 0.9388,
      "step": 4004
    },
    {
      "epoch": 1.0289,
      "grad_norm": 4.888814449310303,
      "learning_rate": 3.0277777777777776e-05,
      "loss": 0.7167,
      "step": 4005
    },
    {
      "epoch": 1.029,
      "grad_norm": 10.642953872680664,
      "learning_rate": 3.0272727272727275e-05,
      "loss": 0.4055,
      "step": 4006
    },
    {
      "epoch": 1.0291,
      "grad_norm": 14.158931732177734,
      "learning_rate": 3.0267676767676767e-05,
      "loss": 1.5176,
      "step": 4007
    },
    {
      "epoch": 1.0292,
      "grad_norm": 5.6501264572143555,
      "learning_rate": 3.0262626262626263e-05,
      "loss": 0.5928,
      "step": 4008
    },
    {
      "epoch": 1.0293,
      "grad_norm": 12.481844902038574,
      "learning_rate": 3.0257575757575756e-05,
      "loss": 1.4228,
      "step": 4009
    },
    {
      "epoch": 1.0294,
      "grad_norm": 6.28749942779541,
      "learning_rate": 3.0252525252525255e-05,
      "loss": 1.0351,
      "step": 4010
    },
    {
      "epoch": 1.0295,
      "grad_norm": 8.865232467651367,
      "learning_rate": 3.0247474747474748e-05,
      "loss": 0.4023,
      "step": 4011
    },
    {
      "epoch": 1.0296,
      "grad_norm": 7.415183067321777,
      "learning_rate": 3.0242424242424244e-05,
      "loss": 0.651,
      "step": 4012
    },
    {
      "epoch": 1.0297,
      "grad_norm": 7.710961818695068,
      "learning_rate": 3.0237373737373737e-05,
      "loss": 0.3413,
      "step": 4013
    },
    {
      "epoch": 1.0298,
      "grad_norm": 5.000598907470703,
      "learning_rate": 3.0232323232323233e-05,
      "loss": 0.3524,
      "step": 4014
    },
    {
      "epoch": 1.0299,
      "grad_norm": 8.237907409667969,
      "learning_rate": 3.0227272727272725e-05,
      "loss": 1.3902,
      "step": 4015
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.733302116394043,
      "learning_rate": 3.0222222222222225e-05,
      "loss": 0.3453,
      "step": 4016
    },
    {
      "epoch": 1.0301,
      "grad_norm": 6.729673862457275,
      "learning_rate": 3.0217171717171717e-05,
      "loss": 0.5614,
      "step": 4017
    },
    {
      "epoch": 1.0302,
      "grad_norm": 12.53017520904541,
      "learning_rate": 3.0212121212121213e-05,
      "loss": 0.8672,
      "step": 4018
    },
    {
      "epoch": 1.0303,
      "grad_norm": 8.291069984436035,
      "learning_rate": 3.0207070707070706e-05,
      "loss": 1.1882,
      "step": 4019
    },
    {
      "epoch": 1.0304,
      "grad_norm": 11.580833435058594,
      "learning_rate": 3.0202020202020205e-05,
      "loss": 1.1306,
      "step": 4020
    },
    {
      "epoch": 1.0305,
      "grad_norm": 3.7210562229156494,
      "learning_rate": 3.0196969696969694e-05,
      "loss": 0.448,
      "step": 4021
    },
    {
      "epoch": 1.0306,
      "grad_norm": 9.084409713745117,
      "learning_rate": 3.0191919191919194e-05,
      "loss": 0.8054,
      "step": 4022
    },
    {
      "epoch": 1.0307,
      "grad_norm": 8.711088180541992,
      "learning_rate": 3.0186868686868686e-05,
      "loss": 0.5227,
      "step": 4023
    },
    {
      "epoch": 1.0308,
      "grad_norm": 3.9583494663238525,
      "learning_rate": 3.0181818181818182e-05,
      "loss": 0.9336,
      "step": 4024
    },
    {
      "epoch": 1.0309,
      "grad_norm": 5.931269645690918,
      "learning_rate": 3.0176767676767675e-05,
      "loss": 0.3962,
      "step": 4025
    },
    {
      "epoch": 1.031,
      "grad_norm": 9.125102043151855,
      "learning_rate": 3.0171717171717174e-05,
      "loss": 0.741,
      "step": 4026
    },
    {
      "epoch": 1.0311,
      "grad_norm": 6.799335479736328,
      "learning_rate": 3.016666666666667e-05,
      "loss": 0.8075,
      "step": 4027
    },
    {
      "epoch": 1.0312,
      "grad_norm": 4.335094451904297,
      "learning_rate": 3.0161616161616163e-05,
      "loss": 0.8808,
      "step": 4028
    },
    {
      "epoch": 1.0313,
      "grad_norm": 6.005122661590576,
      "learning_rate": 3.0156565656565662e-05,
      "loss": 0.7002,
      "step": 4029
    },
    {
      "epoch": 1.0314,
      "grad_norm": 5.853137493133545,
      "learning_rate": 3.015151515151515e-05,
      "loss": 0.4227,
      "step": 4030
    },
    {
      "epoch": 1.0315,
      "grad_norm": 9.792613983154297,
      "learning_rate": 3.014646464646465e-05,
      "loss": 0.9971,
      "step": 4031
    },
    {
      "epoch": 1.0316,
      "grad_norm": 3.200700283050537,
      "learning_rate": 3.0141414141414144e-05,
      "loss": 0.2138,
      "step": 4032
    },
    {
      "epoch": 1.0317,
      "grad_norm": 7.40233039855957,
      "learning_rate": 3.013636363636364e-05,
      "loss": 0.6759,
      "step": 4033
    },
    {
      "epoch": 1.0318,
      "grad_norm": 6.500619411468506,
      "learning_rate": 3.0131313131313132e-05,
      "loss": 0.4233,
      "step": 4034
    },
    {
      "epoch": 1.0319,
      "grad_norm": 5.793773651123047,
      "learning_rate": 3.012626262626263e-05,
      "loss": 0.433,
      "step": 4035
    },
    {
      "epoch": 1.032,
      "grad_norm": 10.664770126342773,
      "learning_rate": 3.012121212121212e-05,
      "loss": 0.2273,
      "step": 4036
    },
    {
      "epoch": 1.0321,
      "grad_norm": 4.939822196960449,
      "learning_rate": 3.011616161616162e-05,
      "loss": 0.857,
      "step": 4037
    },
    {
      "epoch": 1.0322,
      "grad_norm": 6.178371429443359,
      "learning_rate": 3.0111111111111113e-05,
      "loss": 0.5408,
      "step": 4038
    },
    {
      "epoch": 1.0323,
      "grad_norm": 6.248154163360596,
      "learning_rate": 3.010606060606061e-05,
      "loss": 0.7334,
      "step": 4039
    },
    {
      "epoch": 1.0324,
      "grad_norm": 2.879568099975586,
      "learning_rate": 3.01010101010101e-05,
      "loss": 0.1471,
      "step": 4040
    },
    {
      "epoch": 1.0325,
      "grad_norm": 4.312027454376221,
      "learning_rate": 3.00959595959596e-05,
      "loss": 0.5686,
      "step": 4041
    },
    {
      "epoch": 1.0326,
      "grad_norm": 5.288506507873535,
      "learning_rate": 3.0090909090909093e-05,
      "loss": 0.4786,
      "step": 4042
    },
    {
      "epoch": 1.0327,
      "grad_norm": 7.600947380065918,
      "learning_rate": 3.008585858585859e-05,
      "loss": 0.517,
      "step": 4043
    },
    {
      "epoch": 1.0328,
      "grad_norm": 8.863582611083984,
      "learning_rate": 3.0080808080808082e-05,
      "loss": 0.2105,
      "step": 4044
    },
    {
      "epoch": 1.0329,
      "grad_norm": 9.737042427062988,
      "learning_rate": 3.0075757575757578e-05,
      "loss": 0.9147,
      "step": 4045
    },
    {
      "epoch": 1.033,
      "grad_norm": 6.147843837738037,
      "learning_rate": 3.007070707070707e-05,
      "loss": 0.4562,
      "step": 4046
    },
    {
      "epoch": 1.0331,
      "grad_norm": 8.165909767150879,
      "learning_rate": 3.006565656565657e-05,
      "loss": 0.3795,
      "step": 4047
    },
    {
      "epoch": 1.0332,
      "grad_norm": 3.592676877975464,
      "learning_rate": 3.0060606060606062e-05,
      "loss": 0.3347,
      "step": 4048
    },
    {
      "epoch": 1.0333,
      "grad_norm": 4.987141132354736,
      "learning_rate": 3.005555555555556e-05,
      "loss": 0.8675,
      "step": 4049
    },
    {
      "epoch": 1.0334,
      "grad_norm": 7.952305793762207,
      "learning_rate": 3.005050505050505e-05,
      "loss": 0.8548,
      "step": 4050
    },
    {
      "epoch": 1.0335,
      "grad_norm": 9.477084159851074,
      "learning_rate": 3.004545454545455e-05,
      "loss": 0.8733,
      "step": 4051
    },
    {
      "epoch": 1.0336,
      "grad_norm": 6.677531719207764,
      "learning_rate": 3.004040404040404e-05,
      "loss": 0.3823,
      "step": 4052
    },
    {
      "epoch": 1.0337,
      "grad_norm": 4.21234655380249,
      "learning_rate": 3.003535353535354e-05,
      "loss": 0.1619,
      "step": 4053
    },
    {
      "epoch": 1.0338,
      "grad_norm": 5.296980857849121,
      "learning_rate": 3.003030303030303e-05,
      "loss": 0.4034,
      "step": 4054
    },
    {
      "epoch": 1.0339,
      "grad_norm": 12.170886993408203,
      "learning_rate": 3.0025252525252528e-05,
      "loss": 0.6982,
      "step": 4055
    },
    {
      "epoch": 1.034,
      "grad_norm": 5.331367015838623,
      "learning_rate": 3.002020202020202e-05,
      "loss": 0.4878,
      "step": 4056
    },
    {
      "epoch": 1.0341,
      "grad_norm": 5.928987503051758,
      "learning_rate": 3.001515151515152e-05,
      "loss": 0.3969,
      "step": 4057
    },
    {
      "epoch": 1.0342,
      "grad_norm": 2.991994857788086,
      "learning_rate": 3.0010101010101012e-05,
      "loss": 0.134,
      "step": 4058
    },
    {
      "epoch": 1.0343,
      "grad_norm": 5.722099781036377,
      "learning_rate": 3.0005050505050508e-05,
      "loss": 0.7091,
      "step": 4059
    },
    {
      "epoch": 1.0344,
      "grad_norm": 12.231318473815918,
      "learning_rate": 3e-05,
      "loss": 0.7988,
      "step": 4060
    },
    {
      "epoch": 1.0345,
      "grad_norm": 5.849119663238525,
      "learning_rate": 2.9994949494949497e-05,
      "loss": 0.9216,
      "step": 4061
    },
    {
      "epoch": 1.0346,
      "grad_norm": 3.5547866821289062,
      "learning_rate": 2.998989898989899e-05,
      "loss": 1.122,
      "step": 4062
    },
    {
      "epoch": 1.0347,
      "grad_norm": 5.451479911804199,
      "learning_rate": 2.998484848484849e-05,
      "loss": 0.4642,
      "step": 4063
    },
    {
      "epoch": 1.0348,
      "grad_norm": 6.448863506317139,
      "learning_rate": 2.997979797979798e-05,
      "loss": 0.9592,
      "step": 4064
    },
    {
      "epoch": 1.0349,
      "grad_norm": 6.03212308883667,
      "learning_rate": 2.9974747474747477e-05,
      "loss": 0.972,
      "step": 4065
    },
    {
      "epoch": 1.035,
      "grad_norm": 7.2537736892700195,
      "learning_rate": 2.996969696969697e-05,
      "loss": 0.4463,
      "step": 4066
    },
    {
      "epoch": 1.0351,
      "grad_norm": 7.429289817810059,
      "learning_rate": 2.9964646464646466e-05,
      "loss": 0.6293,
      "step": 4067
    },
    {
      "epoch": 1.0352,
      "grad_norm": 7.728455066680908,
      "learning_rate": 2.995959595959596e-05,
      "loss": 0.8631,
      "step": 4068
    },
    {
      "epoch": 1.0352999999999999,
      "grad_norm": 5.054967880249023,
      "learning_rate": 2.9954545454545458e-05,
      "loss": 0.3701,
      "step": 4069
    },
    {
      "epoch": 1.0354,
      "grad_norm": 12.66554069519043,
      "learning_rate": 2.994949494949495e-05,
      "loss": 0.5379,
      "step": 4070
    },
    {
      "epoch": 1.0355,
      "grad_norm": 4.5191144943237305,
      "learning_rate": 2.9944444444444446e-05,
      "loss": 0.5189,
      "step": 4071
    },
    {
      "epoch": 1.0356,
      "grad_norm": 21.06710433959961,
      "learning_rate": 2.993939393939394e-05,
      "loss": 0.614,
      "step": 4072
    },
    {
      "epoch": 1.0357,
      "grad_norm": 4.763858795166016,
      "learning_rate": 2.993434343434344e-05,
      "loss": 0.2368,
      "step": 4073
    },
    {
      "epoch": 1.0358,
      "grad_norm": 5.604577541351318,
      "learning_rate": 2.9929292929292928e-05,
      "loss": 0.783,
      "step": 4074
    },
    {
      "epoch": 1.0359,
      "grad_norm": 4.920997142791748,
      "learning_rate": 2.9924242424242427e-05,
      "loss": 0.5,
      "step": 4075
    },
    {
      "epoch": 1.036,
      "grad_norm": 5.150906562805176,
      "learning_rate": 2.991919191919192e-05,
      "loss": 0.4271,
      "step": 4076
    },
    {
      "epoch": 1.0361,
      "grad_norm": 7.251683235168457,
      "learning_rate": 2.9914141414141416e-05,
      "loss": 0.5465,
      "step": 4077
    },
    {
      "epoch": 1.0362,
      "grad_norm": 5.280575275421143,
      "learning_rate": 2.9909090909090908e-05,
      "loss": 1.1242,
      "step": 4078
    },
    {
      "epoch": 1.0363,
      "grad_norm": 8.124526023864746,
      "learning_rate": 2.9904040404040408e-05,
      "loss": 0.3448,
      "step": 4079
    },
    {
      "epoch": 1.0364,
      "grad_norm": 26.662755966186523,
      "learning_rate": 2.98989898989899e-05,
      "loss": 0.5449,
      "step": 4080
    },
    {
      "epoch": 1.0365,
      "grad_norm": 9.793127059936523,
      "learning_rate": 2.9893939393939396e-05,
      "loss": 0.5659,
      "step": 4081
    },
    {
      "epoch": 1.0366,
      "grad_norm": 7.197762489318848,
      "learning_rate": 2.988888888888889e-05,
      "loss": 0.5586,
      "step": 4082
    },
    {
      "epoch": 1.0367,
      "grad_norm": 8.751152038574219,
      "learning_rate": 2.9883838383838385e-05,
      "loss": 0.5387,
      "step": 4083
    },
    {
      "epoch": 1.0368,
      "grad_norm": 4.54669713973999,
      "learning_rate": 2.9878787878787877e-05,
      "loss": 0.6451,
      "step": 4084
    },
    {
      "epoch": 1.0369,
      "grad_norm": 4.352379322052002,
      "learning_rate": 2.9873737373737377e-05,
      "loss": 0.2727,
      "step": 4085
    },
    {
      "epoch": 1.037,
      "grad_norm": 7.4315361976623535,
      "learning_rate": 2.986868686868687e-05,
      "loss": 0.3967,
      "step": 4086
    },
    {
      "epoch": 1.0371,
      "grad_norm": 4.984579563140869,
      "learning_rate": 2.9863636363636365e-05,
      "loss": 0.7419,
      "step": 4087
    },
    {
      "epoch": 1.0372,
      "grad_norm": 8.527029991149902,
      "learning_rate": 2.9858585858585858e-05,
      "loss": 0.5862,
      "step": 4088
    },
    {
      "epoch": 1.0373,
      "grad_norm": 16.58534812927246,
      "learning_rate": 2.9853535353535357e-05,
      "loss": 0.6514,
      "step": 4089
    },
    {
      "epoch": 1.0374,
      "grad_norm": 7.799651622772217,
      "learning_rate": 2.9848484848484847e-05,
      "loss": 1.0063,
      "step": 4090
    },
    {
      "epoch": 1.0375,
      "grad_norm": 12.443809509277344,
      "learning_rate": 2.9843434343434346e-05,
      "loss": 0.7842,
      "step": 4091
    },
    {
      "epoch": 1.0376,
      "grad_norm": 7.543701648712158,
      "learning_rate": 2.983838383838384e-05,
      "loss": 1.2635,
      "step": 4092
    },
    {
      "epoch": 1.0377,
      "grad_norm": 21.397493362426758,
      "learning_rate": 2.9833333333333335e-05,
      "loss": 0.8599,
      "step": 4093
    },
    {
      "epoch": 1.0378,
      "grad_norm": 7.0754523277282715,
      "learning_rate": 2.9828282828282827e-05,
      "loss": 0.7818,
      "step": 4094
    },
    {
      "epoch": 1.0379,
      "grad_norm": 13.105072975158691,
      "learning_rate": 2.9823232323232327e-05,
      "loss": 0.5939,
      "step": 4095
    },
    {
      "epoch": 1.038,
      "grad_norm": 15.526625633239746,
      "learning_rate": 2.9818181818181816e-05,
      "loss": 0.276,
      "step": 4096
    },
    {
      "epoch": 1.0381,
      "grad_norm": 8.034290313720703,
      "learning_rate": 2.9813131313131315e-05,
      "loss": 0.2222,
      "step": 4097
    },
    {
      "epoch": 1.0382,
      "grad_norm": 19.38471031188965,
      "learning_rate": 2.9808080808080808e-05,
      "loss": 0.8779,
      "step": 4098
    },
    {
      "epoch": 1.0383,
      "grad_norm": 6.198659420013428,
      "learning_rate": 2.9803030303030304e-05,
      "loss": 0.3898,
      "step": 4099
    },
    {
      "epoch": 1.0384,
      "grad_norm": 3.3419666290283203,
      "learning_rate": 2.9797979797979796e-05,
      "loss": 0.1646,
      "step": 4100
    },
    {
      "epoch": 1.0385,
      "grad_norm": 5.897149085998535,
      "learning_rate": 2.9792929292929296e-05,
      "loss": 0.4008,
      "step": 4101
    },
    {
      "epoch": 1.0386,
      "grad_norm": 6.503688335418701,
      "learning_rate": 2.978787878787879e-05,
      "loss": 0.8696,
      "step": 4102
    },
    {
      "epoch": 1.0387,
      "grad_norm": 2.4210050106048584,
      "learning_rate": 2.9782828282828284e-05,
      "loss": 0.3056,
      "step": 4103
    },
    {
      "epoch": 1.0388,
      "grad_norm": 10.093828201293945,
      "learning_rate": 2.9777777777777777e-05,
      "loss": 0.5128,
      "step": 4104
    },
    {
      "epoch": 1.0389,
      "grad_norm": 8.148223876953125,
      "learning_rate": 2.9772727272727273e-05,
      "loss": 0.4008,
      "step": 4105
    },
    {
      "epoch": 1.039,
      "grad_norm": 4.817931175231934,
      "learning_rate": 2.9767676767676765e-05,
      "loss": 0.2986,
      "step": 4106
    },
    {
      "epoch": 1.0391,
      "grad_norm": 8.451888084411621,
      "learning_rate": 2.9762626262626265e-05,
      "loss": 0.3298,
      "step": 4107
    },
    {
      "epoch": 1.0392,
      "grad_norm": 6.94677209854126,
      "learning_rate": 2.9757575757575757e-05,
      "loss": 0.7588,
      "step": 4108
    },
    {
      "epoch": 1.0393,
      "grad_norm": 18.81797981262207,
      "learning_rate": 2.9752525252525253e-05,
      "loss": 0.7013,
      "step": 4109
    },
    {
      "epoch": 1.0394,
      "grad_norm": 8.786355018615723,
      "learning_rate": 2.9747474747474746e-05,
      "loss": 0.3671,
      "step": 4110
    },
    {
      "epoch": 1.0395,
      "grad_norm": 8.0643310546875,
      "learning_rate": 2.9742424242424245e-05,
      "loss": 1.0616,
      "step": 4111
    },
    {
      "epoch": 1.0396,
      "grad_norm": 7.77587890625,
      "learning_rate": 2.973737373737374e-05,
      "loss": 0.6731,
      "step": 4112
    },
    {
      "epoch": 1.0397,
      "grad_norm": 5.746932029724121,
      "learning_rate": 2.9732323232323234e-05,
      "loss": 0.9833,
      "step": 4113
    },
    {
      "epoch": 1.0398,
      "grad_norm": 9.74528980255127,
      "learning_rate": 2.972727272727273e-05,
      "loss": 0.3355,
      "step": 4114
    },
    {
      "epoch": 1.0399,
      "grad_norm": 6.8110527992248535,
      "learning_rate": 2.9722222222222223e-05,
      "loss": 0.9428,
      "step": 4115
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.212206840515137,
      "learning_rate": 2.9717171717171722e-05,
      "loss": 0.6729,
      "step": 4116
    },
    {
      "epoch": 1.0401,
      "grad_norm": 10.913423538208008,
      "learning_rate": 2.9712121212121215e-05,
      "loss": 0.3581,
      "step": 4117
    },
    {
      "epoch": 1.0402,
      "grad_norm": 9.16326904296875,
      "learning_rate": 2.970707070707071e-05,
      "loss": 1.0239,
      "step": 4118
    },
    {
      "epoch": 1.0403,
      "grad_norm": 5.541441917419434,
      "learning_rate": 2.9702020202020203e-05,
      "loss": 0.2684,
      "step": 4119
    },
    {
      "epoch": 1.0404,
      "grad_norm": 17.40587615966797,
      "learning_rate": 2.96969696969697e-05,
      "loss": 0.9491,
      "step": 4120
    },
    {
      "epoch": 1.0405,
      "grad_norm": 7.430034160614014,
      "learning_rate": 2.9691919191919192e-05,
      "loss": 0.5241,
      "step": 4121
    },
    {
      "epoch": 1.0406,
      "grad_norm": 15.721673965454102,
      "learning_rate": 2.968686868686869e-05,
      "loss": 0.8825,
      "step": 4122
    },
    {
      "epoch": 1.0407,
      "grad_norm": 9.215581893920898,
      "learning_rate": 2.9681818181818184e-05,
      "loss": 0.8408,
      "step": 4123
    },
    {
      "epoch": 1.0408,
      "grad_norm": 11.3944673538208,
      "learning_rate": 2.967676767676768e-05,
      "loss": 2.2838,
      "step": 4124
    },
    {
      "epoch": 1.0409,
      "grad_norm": 7.256676197052002,
      "learning_rate": 2.9671717171717172e-05,
      "loss": 0.4467,
      "step": 4125
    },
    {
      "epoch": 1.041,
      "grad_norm": 6.775428295135498,
      "learning_rate": 2.9666666666666672e-05,
      "loss": 1.0605,
      "step": 4126
    },
    {
      "epoch": 1.0411,
      "grad_norm": 7.416471481323242,
      "learning_rate": 2.966161616161616e-05,
      "loss": 0.8572,
      "step": 4127
    },
    {
      "epoch": 1.0412,
      "grad_norm": 8.358333587646484,
      "learning_rate": 2.965656565656566e-05,
      "loss": 0.6073,
      "step": 4128
    },
    {
      "epoch": 1.0413000000000001,
      "grad_norm": 10.257528305053711,
      "learning_rate": 2.9651515151515153e-05,
      "loss": 0.7922,
      "step": 4129
    },
    {
      "epoch": 1.0414,
      "grad_norm": 24.881980895996094,
      "learning_rate": 2.964646464646465e-05,
      "loss": 1.27,
      "step": 4130
    },
    {
      "epoch": 1.0415,
      "grad_norm": 6.534733772277832,
      "learning_rate": 2.964141414141414e-05,
      "loss": 1.2551,
      "step": 4131
    },
    {
      "epoch": 1.0416,
      "grad_norm": 7.046194076538086,
      "learning_rate": 2.963636363636364e-05,
      "loss": 0.2514,
      "step": 4132
    },
    {
      "epoch": 1.0417,
      "grad_norm": 4.762508869171143,
      "learning_rate": 2.9631313131313134e-05,
      "loss": 0.3786,
      "step": 4133
    },
    {
      "epoch": 1.0418,
      "grad_norm": 6.053603649139404,
      "learning_rate": 2.962626262626263e-05,
      "loss": 0.7543,
      "step": 4134
    },
    {
      "epoch": 1.0419,
      "grad_norm": 7.666287899017334,
      "learning_rate": 2.9621212121212122e-05,
      "loss": 0.7561,
      "step": 4135
    },
    {
      "epoch": 1.042,
      "grad_norm": 13.343504905700684,
      "learning_rate": 2.9616161616161618e-05,
      "loss": 0.8841,
      "step": 4136
    },
    {
      "epoch": 1.0421,
      "grad_norm": 8.444512367248535,
      "learning_rate": 2.961111111111111e-05,
      "loss": 0.5951,
      "step": 4137
    },
    {
      "epoch": 1.0422,
      "grad_norm": 6.113968372344971,
      "learning_rate": 2.960606060606061e-05,
      "loss": 0.4066,
      "step": 4138
    },
    {
      "epoch": 1.0423,
      "grad_norm": 3.0951993465423584,
      "learning_rate": 2.9601010101010103e-05,
      "loss": 0.2985,
      "step": 4139
    },
    {
      "epoch": 1.0424,
      "grad_norm": 15.125493049621582,
      "learning_rate": 2.95959595959596e-05,
      "loss": 0.4583,
      "step": 4140
    },
    {
      "epoch": 1.0425,
      "grad_norm": 5.1565351486206055,
      "learning_rate": 2.959090909090909e-05,
      "loss": 1.205,
      "step": 4141
    },
    {
      "epoch": 1.0426,
      "grad_norm": 5.573177337646484,
      "learning_rate": 2.958585858585859e-05,
      "loss": 0.7562,
      "step": 4142
    },
    {
      "epoch": 1.0427,
      "grad_norm": 6.750942230224609,
      "learning_rate": 2.958080808080808e-05,
      "loss": 1.3416,
      "step": 4143
    },
    {
      "epoch": 1.0428,
      "grad_norm": 6.916966438293457,
      "learning_rate": 2.957575757575758e-05,
      "loss": 0.5646,
      "step": 4144
    },
    {
      "epoch": 1.0429,
      "grad_norm": 8.7824125289917,
      "learning_rate": 2.9570707070707072e-05,
      "loss": 0.5823,
      "step": 4145
    },
    {
      "epoch": 1.043,
      "grad_norm": 7.132483959197998,
      "learning_rate": 2.9565656565656568e-05,
      "loss": 0.8007,
      "step": 4146
    },
    {
      "epoch": 1.0431,
      "grad_norm": 7.306180000305176,
      "learning_rate": 2.956060606060606e-05,
      "loss": 0.7224,
      "step": 4147
    },
    {
      "epoch": 1.0432,
      "grad_norm": 8.755817413330078,
      "learning_rate": 2.955555555555556e-05,
      "loss": 0.6651,
      "step": 4148
    },
    {
      "epoch": 1.0433,
      "grad_norm": 5.711585998535156,
      "learning_rate": 2.955050505050505e-05,
      "loss": 0.3146,
      "step": 4149
    },
    {
      "epoch": 1.0434,
      "grad_norm": 4.43522834777832,
      "learning_rate": 2.954545454545455e-05,
      "loss": 0.3754,
      "step": 4150
    },
    {
      "epoch": 1.0435,
      "grad_norm": 6.8158040046691895,
      "learning_rate": 2.954040404040404e-05,
      "loss": 0.412,
      "step": 4151
    },
    {
      "epoch": 1.0436,
      "grad_norm": 4.1484599113464355,
      "learning_rate": 2.9535353535353537e-05,
      "loss": 0.285,
      "step": 4152
    },
    {
      "epoch": 1.0437,
      "grad_norm": 8.6920804977417,
      "learning_rate": 2.953030303030303e-05,
      "loss": 0.4202,
      "step": 4153
    },
    {
      "epoch": 1.0438,
      "grad_norm": 4.879558086395264,
      "learning_rate": 2.952525252525253e-05,
      "loss": 0.3469,
      "step": 4154
    },
    {
      "epoch": 1.0439,
      "grad_norm": 6.556965351104736,
      "learning_rate": 2.952020202020202e-05,
      "loss": 0.3366,
      "step": 4155
    },
    {
      "epoch": 1.044,
      "grad_norm": 4.702517509460449,
      "learning_rate": 2.9515151515151518e-05,
      "loss": 0.3396,
      "step": 4156
    },
    {
      "epoch": 1.0441,
      "grad_norm": 3.3897950649261475,
      "learning_rate": 2.951010101010101e-05,
      "loss": 1.1671,
      "step": 4157
    },
    {
      "epoch": 1.0442,
      "grad_norm": 9.993147850036621,
      "learning_rate": 2.9505050505050506e-05,
      "loss": 0.5215,
      "step": 4158
    },
    {
      "epoch": 1.0443,
      "grad_norm": 6.592219829559326,
      "learning_rate": 2.95e-05,
      "loss": 0.5426,
      "step": 4159
    },
    {
      "epoch": 1.0444,
      "grad_norm": 5.768011093139648,
      "learning_rate": 2.9494949494949498e-05,
      "loss": 0.6741,
      "step": 4160
    },
    {
      "epoch": 1.0445,
      "grad_norm": 4.840683937072754,
      "learning_rate": 2.948989898989899e-05,
      "loss": 0.3993,
      "step": 4161
    },
    {
      "epoch": 1.0446,
      "grad_norm": 13.70383358001709,
      "learning_rate": 2.9484848484848487e-05,
      "loss": 0.3569,
      "step": 4162
    },
    {
      "epoch": 1.0447,
      "grad_norm": 6.604201793670654,
      "learning_rate": 2.947979797979798e-05,
      "loss": 0.393,
      "step": 4163
    },
    {
      "epoch": 1.0448,
      "grad_norm": 12.80690860748291,
      "learning_rate": 2.947474747474748e-05,
      "loss": 1.0822,
      "step": 4164
    },
    {
      "epoch": 1.0449,
      "grad_norm": 6.629452705383301,
      "learning_rate": 2.9469696969696968e-05,
      "loss": 0.3391,
      "step": 4165
    },
    {
      "epoch": 1.045,
      "grad_norm": 4.8941650390625,
      "learning_rate": 2.9464646464646467e-05,
      "loss": 0.6262,
      "step": 4166
    },
    {
      "epoch": 1.0451,
      "grad_norm": 7.90565299987793,
      "learning_rate": 2.945959595959596e-05,
      "loss": 0.5724,
      "step": 4167
    },
    {
      "epoch": 1.0452,
      "grad_norm": 9.535300254821777,
      "learning_rate": 2.9454545454545456e-05,
      "loss": 0.5491,
      "step": 4168
    },
    {
      "epoch": 1.0453,
      "grad_norm": 3.99265718460083,
      "learning_rate": 2.944949494949495e-05,
      "loss": 0.2175,
      "step": 4169
    },
    {
      "epoch": 1.0454,
      "grad_norm": 7.831755638122559,
      "learning_rate": 2.9444444444444448e-05,
      "loss": 0.7294,
      "step": 4170
    },
    {
      "epoch": 1.0455,
      "grad_norm": 7.5254716873168945,
      "learning_rate": 2.943939393939394e-05,
      "loss": 0.6805,
      "step": 4171
    },
    {
      "epoch": 1.0456,
      "grad_norm": 5.043066501617432,
      "learning_rate": 2.9434343434343436e-05,
      "loss": 0.4466,
      "step": 4172
    },
    {
      "epoch": 1.0457,
      "grad_norm": 9.933745384216309,
      "learning_rate": 2.942929292929293e-05,
      "loss": 1.0681,
      "step": 4173
    },
    {
      "epoch": 1.0458,
      "grad_norm": 4.495931148529053,
      "learning_rate": 2.9424242424242425e-05,
      "loss": 0.5345,
      "step": 4174
    },
    {
      "epoch": 1.0459,
      "grad_norm": 6.620901584625244,
      "learning_rate": 2.9419191919191918e-05,
      "loss": 0.4018,
      "step": 4175
    },
    {
      "epoch": 1.046,
      "grad_norm": 8.848827362060547,
      "learning_rate": 2.9414141414141417e-05,
      "loss": 0.4871,
      "step": 4176
    },
    {
      "epoch": 1.0461,
      "grad_norm": 16.481487274169922,
      "learning_rate": 2.940909090909091e-05,
      "loss": 0.5446,
      "step": 4177
    },
    {
      "epoch": 1.0462,
      "grad_norm": 5.052426338195801,
      "learning_rate": 2.9404040404040406e-05,
      "loss": 0.8542,
      "step": 4178
    },
    {
      "epoch": 1.0463,
      "grad_norm": 8.675429344177246,
      "learning_rate": 2.9398989898989898e-05,
      "loss": 0.5146,
      "step": 4179
    },
    {
      "epoch": 1.0464,
      "grad_norm": 5.696794033050537,
      "learning_rate": 2.9393939393939394e-05,
      "loss": 0.5008,
      "step": 4180
    },
    {
      "epoch": 1.0465,
      "grad_norm": 6.092592239379883,
      "learning_rate": 2.9388888888888887e-05,
      "loss": 0.4199,
      "step": 4181
    },
    {
      "epoch": 1.0466,
      "grad_norm": 4.1852312088012695,
      "learning_rate": 2.9383838383838386e-05,
      "loss": 0.3844,
      "step": 4182
    },
    {
      "epoch": 1.0467,
      "grad_norm": 5.333979606628418,
      "learning_rate": 2.937878787878788e-05,
      "loss": 0.3951,
      "step": 4183
    },
    {
      "epoch": 1.0468,
      "grad_norm": 4.957526206970215,
      "learning_rate": 2.9373737373737375e-05,
      "loss": 0.3564,
      "step": 4184
    },
    {
      "epoch": 1.0469,
      "grad_norm": 7.4872517585754395,
      "learning_rate": 2.9368686868686867e-05,
      "loss": 1.0139,
      "step": 4185
    },
    {
      "epoch": 1.047,
      "grad_norm": 4.100264072418213,
      "learning_rate": 2.9363636363636367e-05,
      "loss": 0.2219,
      "step": 4186
    },
    {
      "epoch": 1.0471,
      "grad_norm": 6.013932704925537,
      "learning_rate": 2.9358585858585856e-05,
      "loss": 0.8121,
      "step": 4187
    },
    {
      "epoch": 1.0472,
      "grad_norm": 6.230075836181641,
      "learning_rate": 2.9353535353535355e-05,
      "loss": 0.3747,
      "step": 4188
    },
    {
      "epoch": 1.0473,
      "grad_norm": 7.44619607925415,
      "learning_rate": 2.9348484848484848e-05,
      "loss": 1.1094,
      "step": 4189
    },
    {
      "epoch": 1.0474,
      "grad_norm": 7.460475444793701,
      "learning_rate": 2.9343434343434344e-05,
      "loss": 0.3475,
      "step": 4190
    },
    {
      "epoch": 1.0475,
      "grad_norm": 6.530074596405029,
      "learning_rate": 2.9338383838383837e-05,
      "loss": 0.5077,
      "step": 4191
    },
    {
      "epoch": 1.0476,
      "grad_norm": 6.624098300933838,
      "learning_rate": 2.9333333333333336e-05,
      "loss": 0.3628,
      "step": 4192
    },
    {
      "epoch": 1.0477,
      "grad_norm": 6.552068710327148,
      "learning_rate": 2.932828282828283e-05,
      "loss": 0.5258,
      "step": 4193
    },
    {
      "epoch": 1.0478,
      "grad_norm": 4.137850284576416,
      "learning_rate": 2.9323232323232325e-05,
      "loss": 0.4241,
      "step": 4194
    },
    {
      "epoch": 1.0479,
      "grad_norm": 6.386784553527832,
      "learning_rate": 2.9318181818181817e-05,
      "loss": 0.3423,
      "step": 4195
    },
    {
      "epoch": 1.048,
      "grad_norm": 6.440220832824707,
      "learning_rate": 2.9313131313131313e-05,
      "loss": 0.4484,
      "step": 4196
    },
    {
      "epoch": 1.0481,
      "grad_norm": 5.273252964019775,
      "learning_rate": 2.9308080808080813e-05,
      "loss": 0.4701,
      "step": 4197
    },
    {
      "epoch": 1.0482,
      "grad_norm": 5.74713134765625,
      "learning_rate": 2.9303030303030305e-05,
      "loss": 0.4368,
      "step": 4198
    },
    {
      "epoch": 1.0483,
      "grad_norm": 7.223015308380127,
      "learning_rate": 2.92979797979798e-05,
      "loss": 0.8719,
      "step": 4199
    },
    {
      "epoch": 1.0484,
      "grad_norm": 7.6462602615356445,
      "learning_rate": 2.9292929292929294e-05,
      "loss": 0.487,
      "step": 4200
    },
    {
      "epoch": 1.0485,
      "grad_norm": 3.84702730178833,
      "learning_rate": 2.9287878787878793e-05,
      "loss": 0.5996,
      "step": 4201
    },
    {
      "epoch": 1.0486,
      "grad_norm": 3.3840649127960205,
      "learning_rate": 2.9282828282828286e-05,
      "loss": 0.3723,
      "step": 4202
    },
    {
      "epoch": 1.0487,
      "grad_norm": 8.651215553283691,
      "learning_rate": 2.927777777777778e-05,
      "loss": 0.4393,
      "step": 4203
    },
    {
      "epoch": 1.0488,
      "grad_norm": 8.691407203674316,
      "learning_rate": 2.9272727272727274e-05,
      "loss": 0.46,
      "step": 4204
    },
    {
      "epoch": 1.0489,
      "grad_norm": 8.427436828613281,
      "learning_rate": 2.926767676767677e-05,
      "loss": 0.6624,
      "step": 4205
    },
    {
      "epoch": 1.049,
      "grad_norm": 7.28903865814209,
      "learning_rate": 2.9262626262626263e-05,
      "loss": 0.3557,
      "step": 4206
    },
    {
      "epoch": 1.0491,
      "grad_norm": 7.397300720214844,
      "learning_rate": 2.9257575757575762e-05,
      "loss": 0.4682,
      "step": 4207
    },
    {
      "epoch": 1.0492,
      "grad_norm": 7.145303726196289,
      "learning_rate": 2.9252525252525255e-05,
      "loss": 0.8417,
      "step": 4208
    },
    {
      "epoch": 1.0493,
      "grad_norm": 9.54004955291748,
      "learning_rate": 2.924747474747475e-05,
      "loss": 1.2752,
      "step": 4209
    },
    {
      "epoch": 1.0493999999999999,
      "grad_norm": 32.736515045166016,
      "learning_rate": 2.9242424242424243e-05,
      "loss": 0.537,
      "step": 4210
    },
    {
      "epoch": 1.0495,
      "grad_norm": 3.823988676071167,
      "learning_rate": 2.923737373737374e-05,
      "loss": 1.2876,
      "step": 4211
    },
    {
      "epoch": 1.0496,
      "grad_norm": 5.3900346755981445,
      "learning_rate": 2.9232323232323232e-05,
      "loss": 0.6121,
      "step": 4212
    },
    {
      "epoch": 1.0497,
      "grad_norm": 12.094615936279297,
      "learning_rate": 2.922727272727273e-05,
      "loss": 1.1522,
      "step": 4213
    },
    {
      "epoch": 1.0498,
      "grad_norm": 7.81538200378418,
      "learning_rate": 2.9222222222222224e-05,
      "loss": 0.9563,
      "step": 4214
    },
    {
      "epoch": 1.0499,
      "grad_norm": 5.276789665222168,
      "learning_rate": 2.921717171717172e-05,
      "loss": 0.8752,
      "step": 4215
    },
    {
      "epoch": 1.05,
      "grad_norm": 5.531773567199707,
      "learning_rate": 2.9212121212121213e-05,
      "loss": 1.0016,
      "step": 4216
    },
    {
      "epoch": 1.0501,
      "grad_norm": 8.825020790100098,
      "learning_rate": 2.9207070707070712e-05,
      "loss": 0.5132,
      "step": 4217
    },
    {
      "epoch": 1.0502,
      "grad_norm": 7.478699684143066,
      "learning_rate": 2.92020202020202e-05,
      "loss": 0.5592,
      "step": 4218
    },
    {
      "epoch": 1.0503,
      "grad_norm": 8.587282180786133,
      "learning_rate": 2.91969696969697e-05,
      "loss": 0.599,
      "step": 4219
    },
    {
      "epoch": 1.0504,
      "grad_norm": 8.801159858703613,
      "learning_rate": 2.9191919191919193e-05,
      "loss": 0.5497,
      "step": 4220
    },
    {
      "epoch": 1.0505,
      "grad_norm": 4.622445106506348,
      "learning_rate": 2.918686868686869e-05,
      "loss": 0.433,
      "step": 4221
    },
    {
      "epoch": 1.0506,
      "grad_norm": 7.428670406341553,
      "learning_rate": 2.9181818181818182e-05,
      "loss": 0.5253,
      "step": 4222
    },
    {
      "epoch": 1.0507,
      "grad_norm": 4.507007598876953,
      "learning_rate": 2.917676767676768e-05,
      "loss": 0.5057,
      "step": 4223
    },
    {
      "epoch": 1.0508,
      "grad_norm": 5.032844543457031,
      "learning_rate": 2.9171717171717174e-05,
      "loss": 0.205,
      "step": 4224
    },
    {
      "epoch": 1.0509,
      "grad_norm": 4.277171611785889,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.2603,
      "step": 4225
    },
    {
      "epoch": 1.051,
      "grad_norm": 10.318939208984375,
      "learning_rate": 2.9161616161616162e-05,
      "loss": 0.2644,
      "step": 4226
    },
    {
      "epoch": 1.0511,
      "grad_norm": 4.987938404083252,
      "learning_rate": 2.915656565656566e-05,
      "loss": 0.4717,
      "step": 4227
    },
    {
      "epoch": 1.0512,
      "grad_norm": 8.314927101135254,
      "learning_rate": 2.915151515151515e-05,
      "loss": 1.2589,
      "step": 4228
    },
    {
      "epoch": 1.0513,
      "grad_norm": 5.172833442687988,
      "learning_rate": 2.914646464646465e-05,
      "loss": 0.4671,
      "step": 4229
    },
    {
      "epoch": 1.0514000000000001,
      "grad_norm": 9.357582092285156,
      "learning_rate": 2.9141414141414143e-05,
      "loss": 0.6353,
      "step": 4230
    },
    {
      "epoch": 1.0515,
      "grad_norm": 9.488184928894043,
      "learning_rate": 2.913636363636364e-05,
      "loss": 0.6576,
      "step": 4231
    },
    {
      "epoch": 1.0516,
      "grad_norm": 7.814052104949951,
      "learning_rate": 2.913131313131313e-05,
      "loss": 0.4758,
      "step": 4232
    },
    {
      "epoch": 1.0517,
      "grad_norm": 4.5167717933654785,
      "learning_rate": 2.912626262626263e-05,
      "loss": 0.3843,
      "step": 4233
    },
    {
      "epoch": 1.0518,
      "grad_norm": 6.040778160095215,
      "learning_rate": 2.912121212121212e-05,
      "loss": 0.6272,
      "step": 4234
    },
    {
      "epoch": 1.0519,
      "grad_norm": 5.324766159057617,
      "learning_rate": 2.911616161616162e-05,
      "loss": 0.4356,
      "step": 4235
    },
    {
      "epoch": 1.052,
      "grad_norm": 7.822668075561523,
      "learning_rate": 2.9111111111111112e-05,
      "loss": 0.8863,
      "step": 4236
    },
    {
      "epoch": 1.0521,
      "grad_norm": 5.715809345245361,
      "learning_rate": 2.9106060606060608e-05,
      "loss": 0.3981,
      "step": 4237
    },
    {
      "epoch": 1.0522,
      "grad_norm": 5.163064956665039,
      "learning_rate": 2.91010101010101e-05,
      "loss": 0.4336,
      "step": 4238
    },
    {
      "epoch": 1.0523,
      "grad_norm": 8.162704467773438,
      "learning_rate": 2.90959595959596e-05,
      "loss": 0.553,
      "step": 4239
    },
    {
      "epoch": 1.0524,
      "grad_norm": 7.495536804199219,
      "learning_rate": 2.909090909090909e-05,
      "loss": 2.0962,
      "step": 4240
    },
    {
      "epoch": 1.0525,
      "grad_norm": 5.927620887756348,
      "learning_rate": 2.908585858585859e-05,
      "loss": 0.6868,
      "step": 4241
    },
    {
      "epoch": 1.0526,
      "grad_norm": 11.82400894165039,
      "learning_rate": 2.908080808080808e-05,
      "loss": 1.2396,
      "step": 4242
    },
    {
      "epoch": 1.0527,
      "grad_norm": 9.139451026916504,
      "learning_rate": 2.9075757575757577e-05,
      "loss": 0.7666,
      "step": 4243
    },
    {
      "epoch": 1.0528,
      "grad_norm": 5.294138431549072,
      "learning_rate": 2.907070707070707e-05,
      "loss": 0.7117,
      "step": 4244
    },
    {
      "epoch": 1.0529,
      "grad_norm": 4.170468330383301,
      "learning_rate": 2.906565656565657e-05,
      "loss": 0.3672,
      "step": 4245
    },
    {
      "epoch": 1.053,
      "grad_norm": 8.490615844726562,
      "learning_rate": 2.9060606060606062e-05,
      "loss": 0.8652,
      "step": 4246
    },
    {
      "epoch": 1.0531,
      "grad_norm": 9.668669700622559,
      "learning_rate": 2.9055555555555558e-05,
      "loss": 0.7453,
      "step": 4247
    },
    {
      "epoch": 1.0532,
      "grad_norm": 8.229999542236328,
      "learning_rate": 2.905050505050505e-05,
      "loss": 1.011,
      "step": 4248
    },
    {
      "epoch": 1.0533,
      "grad_norm": 7.512782096862793,
      "learning_rate": 2.9045454545454546e-05,
      "loss": 0.1274,
      "step": 4249
    },
    {
      "epoch": 1.0534,
      "grad_norm": 10.428617477416992,
      "learning_rate": 2.904040404040404e-05,
      "loss": 0.664,
      "step": 4250
    },
    {
      "epoch": 1.0535,
      "grad_norm": 6.133363723754883,
      "learning_rate": 2.903535353535354e-05,
      "loss": 0.5192,
      "step": 4251
    },
    {
      "epoch": 1.0536,
      "grad_norm": 6.366739273071289,
      "learning_rate": 2.903030303030303e-05,
      "loss": 0.3356,
      "step": 4252
    },
    {
      "epoch": 1.0537,
      "grad_norm": 6.101811408996582,
      "learning_rate": 2.9025252525252527e-05,
      "loss": 1.3395,
      "step": 4253
    },
    {
      "epoch": 1.0538,
      "grad_norm": 15.56371021270752,
      "learning_rate": 2.902020202020202e-05,
      "loss": 0.6188,
      "step": 4254
    },
    {
      "epoch": 1.0539,
      "grad_norm": 8.078539848327637,
      "learning_rate": 2.901515151515152e-05,
      "loss": 0.5009,
      "step": 4255
    },
    {
      "epoch": 1.054,
      "grad_norm": 11.693065643310547,
      "learning_rate": 2.9010101010101008e-05,
      "loss": 0.2998,
      "step": 4256
    },
    {
      "epoch": 1.0541,
      "grad_norm": 11.92123031616211,
      "learning_rate": 2.9005050505050508e-05,
      "loss": 1.0337,
      "step": 4257
    },
    {
      "epoch": 1.0542,
      "grad_norm": 5.885799884796143,
      "learning_rate": 2.9e-05,
      "loss": 1.8413,
      "step": 4258
    },
    {
      "epoch": 1.0543,
      "grad_norm": 5.770583152770996,
      "learning_rate": 2.8994949494949496e-05,
      "loss": 0.5538,
      "step": 4259
    },
    {
      "epoch": 1.0544,
      "grad_norm": 5.419753551483154,
      "learning_rate": 2.898989898989899e-05,
      "loss": 0.7502,
      "step": 4260
    },
    {
      "epoch": 1.0545,
      "grad_norm": 11.180257797241211,
      "learning_rate": 2.8984848484848488e-05,
      "loss": 0.7484,
      "step": 4261
    },
    {
      "epoch": 1.0546,
      "grad_norm": 7.98895263671875,
      "learning_rate": 2.897979797979798e-05,
      "loss": 0.6353,
      "step": 4262
    },
    {
      "epoch": 1.0547,
      "grad_norm": 7.1525492668151855,
      "learning_rate": 2.8974747474747477e-05,
      "loss": 0.5372,
      "step": 4263
    },
    {
      "epoch": 1.0548,
      "grad_norm": 11.220785140991211,
      "learning_rate": 2.896969696969697e-05,
      "loss": 0.542,
      "step": 4264
    },
    {
      "epoch": 1.0549,
      "grad_norm": 6.63787317276001,
      "learning_rate": 2.8964646464646465e-05,
      "loss": 0.2914,
      "step": 4265
    },
    {
      "epoch": 1.055,
      "grad_norm": 5.306471347808838,
      "learning_rate": 2.8959595959595958e-05,
      "loss": 0.2749,
      "step": 4266
    },
    {
      "epoch": 1.0551,
      "grad_norm": 6.5405192375183105,
      "learning_rate": 2.8954545454545457e-05,
      "loss": 0.428,
      "step": 4267
    },
    {
      "epoch": 1.0552,
      "grad_norm": 4.928348064422607,
      "learning_rate": 2.894949494949495e-05,
      "loss": 0.3598,
      "step": 4268
    },
    {
      "epoch": 1.0553,
      "grad_norm": 5.990396976470947,
      "learning_rate": 2.8944444444444446e-05,
      "loss": 0.3647,
      "step": 4269
    },
    {
      "epoch": 1.0554,
      "grad_norm": 5.637720108032227,
      "learning_rate": 2.893939393939394e-05,
      "loss": 0.3931,
      "step": 4270
    },
    {
      "epoch": 1.0555,
      "grad_norm": 10.560224533081055,
      "learning_rate": 2.8934343434343434e-05,
      "loss": 0.4798,
      "step": 4271
    },
    {
      "epoch": 1.0556,
      "grad_norm": 5.267584800720215,
      "learning_rate": 2.8929292929292927e-05,
      "loss": 0.9089,
      "step": 4272
    },
    {
      "epoch": 1.0557,
      "grad_norm": 8.83423137664795,
      "learning_rate": 2.8924242424242426e-05,
      "loss": 0.435,
      "step": 4273
    },
    {
      "epoch": 1.0558,
      "grad_norm": 6.078190803527832,
      "learning_rate": 2.891919191919192e-05,
      "loss": 0.5383,
      "step": 4274
    },
    {
      "epoch": 1.0559,
      "grad_norm": 3.9054031372070312,
      "learning_rate": 2.8914141414141415e-05,
      "loss": 0.3022,
      "step": 4275
    },
    {
      "epoch": 1.056,
      "grad_norm": 8.6697359085083,
      "learning_rate": 2.8909090909090908e-05,
      "loss": 0.8995,
      "step": 4276
    },
    {
      "epoch": 1.0561,
      "grad_norm": 4.065650939941406,
      "learning_rate": 2.8904040404040407e-05,
      "loss": 0.6998,
      "step": 4277
    },
    {
      "epoch": 1.0562,
      "grad_norm": 4.343203544616699,
      "learning_rate": 2.8898989898989896e-05,
      "loss": 0.6771,
      "step": 4278
    },
    {
      "epoch": 1.0563,
      "grad_norm": 9.253653526306152,
      "learning_rate": 2.8893939393939396e-05,
      "loss": 0.6667,
      "step": 4279
    },
    {
      "epoch": 1.0564,
      "grad_norm": 12.466538429260254,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 0.3136,
      "step": 4280
    },
    {
      "epoch": 1.0565,
      "grad_norm": 3.075634002685547,
      "learning_rate": 2.8883838383838384e-05,
      "loss": 0.6649,
      "step": 4281
    },
    {
      "epoch": 1.0566,
      "grad_norm": 31.251684188842773,
      "learning_rate": 2.8878787878787884e-05,
      "loss": 0.643,
      "step": 4282
    },
    {
      "epoch": 1.0567,
      "grad_norm": 6.8457350730896,
      "learning_rate": 2.8873737373737376e-05,
      "loss": 0.7535,
      "step": 4283
    },
    {
      "epoch": 1.0568,
      "grad_norm": 7.192487716674805,
      "learning_rate": 2.8868686868686872e-05,
      "loss": 0.6565,
      "step": 4284
    },
    {
      "epoch": 1.0569,
      "grad_norm": 6.887092590332031,
      "learning_rate": 2.8863636363636365e-05,
      "loss": 0.7407,
      "step": 4285
    },
    {
      "epoch": 1.057,
      "grad_norm": 5.664918899536133,
      "learning_rate": 2.8858585858585864e-05,
      "loss": 0.3391,
      "step": 4286
    },
    {
      "epoch": 1.0571,
      "grad_norm": 7.074258327484131,
      "learning_rate": 2.8853535353535353e-05,
      "loss": 0.7322,
      "step": 4287
    },
    {
      "epoch": 1.0572,
      "grad_norm": 6.095677852630615,
      "learning_rate": 2.8848484848484853e-05,
      "loss": 0.3237,
      "step": 4288
    },
    {
      "epoch": 1.0573,
      "grad_norm": 7.928744792938232,
      "learning_rate": 2.8843434343434345e-05,
      "loss": 1.4838,
      "step": 4289
    },
    {
      "epoch": 1.0574,
      "grad_norm": 9.901263236999512,
      "learning_rate": 2.883838383838384e-05,
      "loss": 0.7729,
      "step": 4290
    },
    {
      "epoch": 1.0575,
      "grad_norm": 7.025125980377197,
      "learning_rate": 2.8833333333333334e-05,
      "loss": 0.384,
      "step": 4291
    },
    {
      "epoch": 1.0576,
      "grad_norm": 4.305107116699219,
      "learning_rate": 2.8828282828282833e-05,
      "loss": 0.2973,
      "step": 4292
    },
    {
      "epoch": 1.0577,
      "grad_norm": 8.400778770446777,
      "learning_rate": 2.8823232323232323e-05,
      "loss": 0.8455,
      "step": 4293
    },
    {
      "epoch": 1.0578,
      "grad_norm": 13.478079795837402,
      "learning_rate": 2.8818181818181822e-05,
      "loss": 0.2737,
      "step": 4294
    },
    {
      "epoch": 1.0579,
      "grad_norm": 8.409096717834473,
      "learning_rate": 2.8813131313131315e-05,
      "loss": 0.6562,
      "step": 4295
    },
    {
      "epoch": 1.058,
      "grad_norm": 10.966156959533691,
      "learning_rate": 2.880808080808081e-05,
      "loss": 0.4955,
      "step": 4296
    },
    {
      "epoch": 1.0581,
      "grad_norm": 8.274195671081543,
      "learning_rate": 2.8803030303030303e-05,
      "loss": 0.6349,
      "step": 4297
    },
    {
      "epoch": 1.0582,
      "grad_norm": 6.498606204986572,
      "learning_rate": 2.8797979797979803e-05,
      "loss": 0.2586,
      "step": 4298
    },
    {
      "epoch": 1.0583,
      "grad_norm": 4.2420172691345215,
      "learning_rate": 2.8792929292929295e-05,
      "loss": 1.1298,
      "step": 4299
    },
    {
      "epoch": 1.0584,
      "grad_norm": 6.3551483154296875,
      "learning_rate": 2.878787878787879e-05,
      "loss": 0.2687,
      "step": 4300
    },
    {
      "epoch": 1.0585,
      "grad_norm": 9.094085693359375,
      "learning_rate": 2.8782828282828284e-05,
      "loss": 1.405,
      "step": 4301
    },
    {
      "epoch": 1.0586,
      "grad_norm": 5.759845733642578,
      "learning_rate": 2.877777777777778e-05,
      "loss": 0.411,
      "step": 4302
    },
    {
      "epoch": 1.0587,
      "grad_norm": 5.886022567749023,
      "learning_rate": 2.8772727272727272e-05,
      "loss": 0.3601,
      "step": 4303
    },
    {
      "epoch": 1.0588,
      "grad_norm": 13.491401672363281,
      "learning_rate": 2.876767676767677e-05,
      "loss": 0.3386,
      "step": 4304
    },
    {
      "epoch": 1.0589,
      "grad_norm": 7.133113384246826,
      "learning_rate": 2.8762626262626264e-05,
      "loss": 0.5201,
      "step": 4305
    },
    {
      "epoch": 1.059,
      "grad_norm": 8.822763442993164,
      "learning_rate": 2.875757575757576e-05,
      "loss": 0.4706,
      "step": 4306
    },
    {
      "epoch": 1.0591,
      "grad_norm": 5.207473278045654,
      "learning_rate": 2.8752525252525253e-05,
      "loss": 0.7347,
      "step": 4307
    },
    {
      "epoch": 1.0592,
      "grad_norm": 3.7004871368408203,
      "learning_rate": 2.8747474747474752e-05,
      "loss": 0.2189,
      "step": 4308
    },
    {
      "epoch": 1.0593,
      "grad_norm": 6.454064846038818,
      "learning_rate": 2.874242424242424e-05,
      "loss": 0.7462,
      "step": 4309
    },
    {
      "epoch": 1.0594,
      "grad_norm": 8.722546577453613,
      "learning_rate": 2.873737373737374e-05,
      "loss": 1.3695,
      "step": 4310
    },
    {
      "epoch": 1.0594999999999999,
      "grad_norm": 3.7144618034362793,
      "learning_rate": 2.8732323232323233e-05,
      "loss": 0.4055,
      "step": 4311
    },
    {
      "epoch": 1.0596,
      "grad_norm": 14.457049369812012,
      "learning_rate": 2.872727272727273e-05,
      "loss": 0.4049,
      "step": 4312
    },
    {
      "epoch": 1.0597,
      "grad_norm": 8.021553039550781,
      "learning_rate": 2.8722222222222222e-05,
      "loss": 0.8134,
      "step": 4313
    },
    {
      "epoch": 1.0598,
      "grad_norm": 5.5703558921813965,
      "learning_rate": 2.871717171717172e-05,
      "loss": 0.708,
      "step": 4314
    },
    {
      "epoch": 1.0599,
      "grad_norm": 8.111011505126953,
      "learning_rate": 2.8712121212121214e-05,
      "loss": 0.3492,
      "step": 4315
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.828624725341797,
      "learning_rate": 2.870707070707071e-05,
      "loss": 0.2801,
      "step": 4316
    },
    {
      "epoch": 1.0601,
      "grad_norm": 5.243246555328369,
      "learning_rate": 2.8702020202020203e-05,
      "loss": 0.5526,
      "step": 4317
    },
    {
      "epoch": 1.0602,
      "grad_norm": 5.223872661590576,
      "learning_rate": 2.86969696969697e-05,
      "loss": 0.4387,
      "step": 4318
    },
    {
      "epoch": 1.0603,
      "grad_norm": 5.3030619621276855,
      "learning_rate": 2.869191919191919e-05,
      "loss": 0.4882,
      "step": 4319
    },
    {
      "epoch": 1.0604,
      "grad_norm": 4.418000221252441,
      "learning_rate": 2.868686868686869e-05,
      "loss": 0.2704,
      "step": 4320
    },
    {
      "epoch": 1.0605,
      "grad_norm": 7.9851460456848145,
      "learning_rate": 2.8681818181818183e-05,
      "loss": 0.9473,
      "step": 4321
    },
    {
      "epoch": 1.0606,
      "grad_norm": 8.718622207641602,
      "learning_rate": 2.867676767676768e-05,
      "loss": 0.2673,
      "step": 4322
    },
    {
      "epoch": 1.0607,
      "grad_norm": 18.140188217163086,
      "learning_rate": 2.8671717171717172e-05,
      "loss": 0.9872,
      "step": 4323
    },
    {
      "epoch": 1.0608,
      "grad_norm": 5.69589376449585,
      "learning_rate": 2.8666666666666668e-05,
      "loss": 0.3257,
      "step": 4324
    },
    {
      "epoch": 1.0609,
      "grad_norm": 6.627404689788818,
      "learning_rate": 2.866161616161616e-05,
      "loss": 0.8757,
      "step": 4325
    },
    {
      "epoch": 1.061,
      "grad_norm": 6.050112247467041,
      "learning_rate": 2.865656565656566e-05,
      "loss": 0.4508,
      "step": 4326
    },
    {
      "epoch": 1.0611,
      "grad_norm": 7.163559913635254,
      "learning_rate": 2.8651515151515152e-05,
      "loss": 0.5988,
      "step": 4327
    },
    {
      "epoch": 1.0612,
      "grad_norm": 4.36680793762207,
      "learning_rate": 2.864646464646465e-05,
      "loss": 0.2562,
      "step": 4328
    },
    {
      "epoch": 1.0613,
      "grad_norm": 9.11523151397705,
      "learning_rate": 2.864141414141414e-05,
      "loss": 0.2171,
      "step": 4329
    },
    {
      "epoch": 1.0614,
      "grad_norm": 4.884328842163086,
      "learning_rate": 2.863636363636364e-05,
      "loss": 0.3222,
      "step": 4330
    },
    {
      "epoch": 1.0615,
      "grad_norm": 4.031778335571289,
      "learning_rate": 2.863131313131313e-05,
      "loss": 0.2852,
      "step": 4331
    },
    {
      "epoch": 1.0616,
      "grad_norm": 5.868741035461426,
      "learning_rate": 2.862626262626263e-05,
      "loss": 0.2687,
      "step": 4332
    },
    {
      "epoch": 1.0617,
      "grad_norm": 11.552743911743164,
      "learning_rate": 2.862121212121212e-05,
      "loss": 0.2817,
      "step": 4333
    },
    {
      "epoch": 1.0618,
      "grad_norm": 9.348404884338379,
      "learning_rate": 2.8616161616161618e-05,
      "loss": 0.362,
      "step": 4334
    },
    {
      "epoch": 1.0619,
      "grad_norm": 4.181041240692139,
      "learning_rate": 2.861111111111111e-05,
      "loss": 0.2627,
      "step": 4335
    },
    {
      "epoch": 1.062,
      "grad_norm": 5.826622486114502,
      "learning_rate": 2.860606060606061e-05,
      "loss": 0.7709,
      "step": 4336
    },
    {
      "epoch": 1.0621,
      "grad_norm": 7.216421127319336,
      "learning_rate": 2.8601010101010102e-05,
      "loss": 0.4398,
      "step": 4337
    },
    {
      "epoch": 1.0622,
      "grad_norm": 4.4797892570495605,
      "learning_rate": 2.8595959595959598e-05,
      "loss": 0.6364,
      "step": 4338
    },
    {
      "epoch": 1.0623,
      "grad_norm": 4.548102855682373,
      "learning_rate": 2.859090909090909e-05,
      "loss": 0.2812,
      "step": 4339
    },
    {
      "epoch": 1.0624,
      "grad_norm": 6.442489147186279,
      "learning_rate": 2.8585858585858587e-05,
      "loss": 0.7178,
      "step": 4340
    },
    {
      "epoch": 1.0625,
      "grad_norm": 4.9091620445251465,
      "learning_rate": 2.858080808080808e-05,
      "loss": 0.2631,
      "step": 4341
    },
    {
      "epoch": 1.0626,
      "grad_norm": 5.721745014190674,
      "learning_rate": 2.857575757575758e-05,
      "loss": 0.4199,
      "step": 4342
    },
    {
      "epoch": 1.0627,
      "grad_norm": 8.338112831115723,
      "learning_rate": 2.857070707070707e-05,
      "loss": 0.2531,
      "step": 4343
    },
    {
      "epoch": 1.0628,
      "grad_norm": 3.337961435317993,
      "learning_rate": 2.8565656565656567e-05,
      "loss": 0.0434,
      "step": 4344
    },
    {
      "epoch": 1.0629,
      "grad_norm": 5.480473518371582,
      "learning_rate": 2.856060606060606e-05,
      "loss": 0.144,
      "step": 4345
    },
    {
      "epoch": 1.063,
      "grad_norm": 13.426666259765625,
      "learning_rate": 2.855555555555556e-05,
      "loss": 0.4491,
      "step": 4346
    },
    {
      "epoch": 1.0631,
      "grad_norm": 6.921092987060547,
      "learning_rate": 2.855050505050505e-05,
      "loss": 0.4994,
      "step": 4347
    },
    {
      "epoch": 1.0632,
      "grad_norm": 2.278573751449585,
      "learning_rate": 2.8545454545454548e-05,
      "loss": 0.1479,
      "step": 4348
    },
    {
      "epoch": 1.0633,
      "grad_norm": 8.850642204284668,
      "learning_rate": 2.854040404040404e-05,
      "loss": 0.3691,
      "step": 4349
    },
    {
      "epoch": 1.0634,
      "grad_norm": 5.347734451293945,
      "learning_rate": 2.8535353535353536e-05,
      "loss": 0.1395,
      "step": 4350
    },
    {
      "epoch": 1.0635,
      "grad_norm": 9.118569374084473,
      "learning_rate": 2.853030303030303e-05,
      "loss": 0.2639,
      "step": 4351
    },
    {
      "epoch": 1.0636,
      "grad_norm": 9.424173355102539,
      "learning_rate": 2.852525252525253e-05,
      "loss": 0.3382,
      "step": 4352
    },
    {
      "epoch": 1.0637,
      "grad_norm": 2.8315727710723877,
      "learning_rate": 2.8520202020202018e-05,
      "loss": 0.0967,
      "step": 4353
    },
    {
      "epoch": 1.0638,
      "grad_norm": 1.865235447883606,
      "learning_rate": 2.8515151515151517e-05,
      "loss": 0.0286,
      "step": 4354
    },
    {
      "epoch": 1.0639,
      "grad_norm": 1.3404150009155273,
      "learning_rate": 2.851010101010101e-05,
      "loss": 0.0332,
      "step": 4355
    },
    {
      "epoch": 1.064,
      "grad_norm": 2.4587242603302,
      "learning_rate": 2.8505050505050506e-05,
      "loss": 0.2409,
      "step": 4356
    },
    {
      "epoch": 1.0641,
      "grad_norm": 10.97844409942627,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.387,
      "step": 4357
    },
    {
      "epoch": 1.0642,
      "grad_norm": 3.774998664855957,
      "learning_rate": 2.8494949494949498e-05,
      "loss": 0.1693,
      "step": 4358
    },
    {
      "epoch": 1.0643,
      "grad_norm": 5.6600847244262695,
      "learning_rate": 2.848989898989899e-05,
      "loss": 0.8209,
      "step": 4359
    },
    {
      "epoch": 1.0644,
      "grad_norm": 10.72606372833252,
      "learning_rate": 2.8484848484848486e-05,
      "loss": 0.3698,
      "step": 4360
    },
    {
      "epoch": 1.0645,
      "grad_norm": 5.405937671661377,
      "learning_rate": 2.847979797979798e-05,
      "loss": 0.4672,
      "step": 4361
    },
    {
      "epoch": 1.0646,
      "grad_norm": 2.0433342456817627,
      "learning_rate": 2.8474747474747475e-05,
      "loss": 0.0741,
      "step": 4362
    },
    {
      "epoch": 1.0647,
      "grad_norm": 3.318584442138672,
      "learning_rate": 2.8469696969696967e-05,
      "loss": 0.5844,
      "step": 4363
    },
    {
      "epoch": 1.0648,
      "grad_norm": 3.773360252380371,
      "learning_rate": 2.8464646464646467e-05,
      "loss": 0.0926,
      "step": 4364
    },
    {
      "epoch": 1.0649,
      "grad_norm": 3.035356044769287,
      "learning_rate": 2.845959595959596e-05,
      "loss": 0.2571,
      "step": 4365
    },
    {
      "epoch": 1.065,
      "grad_norm": 2.3809945583343506,
      "learning_rate": 2.8454545454545455e-05,
      "loss": 0.0676,
      "step": 4366
    },
    {
      "epoch": 1.0651,
      "grad_norm": 9.354891777038574,
      "learning_rate": 2.8449494949494948e-05,
      "loss": 0.3057,
      "step": 4367
    },
    {
      "epoch": 1.0652,
      "grad_norm": 11.223954200744629,
      "learning_rate": 2.8444444444444447e-05,
      "loss": 1.3553,
      "step": 4368
    },
    {
      "epoch": 1.0653,
      "grad_norm": 5.111243724822998,
      "learning_rate": 2.8439393939393943e-05,
      "loss": 1.6014,
      "step": 4369
    },
    {
      "epoch": 1.0654,
      "grad_norm": 10.86934757232666,
      "learning_rate": 2.8434343434343436e-05,
      "loss": 0.535,
      "step": 4370
    },
    {
      "epoch": 1.0655000000000001,
      "grad_norm": 7.127585411071777,
      "learning_rate": 2.8429292929292932e-05,
      "loss": 0.6159,
      "step": 4371
    },
    {
      "epoch": 1.0656,
      "grad_norm": 4.838507652282715,
      "learning_rate": 2.8424242424242424e-05,
      "loss": 0.2876,
      "step": 4372
    },
    {
      "epoch": 1.0657,
      "grad_norm": 5.125414848327637,
      "learning_rate": 2.8419191919191924e-05,
      "loss": 0.8315,
      "step": 4373
    },
    {
      "epoch": 1.0658,
      "grad_norm": 11.257477760314941,
      "learning_rate": 2.8414141414141416e-05,
      "loss": 0.1356,
      "step": 4374
    },
    {
      "epoch": 1.0659,
      "grad_norm": 4.515442371368408,
      "learning_rate": 2.8409090909090912e-05,
      "loss": 0.3486,
      "step": 4375
    },
    {
      "epoch": 1.066,
      "grad_norm": 1.041548252105713,
      "learning_rate": 2.8404040404040405e-05,
      "loss": 0.0188,
      "step": 4376
    },
    {
      "epoch": 1.0661,
      "grad_norm": 7.4631547927856445,
      "learning_rate": 2.8398989898989904e-05,
      "loss": 0.5184,
      "step": 4377
    },
    {
      "epoch": 1.0662,
      "grad_norm": 2.4721920490264893,
      "learning_rate": 2.8393939393939394e-05,
      "loss": 0.0873,
      "step": 4378
    },
    {
      "epoch": 1.0663,
      "grad_norm": 4.9490861892700195,
      "learning_rate": 2.8388888888888893e-05,
      "loss": 0.4824,
      "step": 4379
    },
    {
      "epoch": 1.0664,
      "grad_norm": 32.52640151977539,
      "learning_rate": 2.8383838383838386e-05,
      "loss": 0.1448,
      "step": 4380
    },
    {
      "epoch": 1.0665,
      "grad_norm": 4.994377613067627,
      "learning_rate": 2.837878787878788e-05,
      "loss": 0.4048,
      "step": 4381
    },
    {
      "epoch": 1.0666,
      "grad_norm": 7.377306938171387,
      "learning_rate": 2.8373737373737374e-05,
      "loss": 0.8947,
      "step": 4382
    },
    {
      "epoch": 1.0667,
      "grad_norm": 5.901076316833496,
      "learning_rate": 2.8368686868686874e-05,
      "loss": 0.1618,
      "step": 4383
    },
    {
      "epoch": 1.0668,
      "grad_norm": 2.829155206680298,
      "learning_rate": 2.8363636363636363e-05,
      "loss": 0.233,
      "step": 4384
    },
    {
      "epoch": 1.0669,
      "grad_norm": 6.718702793121338,
      "learning_rate": 2.8358585858585862e-05,
      "loss": 0.6652,
      "step": 4385
    },
    {
      "epoch": 1.067,
      "grad_norm": 2.927546739578247,
      "learning_rate": 2.8353535353535355e-05,
      "loss": 0.0602,
      "step": 4386
    },
    {
      "epoch": 1.0671,
      "grad_norm": 4.367210865020752,
      "learning_rate": 2.834848484848485e-05,
      "loss": 0.5696,
      "step": 4387
    },
    {
      "epoch": 1.0672,
      "grad_norm": 4.0259480476379395,
      "learning_rate": 2.8343434343434343e-05,
      "loss": 0.1805,
      "step": 4388
    },
    {
      "epoch": 1.0673,
      "grad_norm": 10.392666816711426,
      "learning_rate": 2.8338383838383843e-05,
      "loss": 0.463,
      "step": 4389
    },
    {
      "epoch": 1.0674,
      "grad_norm": 35.31693649291992,
      "learning_rate": 2.8333333333333335e-05,
      "loss": 0.1527,
      "step": 4390
    },
    {
      "epoch": 1.0675,
      "grad_norm": 4.0556464195251465,
      "learning_rate": 2.832828282828283e-05,
      "loss": 0.3366,
      "step": 4391
    },
    {
      "epoch": 1.0676,
      "grad_norm": 10.443205833435059,
      "learning_rate": 2.8323232323232324e-05,
      "loss": 0.4568,
      "step": 4392
    },
    {
      "epoch": 1.0677,
      "grad_norm": 10.120539665222168,
      "learning_rate": 2.831818181818182e-05,
      "loss": 0.2475,
      "step": 4393
    },
    {
      "epoch": 1.0678,
      "grad_norm": 13.846693992614746,
      "learning_rate": 2.8313131313131313e-05,
      "loss": 0.2374,
      "step": 4394
    },
    {
      "epoch": 1.0679,
      "grad_norm": 1.9835753440856934,
      "learning_rate": 2.8308080808080812e-05,
      "loss": 0.0437,
      "step": 4395
    },
    {
      "epoch": 1.068,
      "grad_norm": 5.784053802490234,
      "learning_rate": 2.8303030303030305e-05,
      "loss": 0.203,
      "step": 4396
    },
    {
      "epoch": 1.0681,
      "grad_norm": 2.819420337677002,
      "learning_rate": 2.82979797979798e-05,
      "loss": 0.0424,
      "step": 4397
    },
    {
      "epoch": 1.0682,
      "grad_norm": 3.9881017208099365,
      "learning_rate": 2.8292929292929293e-05,
      "loss": 0.0668,
      "step": 4398
    },
    {
      "epoch": 1.0683,
      "grad_norm": 5.9591827392578125,
      "learning_rate": 2.8287878787878793e-05,
      "loss": 1.6598,
      "step": 4399
    },
    {
      "epoch": 1.0684,
      "grad_norm": 6.74967098236084,
      "learning_rate": 2.8282828282828282e-05,
      "loss": 0.4266,
      "step": 4400
    },
    {
      "epoch": 1.0685,
      "grad_norm": 8.821608543395996,
      "learning_rate": 2.827777777777778e-05,
      "loss": 0.296,
      "step": 4401
    },
    {
      "epoch": 1.0686,
      "grad_norm": 4.5018310546875,
      "learning_rate": 2.8272727272727274e-05,
      "loss": 0.1838,
      "step": 4402
    },
    {
      "epoch": 1.0687,
      "grad_norm": 4.537114143371582,
      "learning_rate": 2.826767676767677e-05,
      "loss": 0.2105,
      "step": 4403
    },
    {
      "epoch": 1.0688,
      "grad_norm": 1.7334765195846558,
      "learning_rate": 2.8262626262626262e-05,
      "loss": 0.4877,
      "step": 4404
    },
    {
      "epoch": 1.0689,
      "grad_norm": 3.8243155479431152,
      "learning_rate": 2.825757575757576e-05,
      "loss": 1.0884,
      "step": 4405
    },
    {
      "epoch": 1.069,
      "grad_norm": 5.708029747009277,
      "learning_rate": 2.8252525252525254e-05,
      "loss": 0.2939,
      "step": 4406
    },
    {
      "epoch": 1.0691,
      "grad_norm": 13.436494827270508,
      "learning_rate": 2.824747474747475e-05,
      "loss": 0.1405,
      "step": 4407
    },
    {
      "epoch": 1.0692,
      "grad_norm": 2.7021048069000244,
      "learning_rate": 2.8242424242424243e-05,
      "loss": 0.0975,
      "step": 4408
    },
    {
      "epoch": 1.0693,
      "grad_norm": 2.386166572570801,
      "learning_rate": 2.823737373737374e-05,
      "loss": 0.0369,
      "step": 4409
    },
    {
      "epoch": 1.0694,
      "grad_norm": 2.982038974761963,
      "learning_rate": 2.823232323232323e-05,
      "loss": 0.2971,
      "step": 4410
    },
    {
      "epoch": 1.0695000000000001,
      "grad_norm": 3.745443105697632,
      "learning_rate": 2.822727272727273e-05,
      "loss": 0.5819,
      "step": 4411
    },
    {
      "epoch": 1.0695999999999999,
      "grad_norm": 6.997735023498535,
      "learning_rate": 2.8222222222222223e-05,
      "loss": 0.3086,
      "step": 4412
    },
    {
      "epoch": 1.0697,
      "grad_norm": 2.2323529720306396,
      "learning_rate": 2.821717171717172e-05,
      "loss": 0.0765,
      "step": 4413
    },
    {
      "epoch": 1.0698,
      "grad_norm": 2.1999106407165527,
      "learning_rate": 2.8212121212121212e-05,
      "loss": 0.0461,
      "step": 4414
    },
    {
      "epoch": 1.0699,
      "grad_norm": 3.8359224796295166,
      "learning_rate": 2.8207070707070708e-05,
      "loss": 0.1081,
      "step": 4415
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.387284278869629,
      "learning_rate": 2.82020202020202e-05,
      "loss": 0.549,
      "step": 4416
    },
    {
      "epoch": 1.0701,
      "grad_norm": 4.948184013366699,
      "learning_rate": 2.81969696969697e-05,
      "loss": 0.6208,
      "step": 4417
    },
    {
      "epoch": 1.0702,
      "grad_norm": 3.0997109413146973,
      "learning_rate": 2.8191919191919193e-05,
      "loss": 0.145,
      "step": 4418
    },
    {
      "epoch": 1.0703,
      "grad_norm": 3.7700700759887695,
      "learning_rate": 2.818686868686869e-05,
      "loss": 0.5688,
      "step": 4419
    },
    {
      "epoch": 1.0704,
      "grad_norm": 6.206628799438477,
      "learning_rate": 2.818181818181818e-05,
      "loss": 1.2963,
      "step": 4420
    },
    {
      "epoch": 1.0705,
      "grad_norm": 2.6677439212799072,
      "learning_rate": 2.817676767676768e-05,
      "loss": 0.2457,
      "step": 4421
    },
    {
      "epoch": 1.0706,
      "grad_norm": 7.225451946258545,
      "learning_rate": 2.817171717171717e-05,
      "loss": 0.7346,
      "step": 4422
    },
    {
      "epoch": 1.0707,
      "grad_norm": 5.4837164878845215,
      "learning_rate": 2.816666666666667e-05,
      "loss": 0.1579,
      "step": 4423
    },
    {
      "epoch": 1.0708,
      "grad_norm": 4.948251724243164,
      "learning_rate": 2.8161616161616162e-05,
      "loss": 0.2177,
      "step": 4424
    },
    {
      "epoch": 1.0709,
      "grad_norm": 3.7058727741241455,
      "learning_rate": 2.8156565656565658e-05,
      "loss": 0.0743,
      "step": 4425
    },
    {
      "epoch": 1.071,
      "grad_norm": 2.8269124031066895,
      "learning_rate": 2.815151515151515e-05,
      "loss": 0.1008,
      "step": 4426
    },
    {
      "epoch": 1.0711,
      "grad_norm": 11.019569396972656,
      "learning_rate": 2.814646464646465e-05,
      "loss": 0.5999,
      "step": 4427
    },
    {
      "epoch": 1.0712,
      "grad_norm": 4.788045406341553,
      "learning_rate": 2.8141414141414142e-05,
      "loss": 0.5845,
      "step": 4428
    },
    {
      "epoch": 1.0713,
      "grad_norm": 7.889639854431152,
      "learning_rate": 2.813636363636364e-05,
      "loss": 0.663,
      "step": 4429
    },
    {
      "epoch": 1.0714,
      "grad_norm": 10.433123588562012,
      "learning_rate": 2.813131313131313e-05,
      "loss": 0.7509,
      "step": 4430
    },
    {
      "epoch": 1.0715,
      "grad_norm": 4.298377513885498,
      "learning_rate": 2.8126262626262627e-05,
      "loss": 1.2686,
      "step": 4431
    },
    {
      "epoch": 1.0716,
      "grad_norm": 3.4478678703308105,
      "learning_rate": 2.812121212121212e-05,
      "loss": 0.0824,
      "step": 4432
    },
    {
      "epoch": 1.0717,
      "grad_norm": 1.2576937675476074,
      "learning_rate": 2.811616161616162e-05,
      "loss": 0.0195,
      "step": 4433
    },
    {
      "epoch": 1.0718,
      "grad_norm": 16.025588989257812,
      "learning_rate": 2.811111111111111e-05,
      "loss": 0.5142,
      "step": 4434
    },
    {
      "epoch": 1.0719,
      "grad_norm": 6.584979057312012,
      "learning_rate": 2.8106060606060607e-05,
      "loss": 0.3712,
      "step": 4435
    },
    {
      "epoch": 1.072,
      "grad_norm": 6.07836389541626,
      "learning_rate": 2.81010101010101e-05,
      "loss": 0.908,
      "step": 4436
    },
    {
      "epoch": 1.0721,
      "grad_norm": 6.564093589782715,
      "learning_rate": 2.80959595959596e-05,
      "loss": 0.206,
      "step": 4437
    },
    {
      "epoch": 1.0722,
      "grad_norm": 5.219066619873047,
      "learning_rate": 2.809090909090909e-05,
      "loss": 0.3207,
      "step": 4438
    },
    {
      "epoch": 1.0723,
      "grad_norm": 2.1587045192718506,
      "learning_rate": 2.8085858585858588e-05,
      "loss": 0.0865,
      "step": 4439
    },
    {
      "epoch": 1.0724,
      "grad_norm": 1.7250285148620605,
      "learning_rate": 2.808080808080808e-05,
      "loss": 0.4637,
      "step": 4440
    },
    {
      "epoch": 1.0725,
      "grad_norm": 3.2207839488983154,
      "learning_rate": 2.8075757575757577e-05,
      "loss": 0.0277,
      "step": 4441
    },
    {
      "epoch": 1.0726,
      "grad_norm": 2.9696576595306396,
      "learning_rate": 2.807070707070707e-05,
      "loss": 0.738,
      "step": 4442
    },
    {
      "epoch": 1.0727,
      "grad_norm": 6.014795303344727,
      "learning_rate": 2.806565656565657e-05,
      "loss": 0.211,
      "step": 4443
    },
    {
      "epoch": 1.0728,
      "grad_norm": 6.061066150665283,
      "learning_rate": 2.8060606060606058e-05,
      "loss": 0.2322,
      "step": 4444
    },
    {
      "epoch": 1.0729,
      "grad_norm": 18.290321350097656,
      "learning_rate": 2.8055555555555557e-05,
      "loss": 0.2278,
      "step": 4445
    },
    {
      "epoch": 1.073,
      "grad_norm": 4.303313732147217,
      "learning_rate": 2.805050505050505e-05,
      "loss": 0.2036,
      "step": 4446
    },
    {
      "epoch": 1.0731,
      "grad_norm": 2.9161174297332764,
      "learning_rate": 2.8045454545454546e-05,
      "loss": 0.6609,
      "step": 4447
    },
    {
      "epoch": 1.0732,
      "grad_norm": 2.4000985622406006,
      "learning_rate": 2.804040404040404e-05,
      "loss": 0.0375,
      "step": 4448
    },
    {
      "epoch": 1.0733,
      "grad_norm": 13.13653564453125,
      "learning_rate": 2.8035353535353538e-05,
      "loss": 0.2235,
      "step": 4449
    },
    {
      "epoch": 1.0734,
      "grad_norm": 7.471441268920898,
      "learning_rate": 2.803030303030303e-05,
      "loss": 0.173,
      "step": 4450
    },
    {
      "epoch": 1.0735,
      "grad_norm": 4.978579998016357,
      "learning_rate": 2.8025252525252526e-05,
      "loss": 0.4428,
      "step": 4451
    },
    {
      "epoch": 1.0735999999999999,
      "grad_norm": 1.3434606790542603,
      "learning_rate": 2.802020202020202e-05,
      "loss": 0.4611,
      "step": 4452
    },
    {
      "epoch": 1.0737,
      "grad_norm": 5.251018047332764,
      "learning_rate": 2.8015151515151515e-05,
      "loss": 0.345,
      "step": 4453
    },
    {
      "epoch": 1.0738,
      "grad_norm": 2.6774890422821045,
      "learning_rate": 2.8010101010101014e-05,
      "loss": 0.5195,
      "step": 4454
    },
    {
      "epoch": 1.0739,
      "grad_norm": 1.8283231258392334,
      "learning_rate": 2.8005050505050507e-05,
      "loss": 0.0372,
      "step": 4455
    },
    {
      "epoch": 1.074,
      "grad_norm": 2.0217654705047607,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.0474,
      "step": 4456
    },
    {
      "epoch": 1.0741,
      "grad_norm": 21.47507095336914,
      "learning_rate": 2.7994949494949496e-05,
      "loss": 0.7539,
      "step": 4457
    },
    {
      "epoch": 1.0742,
      "grad_norm": 4.061133861541748,
      "learning_rate": 2.7989898989898995e-05,
      "loss": 0.2736,
      "step": 4458
    },
    {
      "epoch": 1.0743,
      "grad_norm": 3.2510056495666504,
      "learning_rate": 2.7984848484848488e-05,
      "loss": 0.3122,
      "step": 4459
    },
    {
      "epoch": 1.0744,
      "grad_norm": 5.277367115020752,
      "learning_rate": 2.7979797979797984e-05,
      "loss": 0.8476,
      "step": 4460
    },
    {
      "epoch": 1.0745,
      "grad_norm": 4.873936653137207,
      "learning_rate": 2.7974747474747476e-05,
      "loss": 0.2355,
      "step": 4461
    },
    {
      "epoch": 1.0746,
      "grad_norm": 4.040995121002197,
      "learning_rate": 2.7969696969696972e-05,
      "loss": 0.3563,
      "step": 4462
    },
    {
      "epoch": 1.0747,
      "grad_norm": 4.282700538635254,
      "learning_rate": 2.7964646464646465e-05,
      "loss": 0.1013,
      "step": 4463
    },
    {
      "epoch": 1.0748,
      "grad_norm": 3.519893169403076,
      "learning_rate": 2.7959595959595964e-05,
      "loss": 0.7203,
      "step": 4464
    },
    {
      "epoch": 1.0749,
      "grad_norm": 1.1456388235092163,
      "learning_rate": 2.7954545454545457e-05,
      "loss": 0.0304,
      "step": 4465
    },
    {
      "epoch": 1.075,
      "grad_norm": 22.547521591186523,
      "learning_rate": 2.7949494949494953e-05,
      "loss": 0.3097,
      "step": 4466
    },
    {
      "epoch": 1.0751,
      "grad_norm": 3.187039613723755,
      "learning_rate": 2.7944444444444445e-05,
      "loss": 0.2969,
      "step": 4467
    },
    {
      "epoch": 1.0752,
      "grad_norm": 1.904062271118164,
      "learning_rate": 2.7939393939393945e-05,
      "loss": 0.0677,
      "step": 4468
    },
    {
      "epoch": 1.0753,
      "grad_norm": 4.271792888641357,
      "learning_rate": 2.7934343434343434e-05,
      "loss": 0.1735,
      "step": 4469
    },
    {
      "epoch": 1.0754,
      "grad_norm": 3.2035248279571533,
      "learning_rate": 2.7929292929292933e-05,
      "loss": 0.1798,
      "step": 4470
    },
    {
      "epoch": 1.0755,
      "grad_norm": 2.3194808959960938,
      "learning_rate": 2.7924242424242426e-05,
      "loss": 0.0879,
      "step": 4471
    },
    {
      "epoch": 1.0756000000000001,
      "grad_norm": 4.155483722686768,
      "learning_rate": 2.7919191919191922e-05,
      "loss": 0.4504,
      "step": 4472
    },
    {
      "epoch": 1.0757,
      "grad_norm": 6.338697910308838,
      "learning_rate": 2.7914141414141414e-05,
      "loss": 0.1374,
      "step": 4473
    },
    {
      "epoch": 1.0758,
      "grad_norm": 4.840360164642334,
      "learning_rate": 2.7909090909090914e-05,
      "loss": 0.5603,
      "step": 4474
    },
    {
      "epoch": 1.0759,
      "grad_norm": 11.224135398864746,
      "learning_rate": 2.7904040404040403e-05,
      "loss": 0.6157,
      "step": 4475
    },
    {
      "epoch": 1.076,
      "grad_norm": 14.17590618133545,
      "learning_rate": 2.7898989898989902e-05,
      "loss": 0.8824,
      "step": 4476
    },
    {
      "epoch": 1.0761,
      "grad_norm": 4.43388557434082,
      "learning_rate": 2.7893939393939395e-05,
      "loss": 0.7116,
      "step": 4477
    },
    {
      "epoch": 1.0762,
      "grad_norm": 7.751522541046143,
      "learning_rate": 2.788888888888889e-05,
      "loss": 0.1601,
      "step": 4478
    },
    {
      "epoch": 1.0763,
      "grad_norm": 3.6110994815826416,
      "learning_rate": 2.7883838383838384e-05,
      "loss": 0.0968,
      "step": 4479
    },
    {
      "epoch": 1.0764,
      "grad_norm": 5.974032402038574,
      "learning_rate": 2.7878787878787883e-05,
      "loss": 0.4391,
      "step": 4480
    },
    {
      "epoch": 1.0765,
      "grad_norm": 6.334407329559326,
      "learning_rate": 2.7873737373737376e-05,
      "loss": 0.8604,
      "step": 4481
    },
    {
      "epoch": 1.0766,
      "grad_norm": 5.407216548919678,
      "learning_rate": 2.786868686868687e-05,
      "loss": 0.2358,
      "step": 4482
    },
    {
      "epoch": 1.0767,
      "grad_norm": 4.285943984985352,
      "learning_rate": 2.7863636363636364e-05,
      "loss": 0.0979,
      "step": 4483
    },
    {
      "epoch": 1.0768,
      "grad_norm": 4.491153240203857,
      "learning_rate": 2.785858585858586e-05,
      "loss": 0.2563,
      "step": 4484
    },
    {
      "epoch": 1.0769,
      "grad_norm": 2.446692943572998,
      "learning_rate": 2.7853535353535353e-05,
      "loss": 0.0982,
      "step": 4485
    },
    {
      "epoch": 1.077,
      "grad_norm": 3.0887210369110107,
      "learning_rate": 2.7848484848484852e-05,
      "loss": 0.1092,
      "step": 4486
    },
    {
      "epoch": 1.0771,
      "grad_norm": 4.3028130531311035,
      "learning_rate": 2.7843434343434345e-05,
      "loss": 0.4097,
      "step": 4487
    },
    {
      "epoch": 1.0772,
      "grad_norm": 4.745615005493164,
      "learning_rate": 2.783838383838384e-05,
      "loss": 0.0564,
      "step": 4488
    },
    {
      "epoch": 1.0773,
      "grad_norm": 7.891913890838623,
      "learning_rate": 2.7833333333333333e-05,
      "loss": 0.5533,
      "step": 4489
    },
    {
      "epoch": 1.0774,
      "grad_norm": 9.438396453857422,
      "learning_rate": 2.7828282828282833e-05,
      "loss": 0.1718,
      "step": 4490
    },
    {
      "epoch": 1.0775,
      "grad_norm": 4.0838189125061035,
      "learning_rate": 2.7823232323232322e-05,
      "loss": 0.1719,
      "step": 4491
    },
    {
      "epoch": 1.0776,
      "grad_norm": 12.88651180267334,
      "learning_rate": 2.781818181818182e-05,
      "loss": 0.43,
      "step": 4492
    },
    {
      "epoch": 1.0777,
      "grad_norm": 5.266528606414795,
      "learning_rate": 2.7813131313131314e-05,
      "loss": 0.5928,
      "step": 4493
    },
    {
      "epoch": 1.0778,
      "grad_norm": 7.787312984466553,
      "learning_rate": 2.780808080808081e-05,
      "loss": 0.7031,
      "step": 4494
    },
    {
      "epoch": 1.0779,
      "grad_norm": 8.67168140411377,
      "learning_rate": 2.7803030303030303e-05,
      "loss": 0.282,
      "step": 4495
    },
    {
      "epoch": 1.078,
      "grad_norm": 6.115739822387695,
      "learning_rate": 2.7797979797979802e-05,
      "loss": 0.1497,
      "step": 4496
    },
    {
      "epoch": 1.0781,
      "grad_norm": 5.7225871086120605,
      "learning_rate": 2.779292929292929e-05,
      "loss": 0.8858,
      "step": 4497
    },
    {
      "epoch": 1.0782,
      "grad_norm": 6.646703243255615,
      "learning_rate": 2.778787878787879e-05,
      "loss": 0.722,
      "step": 4498
    },
    {
      "epoch": 1.0783,
      "grad_norm": 4.672988414764404,
      "learning_rate": 2.7782828282828283e-05,
      "loss": 0.4085,
      "step": 4499
    },
    {
      "epoch": 1.0784,
      "grad_norm": 2.637394428253174,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.246,
      "step": 4500
    },
    {
      "epoch": 1.0785,
      "grad_norm": 2.498063087463379,
      "learning_rate": 2.7772727272727272e-05,
      "loss": 0.0958,
      "step": 4501
    },
    {
      "epoch": 1.0786,
      "grad_norm": 5.903378486633301,
      "learning_rate": 2.776767676767677e-05,
      "loss": 0.1526,
      "step": 4502
    },
    {
      "epoch": 1.0787,
      "grad_norm": 6.96174955368042,
      "learning_rate": 2.7762626262626264e-05,
      "loss": 0.0475,
      "step": 4503
    },
    {
      "epoch": 1.0788,
      "grad_norm": 9.666726112365723,
      "learning_rate": 2.775757575757576e-05,
      "loss": 0.1084,
      "step": 4504
    },
    {
      "epoch": 1.0789,
      "grad_norm": 4.222426414489746,
      "learning_rate": 2.7752525252525252e-05,
      "loss": 0.5791,
      "step": 4505
    },
    {
      "epoch": 1.079,
      "grad_norm": 19.24481964111328,
      "learning_rate": 2.7747474747474748e-05,
      "loss": 0.2203,
      "step": 4506
    },
    {
      "epoch": 1.0791,
      "grad_norm": 17.132062911987305,
      "learning_rate": 2.774242424242424e-05,
      "loss": 0.1177,
      "step": 4507
    },
    {
      "epoch": 1.0792,
      "grad_norm": 3.599916458129883,
      "learning_rate": 2.773737373737374e-05,
      "loss": 0.096,
      "step": 4508
    },
    {
      "epoch": 1.0793,
      "grad_norm": 12.777718544006348,
      "learning_rate": 2.7732323232323233e-05,
      "loss": 0.4618,
      "step": 4509
    },
    {
      "epoch": 1.0794,
      "grad_norm": 4.060608863830566,
      "learning_rate": 2.772727272727273e-05,
      "loss": 0.1709,
      "step": 4510
    },
    {
      "epoch": 1.0795,
      "grad_norm": 5.172435760498047,
      "learning_rate": 2.772222222222222e-05,
      "loss": 0.3878,
      "step": 4511
    },
    {
      "epoch": 1.0796000000000001,
      "grad_norm": 1.484810471534729,
      "learning_rate": 2.771717171717172e-05,
      "loss": 0.0572,
      "step": 4512
    },
    {
      "epoch": 1.0796999999999999,
      "grad_norm": 5.558531761169434,
      "learning_rate": 2.771212121212121e-05,
      "loss": 0.27,
      "step": 4513
    },
    {
      "epoch": 1.0798,
      "grad_norm": 5.569042205810547,
      "learning_rate": 2.770707070707071e-05,
      "loss": 0.1912,
      "step": 4514
    },
    {
      "epoch": 1.0799,
      "grad_norm": 12.576812744140625,
      "learning_rate": 2.7702020202020202e-05,
      "loss": 0.4527,
      "step": 4515
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.7201738357543945,
      "learning_rate": 2.7696969696969698e-05,
      "loss": 0.1352,
      "step": 4516
    },
    {
      "epoch": 1.0801,
      "grad_norm": 3.643437623977661,
      "learning_rate": 2.769191919191919e-05,
      "loss": 0.5947,
      "step": 4517
    },
    {
      "epoch": 1.0802,
      "grad_norm": 11.3472318649292,
      "learning_rate": 2.768686868686869e-05,
      "loss": 1.3834,
      "step": 4518
    },
    {
      "epoch": 1.0803,
      "grad_norm": 5.562522888183594,
      "learning_rate": 2.7681818181818183e-05,
      "loss": 0.3678,
      "step": 4519
    },
    {
      "epoch": 1.0804,
      "grad_norm": 7.411900997161865,
      "learning_rate": 2.767676767676768e-05,
      "loss": 0.3533,
      "step": 4520
    },
    {
      "epoch": 1.0805,
      "grad_norm": 9.364681243896484,
      "learning_rate": 2.767171717171717e-05,
      "loss": 0.3623,
      "step": 4521
    },
    {
      "epoch": 1.0806,
      "grad_norm": 2.5941429138183594,
      "learning_rate": 2.7666666666666667e-05,
      "loss": 0.1014,
      "step": 4522
    },
    {
      "epoch": 1.0807,
      "grad_norm": 3.0987730026245117,
      "learning_rate": 2.766161616161616e-05,
      "loss": 0.1033,
      "step": 4523
    },
    {
      "epoch": 1.0808,
      "grad_norm": 6.796581268310547,
      "learning_rate": 2.765656565656566e-05,
      "loss": 0.8171,
      "step": 4524
    },
    {
      "epoch": 1.0809,
      "grad_norm": 2.4386985301971436,
      "learning_rate": 2.7651515151515152e-05,
      "loss": 0.0698,
      "step": 4525
    },
    {
      "epoch": 1.081,
      "grad_norm": 4.233391284942627,
      "learning_rate": 2.7646464646464648e-05,
      "loss": 0.0723,
      "step": 4526
    },
    {
      "epoch": 1.0811,
      "grad_norm": 5.997183799743652,
      "learning_rate": 2.764141414141414e-05,
      "loss": 0.2005,
      "step": 4527
    },
    {
      "epoch": 1.0812,
      "grad_norm": 5.026284217834473,
      "learning_rate": 2.7636363636363636e-05,
      "loss": 0.7034,
      "step": 4528
    },
    {
      "epoch": 1.0813,
      "grad_norm": 13.51257610321045,
      "learning_rate": 2.763131313131313e-05,
      "loss": 0.3913,
      "step": 4529
    },
    {
      "epoch": 1.0814,
      "grad_norm": 4.365715026855469,
      "learning_rate": 2.762626262626263e-05,
      "loss": 0.4095,
      "step": 4530
    },
    {
      "epoch": 1.0815,
      "grad_norm": 6.396525859832764,
      "learning_rate": 2.762121212121212e-05,
      "loss": 0.3288,
      "step": 4531
    },
    {
      "epoch": 1.0816,
      "grad_norm": 4.0073370933532715,
      "learning_rate": 2.7616161616161617e-05,
      "loss": 0.1323,
      "step": 4532
    },
    {
      "epoch": 1.0817,
      "grad_norm": 14.901534080505371,
      "learning_rate": 2.761111111111111e-05,
      "loss": 0.817,
      "step": 4533
    },
    {
      "epoch": 1.0818,
      "grad_norm": 5.303347587585449,
      "learning_rate": 2.760606060606061e-05,
      "loss": 0.1614,
      "step": 4534
    },
    {
      "epoch": 1.0819,
      "grad_norm": 1.5548290014266968,
      "learning_rate": 2.7601010101010098e-05,
      "loss": 0.0538,
      "step": 4535
    },
    {
      "epoch": 1.082,
      "grad_norm": 1.9665848016738892,
      "learning_rate": 2.7595959595959597e-05,
      "loss": 0.031,
      "step": 4536
    },
    {
      "epoch": 1.0821,
      "grad_norm": 10.82552433013916,
      "learning_rate": 2.759090909090909e-05,
      "loss": 0.2794,
      "step": 4537
    },
    {
      "epoch": 1.0822,
      "grad_norm": 4.5636186599731445,
      "learning_rate": 2.7585858585858586e-05,
      "loss": 0.0447,
      "step": 4538
    },
    {
      "epoch": 1.0823,
      "grad_norm": 5.0019636154174805,
      "learning_rate": 2.7580808080808085e-05,
      "loss": 0.644,
      "step": 4539
    },
    {
      "epoch": 1.0824,
      "grad_norm": 5.01356840133667,
      "learning_rate": 2.7575757575757578e-05,
      "loss": 1.0847,
      "step": 4540
    },
    {
      "epoch": 1.0825,
      "grad_norm": 3.9238698482513428,
      "learning_rate": 2.7570707070707074e-05,
      "loss": 0.6221,
      "step": 4541
    },
    {
      "epoch": 1.0826,
      "grad_norm": 3.4767627716064453,
      "learning_rate": 2.7565656565656567e-05,
      "loss": 0.0758,
      "step": 4542
    },
    {
      "epoch": 1.0827,
      "grad_norm": 4.9264960289001465,
      "learning_rate": 2.7560606060606066e-05,
      "loss": 0.326,
      "step": 4543
    },
    {
      "epoch": 1.0828,
      "grad_norm": 5.187155723571777,
      "learning_rate": 2.7555555555555555e-05,
      "loss": 0.2277,
      "step": 4544
    },
    {
      "epoch": 1.0829,
      "grad_norm": 4.748517036437988,
      "learning_rate": 2.7550505050505055e-05,
      "loss": 0.575,
      "step": 4545
    },
    {
      "epoch": 1.083,
      "grad_norm": 1.3097610473632812,
      "learning_rate": 2.7545454545454547e-05,
      "loss": 0.0289,
      "step": 4546
    },
    {
      "epoch": 1.0831,
      "grad_norm": 7.82279109954834,
      "learning_rate": 2.7540404040404043e-05,
      "loss": 0.3211,
      "step": 4547
    },
    {
      "epoch": 1.0832,
      "grad_norm": 11.1301908493042,
      "learning_rate": 2.7535353535353536e-05,
      "loss": 0.163,
      "step": 4548
    },
    {
      "epoch": 1.0833,
      "grad_norm": 5.770456790924072,
      "learning_rate": 2.7530303030303035e-05,
      "loss": 0.2323,
      "step": 4549
    },
    {
      "epoch": 1.0834,
      "grad_norm": 6.886397838592529,
      "learning_rate": 2.7525252525252528e-05,
      "loss": 0.5342,
      "step": 4550
    },
    {
      "epoch": 1.0835,
      "grad_norm": 3.1356723308563232,
      "learning_rate": 2.7520202020202024e-05,
      "loss": 0.087,
      "step": 4551
    },
    {
      "epoch": 1.0836,
      "grad_norm": 5.199172496795654,
      "learning_rate": 2.7515151515151516e-05,
      "loss": 0.2752,
      "step": 4552
    },
    {
      "epoch": 1.0836999999999999,
      "grad_norm": 4.715618133544922,
      "learning_rate": 2.7510101010101012e-05,
      "loss": 0.1175,
      "step": 4553
    },
    {
      "epoch": 1.0838,
      "grad_norm": 7.36366605758667,
      "learning_rate": 2.7505050505050505e-05,
      "loss": 0.1304,
      "step": 4554
    },
    {
      "epoch": 1.0839,
      "grad_norm": 4.032872676849365,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2747,
      "step": 4555
    },
    {
      "epoch": 1.084,
      "grad_norm": 1.756887674331665,
      "learning_rate": 2.7494949494949497e-05,
      "loss": 0.0534,
      "step": 4556
    },
    {
      "epoch": 1.0841,
      "grad_norm": 46.683380126953125,
      "learning_rate": 2.7489898989898993e-05,
      "loss": 0.3122,
      "step": 4557
    },
    {
      "epoch": 1.0842,
      "grad_norm": 2.600106716156006,
      "learning_rate": 2.7484848484848486e-05,
      "loss": 0.525,
      "step": 4558
    },
    {
      "epoch": 1.0843,
      "grad_norm": 3.6260180473327637,
      "learning_rate": 2.747979797979798e-05,
      "loss": 0.3613,
      "step": 4559
    },
    {
      "epoch": 1.0844,
      "grad_norm": 4.167929649353027,
      "learning_rate": 2.7474747474747474e-05,
      "loss": 0.1587,
      "step": 4560
    },
    {
      "epoch": 1.0845,
      "grad_norm": 5.977355003356934,
      "learning_rate": 2.7469696969696974e-05,
      "loss": 0.1328,
      "step": 4561
    },
    {
      "epoch": 1.0846,
      "grad_norm": 3.858808755874634,
      "learning_rate": 2.7464646464646466e-05,
      "loss": 0.276,
      "step": 4562
    },
    {
      "epoch": 1.0847,
      "grad_norm": 5.957200527191162,
      "learning_rate": 2.7459595959595962e-05,
      "loss": 0.1242,
      "step": 4563
    },
    {
      "epoch": 1.0848,
      "grad_norm": 31.594619750976562,
      "learning_rate": 2.7454545454545455e-05,
      "loss": 0.2064,
      "step": 4564
    },
    {
      "epoch": 1.0849,
      "grad_norm": 1.9893351793289185,
      "learning_rate": 2.7449494949494954e-05,
      "loss": 0.5113,
      "step": 4565
    },
    {
      "epoch": 1.085,
      "grad_norm": 8.057888984680176,
      "learning_rate": 2.7444444444444443e-05,
      "loss": 0.353,
      "step": 4566
    },
    {
      "epoch": 1.0851,
      "grad_norm": 8.596343994140625,
      "learning_rate": 2.7439393939393943e-05,
      "loss": 0.3788,
      "step": 4567
    },
    {
      "epoch": 1.0852,
      "grad_norm": 3.6157572269439697,
      "learning_rate": 2.7434343434343435e-05,
      "loss": 0.2819,
      "step": 4568
    },
    {
      "epoch": 1.0853,
      "grad_norm": 4.778983116149902,
      "learning_rate": 2.742929292929293e-05,
      "loss": 0.5734,
      "step": 4569
    },
    {
      "epoch": 1.0854,
      "grad_norm": 5.066657066345215,
      "learning_rate": 2.7424242424242424e-05,
      "loss": 0.2581,
      "step": 4570
    },
    {
      "epoch": 1.0855,
      "grad_norm": 13.723637580871582,
      "learning_rate": 2.7419191919191923e-05,
      "loss": 0.8749,
      "step": 4571
    },
    {
      "epoch": 1.0856,
      "grad_norm": 4.7164626121521,
      "learning_rate": 2.7414141414141416e-05,
      "loss": 0.2277,
      "step": 4572
    },
    {
      "epoch": 1.0857,
      "grad_norm": 7.394711971282959,
      "learning_rate": 2.7409090909090912e-05,
      "loss": 0.6309,
      "step": 4573
    },
    {
      "epoch": 1.0858,
      "grad_norm": 4.1672682762146,
      "learning_rate": 2.7404040404040404e-05,
      "loss": 0.2744,
      "step": 4574
    },
    {
      "epoch": 1.0859,
      "grad_norm": 4.047881126403809,
      "learning_rate": 2.73989898989899e-05,
      "loss": 0.3588,
      "step": 4575
    },
    {
      "epoch": 1.086,
      "grad_norm": 2.8810458183288574,
      "learning_rate": 2.7393939393939393e-05,
      "loss": 0.0824,
      "step": 4576
    },
    {
      "epoch": 1.0861,
      "grad_norm": 6.530353546142578,
      "learning_rate": 2.7388888888888892e-05,
      "loss": 0.1587,
      "step": 4577
    },
    {
      "epoch": 1.0862,
      "grad_norm": 2.325427532196045,
      "learning_rate": 2.7383838383838385e-05,
      "loss": 0.0296,
      "step": 4578
    },
    {
      "epoch": 1.0863,
      "grad_norm": 1.8143789768218994,
      "learning_rate": 2.737878787878788e-05,
      "loss": 0.056,
      "step": 4579
    },
    {
      "epoch": 1.0864,
      "grad_norm": 3.867959976196289,
      "learning_rate": 2.7373737373737374e-05,
      "loss": 0.1244,
      "step": 4580
    },
    {
      "epoch": 1.0865,
      "grad_norm": 4.980330944061279,
      "learning_rate": 2.7368686868686873e-05,
      "loss": 0.6019,
      "step": 4581
    },
    {
      "epoch": 1.0866,
      "grad_norm": 0.8483951091766357,
      "learning_rate": 2.7363636363636362e-05,
      "loss": 0.0114,
      "step": 4582
    },
    {
      "epoch": 1.0867,
      "grad_norm": 5.016300678253174,
      "learning_rate": 2.735858585858586e-05,
      "loss": 0.2098,
      "step": 4583
    },
    {
      "epoch": 1.0868,
      "grad_norm": 4.1486005783081055,
      "learning_rate": 2.7353535353535354e-05,
      "loss": 0.1546,
      "step": 4584
    },
    {
      "epoch": 1.0869,
      "grad_norm": 5.181750774383545,
      "learning_rate": 2.734848484848485e-05,
      "loss": 0.1414,
      "step": 4585
    },
    {
      "epoch": 1.087,
      "grad_norm": 9.538479804992676,
      "learning_rate": 2.7343434343434343e-05,
      "loss": 0.1073,
      "step": 4586
    },
    {
      "epoch": 1.0871,
      "grad_norm": 6.239503383636475,
      "learning_rate": 2.7338383838383842e-05,
      "loss": 0.6784,
      "step": 4587
    },
    {
      "epoch": 1.0872,
      "grad_norm": 1.2643758058547974,
      "learning_rate": 2.733333333333333e-05,
      "loss": 0.0224,
      "step": 4588
    },
    {
      "epoch": 1.0873,
      "grad_norm": 3.3596627712249756,
      "learning_rate": 2.732828282828283e-05,
      "loss": 0.2668,
      "step": 4589
    },
    {
      "epoch": 1.0874,
      "grad_norm": 12.461421012878418,
      "learning_rate": 2.7323232323232323e-05,
      "loss": 0.6737,
      "step": 4590
    },
    {
      "epoch": 1.0875,
      "grad_norm": 5.01960563659668,
      "learning_rate": 2.731818181818182e-05,
      "loss": 0.7002,
      "step": 4591
    },
    {
      "epoch": 1.0876,
      "grad_norm": 4.413238048553467,
      "learning_rate": 2.7313131313131312e-05,
      "loss": 0.1455,
      "step": 4592
    },
    {
      "epoch": 1.0877,
      "grad_norm": 2.740668773651123,
      "learning_rate": 2.730808080808081e-05,
      "loss": 0.5088,
      "step": 4593
    },
    {
      "epoch": 1.0878,
      "grad_norm": 5.771108150482178,
      "learning_rate": 2.7303030303030304e-05,
      "loss": 0.2456,
      "step": 4594
    },
    {
      "epoch": 1.0879,
      "grad_norm": 5.270373344421387,
      "learning_rate": 2.72979797979798e-05,
      "loss": 0.3007,
      "step": 4595
    },
    {
      "epoch": 1.088,
      "grad_norm": 5.082179546356201,
      "learning_rate": 2.7292929292929293e-05,
      "loss": 0.3325,
      "step": 4596
    },
    {
      "epoch": 1.0881,
      "grad_norm": 3.1292903423309326,
      "learning_rate": 2.728787878787879e-05,
      "loss": 0.1445,
      "step": 4597
    },
    {
      "epoch": 1.0882,
      "grad_norm": 2.518357992172241,
      "learning_rate": 2.728282828282828e-05,
      "loss": 0.5269,
      "step": 4598
    },
    {
      "epoch": 1.0883,
      "grad_norm": 2.9777560234069824,
      "learning_rate": 2.727777777777778e-05,
      "loss": 0.0972,
      "step": 4599
    },
    {
      "epoch": 1.0884,
      "grad_norm": 5.875877857208252,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.3274,
      "step": 4600
    },
    {
      "epoch": 1.0885,
      "grad_norm": 8.684350967407227,
      "learning_rate": 2.726767676767677e-05,
      "loss": 0.7824,
      "step": 4601
    },
    {
      "epoch": 1.0886,
      "grad_norm": 3.4946091175079346,
      "learning_rate": 2.7262626262626262e-05,
      "loss": 0.0497,
      "step": 4602
    },
    {
      "epoch": 1.0887,
      "grad_norm": 25.43962287902832,
      "learning_rate": 2.725757575757576e-05,
      "loss": 0.1418,
      "step": 4603
    },
    {
      "epoch": 1.0888,
      "grad_norm": 5.324992656707764,
      "learning_rate": 2.725252525252525e-05,
      "loss": 0.1334,
      "step": 4604
    },
    {
      "epoch": 1.0889,
      "grad_norm": 4.155129432678223,
      "learning_rate": 2.724747474747475e-05,
      "loss": 0.2094,
      "step": 4605
    },
    {
      "epoch": 1.089,
      "grad_norm": 1.5673807859420776,
      "learning_rate": 2.7242424242424242e-05,
      "loss": 0.0427,
      "step": 4606
    },
    {
      "epoch": 1.0891,
      "grad_norm": 1.2649624347686768,
      "learning_rate": 2.7237373737373738e-05,
      "loss": 0.0131,
      "step": 4607
    },
    {
      "epoch": 1.0892,
      "grad_norm": 5.458670616149902,
      "learning_rate": 2.723232323232323e-05,
      "loss": 0.0929,
      "step": 4608
    },
    {
      "epoch": 1.0893,
      "grad_norm": 8.956260681152344,
      "learning_rate": 2.722727272727273e-05,
      "loss": 0.0888,
      "step": 4609
    },
    {
      "epoch": 1.0894,
      "grad_norm": 24.12245750427246,
      "learning_rate": 2.7222222222222223e-05,
      "loss": 0.5233,
      "step": 4610
    },
    {
      "epoch": 1.0895,
      "grad_norm": 4.2306437492370605,
      "learning_rate": 2.721717171717172e-05,
      "loss": 0.6102,
      "step": 4611
    },
    {
      "epoch": 1.0896,
      "grad_norm": 4.421792507171631,
      "learning_rate": 2.721212121212121e-05,
      "loss": 0.1123,
      "step": 4612
    },
    {
      "epoch": 1.0897000000000001,
      "grad_norm": 6.966654300689697,
      "learning_rate": 2.7207070707070707e-05,
      "loss": 0.1022,
      "step": 4613
    },
    {
      "epoch": 1.0898,
      "grad_norm": 5.583467960357666,
      "learning_rate": 2.72020202020202e-05,
      "loss": 0.331,
      "step": 4614
    },
    {
      "epoch": 1.0899,
      "grad_norm": 2.1321117877960205,
      "learning_rate": 2.71969696969697e-05,
      "loss": 0.4858,
      "step": 4615
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.836759567260742,
      "learning_rate": 2.7191919191919192e-05,
      "loss": 0.7779,
      "step": 4616
    },
    {
      "epoch": 1.0901,
      "grad_norm": 11.85673999786377,
      "learning_rate": 2.7186868686868688e-05,
      "loss": 0.3978,
      "step": 4617
    },
    {
      "epoch": 1.0902,
      "grad_norm": 3.864471435546875,
      "learning_rate": 2.718181818181818e-05,
      "loss": 0.1442,
      "step": 4618
    },
    {
      "epoch": 1.0903,
      "grad_norm": 7.8743672370910645,
      "learning_rate": 2.7176767676767677e-05,
      "loss": 0.081,
      "step": 4619
    },
    {
      "epoch": 1.0904,
      "grad_norm": 2.815925359725952,
      "learning_rate": 2.717171717171717e-05,
      "loss": 0.2157,
      "step": 4620
    },
    {
      "epoch": 1.0905,
      "grad_norm": 5.72878885269165,
      "learning_rate": 2.716666666666667e-05,
      "loss": 0.6712,
      "step": 4621
    },
    {
      "epoch": 1.0906,
      "grad_norm": 2.5734410285949707,
      "learning_rate": 2.716161616161616e-05,
      "loss": 0.0984,
      "step": 4622
    },
    {
      "epoch": 1.0907,
      "grad_norm": 1.758327841758728,
      "learning_rate": 2.7156565656565657e-05,
      "loss": 0.054,
      "step": 4623
    },
    {
      "epoch": 1.0908,
      "grad_norm": 5.060421943664551,
      "learning_rate": 2.7151515151515157e-05,
      "loss": 0.1145,
      "step": 4624
    },
    {
      "epoch": 1.0909,
      "grad_norm": 4.769593715667725,
      "learning_rate": 2.714646464646465e-05,
      "loss": 0.5553,
      "step": 4625
    },
    {
      "epoch": 1.091,
      "grad_norm": 5.55781364440918,
      "learning_rate": 2.7141414141414145e-05,
      "loss": 0.6003,
      "step": 4626
    },
    {
      "epoch": 1.0911,
      "grad_norm": 9.701210975646973,
      "learning_rate": 2.7136363636363638e-05,
      "loss": 0.1326,
      "step": 4627
    },
    {
      "epoch": 1.0912,
      "grad_norm": 6.771081924438477,
      "learning_rate": 2.7131313131313134e-05,
      "loss": 0.5986,
      "step": 4628
    },
    {
      "epoch": 1.0913,
      "grad_norm": 4.255305767059326,
      "learning_rate": 2.7126262626262626e-05,
      "loss": 0.1836,
      "step": 4629
    },
    {
      "epoch": 1.0914,
      "grad_norm": 4.614716053009033,
      "learning_rate": 2.7121212121212126e-05,
      "loss": 0.1272,
      "step": 4630
    },
    {
      "epoch": 1.0915,
      "grad_norm": 5.065900802612305,
      "learning_rate": 2.711616161616162e-05,
      "loss": 0.7681,
      "step": 4631
    },
    {
      "epoch": 1.0916,
      "grad_norm": 4.657528877258301,
      "learning_rate": 2.7111111111111114e-05,
      "loss": 0.2073,
      "step": 4632
    },
    {
      "epoch": 1.0917,
      "grad_norm": 3.6939823627471924,
      "learning_rate": 2.7106060606060607e-05,
      "loss": 0.16,
      "step": 4633
    },
    {
      "epoch": 1.0918,
      "grad_norm": 2.2165753841400146,
      "learning_rate": 2.7101010101010106e-05,
      "loss": 0.0628,
      "step": 4634
    },
    {
      "epoch": 1.0919,
      "grad_norm": 4.813558101654053,
      "learning_rate": 2.7095959595959596e-05,
      "loss": 0.7803,
      "step": 4635
    },
    {
      "epoch": 1.092,
      "grad_norm": 3.9919021129608154,
      "learning_rate": 2.7090909090909095e-05,
      "loss": 0.3464,
      "step": 4636
    },
    {
      "epoch": 1.0921,
      "grad_norm": 7.0689873695373535,
      "learning_rate": 2.7085858585858587e-05,
      "loss": 0.1004,
      "step": 4637
    },
    {
      "epoch": 1.0922,
      "grad_norm": 2.207122325897217,
      "learning_rate": 2.7080808080808083e-05,
      "loss": 0.058,
      "step": 4638
    },
    {
      "epoch": 1.0923,
      "grad_norm": 1.6250667572021484,
      "learning_rate": 2.7075757575757576e-05,
      "loss": 0.0249,
      "step": 4639
    },
    {
      "epoch": 1.0924,
      "grad_norm": 10.26284122467041,
      "learning_rate": 2.7070707070707075e-05,
      "loss": 0.6696,
      "step": 4640
    },
    {
      "epoch": 1.0925,
      "grad_norm": 2.7226457595825195,
      "learning_rate": 2.7065656565656568e-05,
      "loss": 0.0679,
      "step": 4641
    },
    {
      "epoch": 1.0926,
      "grad_norm": 3.314957857131958,
      "learning_rate": 2.7060606060606064e-05,
      "loss": 0.3482,
      "step": 4642
    },
    {
      "epoch": 1.0927,
      "grad_norm": 6.564191818237305,
      "learning_rate": 2.7055555555555557e-05,
      "loss": 0.3951,
      "step": 4643
    },
    {
      "epoch": 1.0928,
      "grad_norm": 8.232748985290527,
      "learning_rate": 2.7050505050505053e-05,
      "loss": 0.6064,
      "step": 4644
    },
    {
      "epoch": 1.0929,
      "grad_norm": 3.425788402557373,
      "learning_rate": 2.7045454545454545e-05,
      "loss": 0.3126,
      "step": 4645
    },
    {
      "epoch": 1.093,
      "grad_norm": 4.342231273651123,
      "learning_rate": 2.7040404040404045e-05,
      "loss": 0.1455,
      "step": 4646
    },
    {
      "epoch": 1.0931,
      "grad_norm": 4.138504981994629,
      "learning_rate": 2.7035353535353537e-05,
      "loss": 0.0556,
      "step": 4647
    },
    {
      "epoch": 1.0932,
      "grad_norm": 4.218878746032715,
      "learning_rate": 2.7030303030303033e-05,
      "loss": 0.1188,
      "step": 4648
    },
    {
      "epoch": 1.0933,
      "grad_norm": 2.7341651916503906,
      "learning_rate": 2.7025252525252526e-05,
      "loss": 0.4856,
      "step": 4649
    },
    {
      "epoch": 1.0934,
      "grad_norm": 2.870980739593506,
      "learning_rate": 2.7020202020202022e-05,
      "loss": 0.0829,
      "step": 4650
    },
    {
      "epoch": 1.0935,
      "grad_norm": 2.028167486190796,
      "learning_rate": 2.7015151515151514e-05,
      "loss": 0.2315,
      "step": 4651
    },
    {
      "epoch": 1.0936,
      "grad_norm": 14.697094917297363,
      "learning_rate": 2.7010101010101014e-05,
      "loss": 0.3967,
      "step": 4652
    },
    {
      "epoch": 1.0937000000000001,
      "grad_norm": 6.385424613952637,
      "learning_rate": 2.7005050505050506e-05,
      "loss": 0.7878,
      "step": 4653
    },
    {
      "epoch": 1.0937999999999999,
      "grad_norm": 4.445223808288574,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.0856,
      "step": 4654
    },
    {
      "epoch": 1.0939,
      "grad_norm": 4.058835983276367,
      "learning_rate": 2.6994949494949495e-05,
      "loss": 0.7065,
      "step": 4655
    },
    {
      "epoch": 1.094,
      "grad_norm": 3.1801095008850098,
      "learning_rate": 2.6989898989898994e-05,
      "loss": 0.1272,
      "step": 4656
    },
    {
      "epoch": 1.0941,
      "grad_norm": 5.968142986297607,
      "learning_rate": 2.6984848484848484e-05,
      "loss": 0.3551,
      "step": 4657
    },
    {
      "epoch": 1.0942,
      "grad_norm": 4.905891418457031,
      "learning_rate": 2.6979797979797983e-05,
      "loss": 0.5595,
      "step": 4658
    },
    {
      "epoch": 1.0943,
      "grad_norm": 3.6735832691192627,
      "learning_rate": 2.6974747474747476e-05,
      "loss": 0.1118,
      "step": 4659
    },
    {
      "epoch": 1.0944,
      "grad_norm": 6.067805767059326,
      "learning_rate": 2.696969696969697e-05,
      "loss": 0.1455,
      "step": 4660
    },
    {
      "epoch": 1.0945,
      "grad_norm": 1.931154727935791,
      "learning_rate": 2.6964646464646464e-05,
      "loss": 0.0367,
      "step": 4661
    },
    {
      "epoch": 1.0946,
      "grad_norm": 4.836620330810547,
      "learning_rate": 2.6959595959595964e-05,
      "loss": 0.2558,
      "step": 4662
    },
    {
      "epoch": 1.0947,
      "grad_norm": 3.893024206161499,
      "learning_rate": 2.6954545454545456e-05,
      "loss": 0.0896,
      "step": 4663
    },
    {
      "epoch": 1.0948,
      "grad_norm": 7.146899223327637,
      "learning_rate": 2.6949494949494952e-05,
      "loss": 0.5896,
      "step": 4664
    },
    {
      "epoch": 1.0949,
      "grad_norm": 3.5448012351989746,
      "learning_rate": 2.6944444444444445e-05,
      "loss": 0.6335,
      "step": 4665
    },
    {
      "epoch": 1.095,
      "grad_norm": 4.290581703186035,
      "learning_rate": 2.693939393939394e-05,
      "loss": 0.5726,
      "step": 4666
    },
    {
      "epoch": 1.0951,
      "grad_norm": 3.2904953956604004,
      "learning_rate": 2.6934343434343433e-05,
      "loss": 0.2502,
      "step": 4667
    },
    {
      "epoch": 1.0952,
      "grad_norm": 10.657032012939453,
      "learning_rate": 2.6929292929292933e-05,
      "loss": 0.3868,
      "step": 4668
    },
    {
      "epoch": 1.0953,
      "grad_norm": 4.15410852432251,
      "learning_rate": 2.6924242424242425e-05,
      "loss": 0.0682,
      "step": 4669
    },
    {
      "epoch": 1.0954,
      "grad_norm": 6.461917877197266,
      "learning_rate": 2.691919191919192e-05,
      "loss": 0.0951,
      "step": 4670
    },
    {
      "epoch": 1.0955,
      "grad_norm": 5.401758670806885,
      "learning_rate": 2.6914141414141414e-05,
      "loss": 0.1711,
      "step": 4671
    },
    {
      "epoch": 1.0956,
      "grad_norm": 6.640973091125488,
      "learning_rate": 2.6909090909090913e-05,
      "loss": 0.2389,
      "step": 4672
    },
    {
      "epoch": 1.0957,
      "grad_norm": 3.9821364879608154,
      "learning_rate": 2.6904040404040402e-05,
      "loss": 0.2943,
      "step": 4673
    },
    {
      "epoch": 1.0958,
      "grad_norm": 5.19147253036499,
      "learning_rate": 2.6898989898989902e-05,
      "loss": 0.4812,
      "step": 4674
    },
    {
      "epoch": 1.0959,
      "grad_norm": 2.2747137546539307,
      "learning_rate": 2.6893939393939394e-05,
      "loss": 0.0721,
      "step": 4675
    },
    {
      "epoch": 1.096,
      "grad_norm": 2.9567625522613525,
      "learning_rate": 2.688888888888889e-05,
      "loss": 0.2967,
      "step": 4676
    },
    {
      "epoch": 1.0961,
      "grad_norm": 9.025284767150879,
      "learning_rate": 2.6883838383838383e-05,
      "loss": 0.3418,
      "step": 4677
    },
    {
      "epoch": 1.0962,
      "grad_norm": 4.676849842071533,
      "learning_rate": 2.6878787878787882e-05,
      "loss": 0.1398,
      "step": 4678
    },
    {
      "epoch": 1.0963,
      "grad_norm": 1.5845946073532104,
      "learning_rate": 2.687373737373737e-05,
      "loss": 0.0337,
      "step": 4679
    },
    {
      "epoch": 1.0964,
      "grad_norm": 1.7855446338653564,
      "learning_rate": 2.686868686868687e-05,
      "loss": 0.0221,
      "step": 4680
    },
    {
      "epoch": 1.0965,
      "grad_norm": 6.108048439025879,
      "learning_rate": 2.6863636363636364e-05,
      "loss": 0.1379,
      "step": 4681
    },
    {
      "epoch": 1.0966,
      "grad_norm": 2.5376269817352295,
      "learning_rate": 2.685858585858586e-05,
      "loss": 0.2455,
      "step": 4682
    },
    {
      "epoch": 1.0967,
      "grad_norm": 2.7299447059631348,
      "learning_rate": 2.6853535353535352e-05,
      "loss": 0.1135,
      "step": 4683
    },
    {
      "epoch": 1.0968,
      "grad_norm": 4.540733337402344,
      "learning_rate": 2.684848484848485e-05,
      "loss": 0.0928,
      "step": 4684
    },
    {
      "epoch": 1.0969,
      "grad_norm": 4.1390485763549805,
      "learning_rate": 2.6843434343434344e-05,
      "loss": 0.2107,
      "step": 4685
    },
    {
      "epoch": 1.097,
      "grad_norm": 7.316167831420898,
      "learning_rate": 2.683838383838384e-05,
      "loss": 0.3787,
      "step": 4686
    },
    {
      "epoch": 1.0971,
      "grad_norm": 2.1619811058044434,
      "learning_rate": 2.6833333333333333e-05,
      "loss": 0.0695,
      "step": 4687
    },
    {
      "epoch": 1.0972,
      "grad_norm": 2.5108931064605713,
      "learning_rate": 2.682828282828283e-05,
      "loss": 0.0635,
      "step": 4688
    },
    {
      "epoch": 1.0973,
      "grad_norm": 7.885334014892578,
      "learning_rate": 2.682323232323232e-05,
      "loss": 0.1692,
      "step": 4689
    },
    {
      "epoch": 1.0974,
      "grad_norm": 1.9064725637435913,
      "learning_rate": 2.681818181818182e-05,
      "loss": 0.0267,
      "step": 4690
    },
    {
      "epoch": 1.0975,
      "grad_norm": 6.8264617919921875,
      "learning_rate": 2.6813131313131313e-05,
      "loss": 0.9885,
      "step": 4691
    },
    {
      "epoch": 1.0976,
      "grad_norm": 7.0961222648620605,
      "learning_rate": 2.680808080808081e-05,
      "loss": 0.2369,
      "step": 4692
    },
    {
      "epoch": 1.0977,
      "grad_norm": 8.284830093383789,
      "learning_rate": 2.6803030303030302e-05,
      "loss": 0.5209,
      "step": 4693
    },
    {
      "epoch": 1.0977999999999999,
      "grad_norm": 10.827559471130371,
      "learning_rate": 2.67979797979798e-05,
      "loss": 0.697,
      "step": 4694
    },
    {
      "epoch": 1.0979,
      "grad_norm": 2.305044412612915,
      "learning_rate": 2.679292929292929e-05,
      "loss": 0.0359,
      "step": 4695
    },
    {
      "epoch": 1.098,
      "grad_norm": 6.810680389404297,
      "learning_rate": 2.678787878787879e-05,
      "loss": 0.3939,
      "step": 4696
    },
    {
      "epoch": 1.0981,
      "grad_norm": 8.063207626342773,
      "learning_rate": 2.6782828282828283e-05,
      "loss": 0.9844,
      "step": 4697
    },
    {
      "epoch": 1.0982,
      "grad_norm": 8.863363265991211,
      "learning_rate": 2.677777777777778e-05,
      "loss": 0.1158,
      "step": 4698
    },
    {
      "epoch": 1.0983,
      "grad_norm": 2.8922369480133057,
      "learning_rate": 2.677272727272727e-05,
      "loss": 0.0714,
      "step": 4699
    },
    {
      "epoch": 1.0984,
      "grad_norm": 3.0637359619140625,
      "learning_rate": 2.676767676767677e-05,
      "loss": 0.2725,
      "step": 4700
    },
    {
      "epoch": 1.0985,
      "grad_norm": 11.083717346191406,
      "learning_rate": 2.676262626262626e-05,
      "loss": 0.1374,
      "step": 4701
    },
    {
      "epoch": 1.0986,
      "grad_norm": 4.997063636779785,
      "learning_rate": 2.675757575757576e-05,
      "loss": 0.7209,
      "step": 4702
    },
    {
      "epoch": 1.0987,
      "grad_norm": 5.824921607971191,
      "learning_rate": 2.6752525252525252e-05,
      "loss": 0.1204,
      "step": 4703
    },
    {
      "epoch": 1.0988,
      "grad_norm": 5.970680236816406,
      "learning_rate": 2.6747474747474748e-05,
      "loss": 0.2329,
      "step": 4704
    },
    {
      "epoch": 1.0989,
      "grad_norm": 4.352779388427734,
      "learning_rate": 2.674242424242424e-05,
      "loss": 0.5395,
      "step": 4705
    },
    {
      "epoch": 1.099,
      "grad_norm": 2.0886614322662354,
      "learning_rate": 2.673737373737374e-05,
      "loss": 0.2288,
      "step": 4706
    },
    {
      "epoch": 1.0991,
      "grad_norm": 1.045113444328308,
      "learning_rate": 2.6732323232323232e-05,
      "loss": 0.0177,
      "step": 4707
    },
    {
      "epoch": 1.0992,
      "grad_norm": 2.68735671043396,
      "learning_rate": 2.6727272727272728e-05,
      "loss": 0.0963,
      "step": 4708
    },
    {
      "epoch": 1.0993,
      "grad_norm": 2.180176019668579,
      "learning_rate": 2.6722222222222228e-05,
      "loss": 0.2008,
      "step": 4709
    },
    {
      "epoch": 1.0994,
      "grad_norm": 13.112486839294434,
      "learning_rate": 2.6717171717171717e-05,
      "loss": 0.3696,
      "step": 4710
    },
    {
      "epoch": 1.0995,
      "grad_norm": 1.3146841526031494,
      "learning_rate": 2.6712121212121216e-05,
      "loss": 0.0042,
      "step": 4711
    },
    {
      "epoch": 1.0996,
      "grad_norm": 4.584450721740723,
      "learning_rate": 2.670707070707071e-05,
      "loss": 0.1284,
      "step": 4712
    },
    {
      "epoch": 1.0997,
      "grad_norm": 15.740701675415039,
      "learning_rate": 2.6702020202020205e-05,
      "loss": 0.2741,
      "step": 4713
    },
    {
      "epoch": 1.0998,
      "grad_norm": 3.0387277603149414,
      "learning_rate": 2.6696969696969697e-05,
      "loss": 0.057,
      "step": 4714
    },
    {
      "epoch": 1.0999,
      "grad_norm": 10.267152786254883,
      "learning_rate": 2.6691919191919197e-05,
      "loss": 0.2278,
      "step": 4715
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.353453278541565,
      "learning_rate": 2.668686868686869e-05,
      "loss": 0.0334,
      "step": 4716
    },
    {
      "epoch": 1.1001,
      "grad_norm": 1.5547707080841064,
      "learning_rate": 2.6681818181818185e-05,
      "loss": 0.0236,
      "step": 4717
    },
    {
      "epoch": 1.1002,
      "grad_norm": 3.0871219635009766,
      "learning_rate": 2.6676767676767678e-05,
      "loss": 0.0897,
      "step": 4718
    },
    {
      "epoch": 1.1003,
      "grad_norm": 7.916445732116699,
      "learning_rate": 2.6671717171717174e-05,
      "loss": 0.3943,
      "step": 4719
    },
    {
      "epoch": 1.1004,
      "grad_norm": 5.209049224853516,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.5568,
      "step": 4720
    },
    {
      "epoch": 1.1005,
      "grad_norm": 4.295945167541504,
      "learning_rate": 2.6661616161616166e-05,
      "loss": 1.0999,
      "step": 4721
    },
    {
      "epoch": 1.1006,
      "grad_norm": 30.853485107421875,
      "learning_rate": 2.665656565656566e-05,
      "loss": 0.5788,
      "step": 4722
    },
    {
      "epoch": 1.1007,
      "grad_norm": 2.9783289432525635,
      "learning_rate": 2.6651515151515155e-05,
      "loss": 0.067,
      "step": 4723
    },
    {
      "epoch": 1.1008,
      "grad_norm": 4.878079891204834,
      "learning_rate": 2.6646464646464647e-05,
      "loss": 0.0665,
      "step": 4724
    },
    {
      "epoch": 1.1009,
      "grad_norm": 9.902593612670898,
      "learning_rate": 2.6641414141414147e-05,
      "loss": 0.1304,
      "step": 4725
    },
    {
      "epoch": 1.101,
      "grad_norm": 3.9378278255462646,
      "learning_rate": 2.6636363636363636e-05,
      "loss": 0.0287,
      "step": 4726
    },
    {
      "epoch": 1.1011,
      "grad_norm": 4.531147480010986,
      "learning_rate": 2.6631313131313135e-05,
      "loss": 0.3662,
      "step": 4727
    },
    {
      "epoch": 1.1012,
      "grad_norm": 2.3164846897125244,
      "learning_rate": 2.6626262626262628e-05,
      "loss": 0.0254,
      "step": 4728
    },
    {
      "epoch": 1.1013,
      "grad_norm": 3.2973008155822754,
      "learning_rate": 2.6621212121212124e-05,
      "loss": 0.1728,
      "step": 4729
    },
    {
      "epoch": 1.1014,
      "grad_norm": 5.798940181732178,
      "learning_rate": 2.6616161616161616e-05,
      "loss": 0.3768,
      "step": 4730
    },
    {
      "epoch": 1.1015,
      "grad_norm": 2.282796859741211,
      "learning_rate": 2.6611111111111116e-05,
      "loss": 0.5076,
      "step": 4731
    },
    {
      "epoch": 1.1016,
      "grad_norm": 5.643033981323242,
      "learning_rate": 2.6606060606060605e-05,
      "loss": 0.3038,
      "step": 4732
    },
    {
      "epoch": 1.1017,
      "grad_norm": 2.214726686477661,
      "learning_rate": 2.6601010101010104e-05,
      "loss": 0.485,
      "step": 4733
    },
    {
      "epoch": 1.1018,
      "grad_norm": 5.289432048797607,
      "learning_rate": 2.6595959595959597e-05,
      "loss": 0.1485,
      "step": 4734
    },
    {
      "epoch": 1.1019,
      "grad_norm": 5.455615043640137,
      "learning_rate": 2.6590909090909093e-05,
      "loss": 0.1766,
      "step": 4735
    },
    {
      "epoch": 1.102,
      "grad_norm": 2.7778663635253906,
      "learning_rate": 2.6585858585858585e-05,
      "loss": 0.2894,
      "step": 4736
    },
    {
      "epoch": 1.1021,
      "grad_norm": 4.727522373199463,
      "learning_rate": 2.6580808080808085e-05,
      "loss": 0.1746,
      "step": 4737
    },
    {
      "epoch": 1.1022,
      "grad_norm": 5.114647388458252,
      "learning_rate": 2.6575757575757577e-05,
      "loss": 0.1683,
      "step": 4738
    },
    {
      "epoch": 1.1023,
      "grad_norm": 2.105909585952759,
      "learning_rate": 2.6570707070707073e-05,
      "loss": 0.6724,
      "step": 4739
    },
    {
      "epoch": 1.1024,
      "grad_norm": 6.956855297088623,
      "learning_rate": 2.6565656565656566e-05,
      "loss": 0.7994,
      "step": 4740
    },
    {
      "epoch": 1.1025,
      "grad_norm": 4.110389232635498,
      "learning_rate": 2.6560606060606062e-05,
      "loss": 0.2267,
      "step": 4741
    },
    {
      "epoch": 1.1026,
      "grad_norm": 2.241722345352173,
      "learning_rate": 2.6555555555555555e-05,
      "loss": 0.0113,
      "step": 4742
    },
    {
      "epoch": 1.1027,
      "grad_norm": 10.859420776367188,
      "learning_rate": 2.6550505050505054e-05,
      "loss": 0.3392,
      "step": 4743
    },
    {
      "epoch": 1.1028,
      "grad_norm": 6.606860160827637,
      "learning_rate": 2.6545454545454547e-05,
      "loss": 0.3444,
      "step": 4744
    },
    {
      "epoch": 1.1029,
      "grad_norm": 4.470326900482178,
      "learning_rate": 2.6540404040404043e-05,
      "loss": 0.6473,
      "step": 4745
    },
    {
      "epoch": 1.103,
      "grad_norm": 3.18007493019104,
      "learning_rate": 2.6535353535353535e-05,
      "loss": 0.0807,
      "step": 4746
    },
    {
      "epoch": 1.1031,
      "grad_norm": 2.9155893325805664,
      "learning_rate": 2.6530303030303035e-05,
      "loss": 0.1199,
      "step": 4747
    },
    {
      "epoch": 1.1032,
      "grad_norm": 4.748797416687012,
      "learning_rate": 2.6525252525252524e-05,
      "loss": 0.1566,
      "step": 4748
    },
    {
      "epoch": 1.1033,
      "grad_norm": 8.342281341552734,
      "learning_rate": 2.6520202020202023e-05,
      "loss": 0.6327,
      "step": 4749
    },
    {
      "epoch": 1.1034,
      "grad_norm": 8.372477531433105,
      "learning_rate": 2.6515151515151516e-05,
      "loss": 0.3317,
      "step": 4750
    },
    {
      "epoch": 1.1035,
      "grad_norm": 3.4988560676574707,
      "learning_rate": 2.6510101010101012e-05,
      "loss": 0.9876,
      "step": 4751
    },
    {
      "epoch": 1.1036,
      "grad_norm": 2.5799055099487305,
      "learning_rate": 2.6505050505050504e-05,
      "loss": 0.0779,
      "step": 4752
    },
    {
      "epoch": 1.1037,
      "grad_norm": 4.161092281341553,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.3332,
      "step": 4753
    },
    {
      "epoch": 1.1038000000000001,
      "grad_norm": 5.108014106750488,
      "learning_rate": 2.6494949494949496e-05,
      "loss": 0.6809,
      "step": 4754
    },
    {
      "epoch": 1.1039,
      "grad_norm": 3.822514295578003,
      "learning_rate": 2.6489898989898992e-05,
      "loss": 0.1535,
      "step": 4755
    },
    {
      "epoch": 1.104,
      "grad_norm": 5.9822235107421875,
      "learning_rate": 2.6484848484848485e-05,
      "loss": 0.4307,
      "step": 4756
    },
    {
      "epoch": 1.1041,
      "grad_norm": 3.3132760524749756,
      "learning_rate": 2.647979797979798e-05,
      "loss": 0.4967,
      "step": 4757
    },
    {
      "epoch": 1.1042,
      "grad_norm": 4.068281650543213,
      "learning_rate": 2.6474747474747474e-05,
      "loss": 0.2584,
      "step": 4758
    },
    {
      "epoch": 1.1043,
      "grad_norm": 0.6794460415840149,
      "learning_rate": 2.6469696969696973e-05,
      "loss": 0.0134,
      "step": 4759
    },
    {
      "epoch": 1.1044,
      "grad_norm": 3.351680040359497,
      "learning_rate": 2.6464646464646466e-05,
      "loss": 0.094,
      "step": 4760
    },
    {
      "epoch": 1.1045,
      "grad_norm": 2.7186951637268066,
      "learning_rate": 2.645959595959596e-05,
      "loss": 0.069,
      "step": 4761
    },
    {
      "epoch": 1.1046,
      "grad_norm": 10.7376070022583,
      "learning_rate": 2.6454545454545454e-05,
      "loss": 0.181,
      "step": 4762
    },
    {
      "epoch": 1.1047,
      "grad_norm": 1.6022799015045166,
      "learning_rate": 2.644949494949495e-05,
      "loss": 0.0291,
      "step": 4763
    },
    {
      "epoch": 1.1048,
      "grad_norm": 7.684820175170898,
      "learning_rate": 2.6444444444444443e-05,
      "loss": 0.1494,
      "step": 4764
    },
    {
      "epoch": 1.1049,
      "grad_norm": 9.063292503356934,
      "learning_rate": 2.6439393939393942e-05,
      "loss": 1.0569,
      "step": 4765
    },
    {
      "epoch": 1.105,
      "grad_norm": 9.320415496826172,
      "learning_rate": 2.6434343434343435e-05,
      "loss": 0.0137,
      "step": 4766
    },
    {
      "epoch": 1.1051,
      "grad_norm": 3.1290462017059326,
      "learning_rate": 2.642929292929293e-05,
      "loss": 0.1326,
      "step": 4767
    },
    {
      "epoch": 1.1052,
      "grad_norm": 4.533987522125244,
      "learning_rate": 2.6424242424242423e-05,
      "loss": 0.2344,
      "step": 4768
    },
    {
      "epoch": 1.1053,
      "grad_norm": 3.423022508621216,
      "learning_rate": 2.6419191919191923e-05,
      "loss": 0.2274,
      "step": 4769
    },
    {
      "epoch": 1.1054,
      "grad_norm": 1.6402535438537598,
      "learning_rate": 2.6414141414141412e-05,
      "loss": 0.0744,
      "step": 4770
    },
    {
      "epoch": 1.1055,
      "grad_norm": 3.4707820415496826,
      "learning_rate": 2.640909090909091e-05,
      "loss": 0.0655,
      "step": 4771
    },
    {
      "epoch": 1.1056,
      "grad_norm": 2.178475856781006,
      "learning_rate": 2.6404040404040404e-05,
      "loss": 0.1164,
      "step": 4772
    },
    {
      "epoch": 1.1057,
      "grad_norm": 0.951576828956604,
      "learning_rate": 2.63989898989899e-05,
      "loss": 0.4818,
      "step": 4773
    },
    {
      "epoch": 1.1058,
      "grad_norm": 5.542392730712891,
      "learning_rate": 2.6393939393939392e-05,
      "loss": 0.0974,
      "step": 4774
    },
    {
      "epoch": 1.1059,
      "grad_norm": 7.491236209869385,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.883,
      "step": 4775
    },
    {
      "epoch": 1.106,
      "grad_norm": 2.9268033504486084,
      "learning_rate": 2.6383838383838384e-05,
      "loss": 0.5233,
      "step": 4776
    },
    {
      "epoch": 1.1061,
      "grad_norm": 1.0053852796554565,
      "learning_rate": 2.637878787878788e-05,
      "loss": 0.0112,
      "step": 4777
    },
    {
      "epoch": 1.1062,
      "grad_norm": 2.9170756340026855,
      "learning_rate": 2.6373737373737373e-05,
      "loss": 0.1127,
      "step": 4778
    },
    {
      "epoch": 1.1063,
      "grad_norm": 3.6656930446624756,
      "learning_rate": 2.636868686868687e-05,
      "loss": 0.247,
      "step": 4779
    },
    {
      "epoch": 1.1064,
      "grad_norm": 1.396333932876587,
      "learning_rate": 2.636363636363636e-05,
      "loss": 0.0254,
      "step": 4780
    },
    {
      "epoch": 1.1065,
      "grad_norm": 3.7425663471221924,
      "learning_rate": 2.635858585858586e-05,
      "loss": 0.0721,
      "step": 4781
    },
    {
      "epoch": 1.1066,
      "grad_norm": 7.974576950073242,
      "learning_rate": 2.6353535353535354e-05,
      "loss": 0.5421,
      "step": 4782
    },
    {
      "epoch": 1.1067,
      "grad_norm": 3.1439709663391113,
      "learning_rate": 2.634848484848485e-05,
      "loss": 0.5428,
      "step": 4783
    },
    {
      "epoch": 1.1068,
      "grad_norm": 13.788487434387207,
      "learning_rate": 2.6343434343434342e-05,
      "loss": 0.3944,
      "step": 4784
    },
    {
      "epoch": 1.1069,
      "grad_norm": 3.675260543823242,
      "learning_rate": 2.633838383838384e-05,
      "loss": 0.9772,
      "step": 4785
    },
    {
      "epoch": 1.107,
      "grad_norm": 7.3754777908325195,
      "learning_rate": 2.633333333333333e-05,
      "loss": 0.0672,
      "step": 4786
    },
    {
      "epoch": 1.1071,
      "grad_norm": 5.616894245147705,
      "learning_rate": 2.632828282828283e-05,
      "loss": 0.2961,
      "step": 4787
    },
    {
      "epoch": 1.1072,
      "grad_norm": 6.757253170013428,
      "learning_rate": 2.6323232323232323e-05,
      "loss": 0.7591,
      "step": 4788
    },
    {
      "epoch": 1.1073,
      "grad_norm": 9.770129203796387,
      "learning_rate": 2.631818181818182e-05,
      "loss": 0.434,
      "step": 4789
    },
    {
      "epoch": 1.1074,
      "grad_norm": 7.589291572570801,
      "learning_rate": 2.631313131313131e-05,
      "loss": 0.3926,
      "step": 4790
    },
    {
      "epoch": 1.1075,
      "grad_norm": 3.9038259983062744,
      "learning_rate": 2.630808080808081e-05,
      "loss": 0.5264,
      "step": 4791
    },
    {
      "epoch": 1.1076,
      "grad_norm": 3.9339683055877686,
      "learning_rate": 2.63030303030303e-05,
      "loss": 0.5421,
      "step": 4792
    },
    {
      "epoch": 1.1077,
      "grad_norm": 3.5902292728424072,
      "learning_rate": 2.62979797979798e-05,
      "loss": 0.0769,
      "step": 4793
    },
    {
      "epoch": 1.1078000000000001,
      "grad_norm": 4.059595584869385,
      "learning_rate": 2.6292929292929295e-05,
      "loss": 0.1823,
      "step": 4794
    },
    {
      "epoch": 1.1078999999999999,
      "grad_norm": 5.427414894104004,
      "learning_rate": 2.6287878787878788e-05,
      "loss": 0.983,
      "step": 4795
    },
    {
      "epoch": 1.108,
      "grad_norm": 8.690467834472656,
      "learning_rate": 2.6282828282828287e-05,
      "loss": 0.8679,
      "step": 4796
    },
    {
      "epoch": 1.1081,
      "grad_norm": 13.304617881774902,
      "learning_rate": 2.627777777777778e-05,
      "loss": 0.4261,
      "step": 4797
    },
    {
      "epoch": 1.1082,
      "grad_norm": 20.028169631958008,
      "learning_rate": 2.6272727272727276e-05,
      "loss": 0.4219,
      "step": 4798
    },
    {
      "epoch": 1.1083,
      "grad_norm": 7.9599289894104,
      "learning_rate": 2.626767676767677e-05,
      "loss": 0.4224,
      "step": 4799
    },
    {
      "epoch": 1.1084,
      "grad_norm": 1.9548494815826416,
      "learning_rate": 2.6262626262626268e-05,
      "loss": 0.0745,
      "step": 4800
    },
    {
      "epoch": 1.1085,
      "grad_norm": 1.4542723894119263,
      "learning_rate": 2.6257575757575757e-05,
      "loss": 0.015,
      "step": 4801
    },
    {
      "epoch": 1.1086,
      "grad_norm": 5.77908992767334,
      "learning_rate": 2.6252525252525256e-05,
      "loss": 0.2875,
      "step": 4802
    },
    {
      "epoch": 1.1087,
      "grad_norm": 4.072074890136719,
      "learning_rate": 2.624747474747475e-05,
      "loss": 0.5288,
      "step": 4803
    },
    {
      "epoch": 1.1088,
      "grad_norm": 3.5278916358947754,
      "learning_rate": 2.6242424242424245e-05,
      "loss": 0.9538,
      "step": 4804
    },
    {
      "epoch": 1.1089,
      "grad_norm": 4.977423667907715,
      "learning_rate": 2.6237373737373738e-05,
      "loss": 0.1535,
      "step": 4805
    },
    {
      "epoch": 1.109,
      "grad_norm": 4.529896259307861,
      "learning_rate": 2.6232323232323237e-05,
      "loss": 0.1415,
      "step": 4806
    },
    {
      "epoch": 1.1091,
      "grad_norm": 6.453147888183594,
      "learning_rate": 2.622727272727273e-05,
      "loss": 0.2562,
      "step": 4807
    },
    {
      "epoch": 1.1092,
      "grad_norm": 4.785303592681885,
      "learning_rate": 2.6222222222222226e-05,
      "loss": 0.1309,
      "step": 4808
    },
    {
      "epoch": 1.1093,
      "grad_norm": 4.911216735839844,
      "learning_rate": 2.6217171717171718e-05,
      "loss": 0.0435,
      "step": 4809
    },
    {
      "epoch": 1.1094,
      "grad_norm": 1.6400123834609985,
      "learning_rate": 2.6212121212121214e-05,
      "loss": 0.0372,
      "step": 4810
    },
    {
      "epoch": 1.1095,
      "grad_norm": 4.180517673492432,
      "learning_rate": 2.6207070707070707e-05,
      "loss": 0.2423,
      "step": 4811
    },
    {
      "epoch": 1.1096,
      "grad_norm": 2.952608823776245,
      "learning_rate": 2.6202020202020206e-05,
      "loss": 0.5218,
      "step": 4812
    },
    {
      "epoch": 1.1097,
      "grad_norm": 4.51439094543457,
      "learning_rate": 2.61969696969697e-05,
      "loss": 0.1744,
      "step": 4813
    },
    {
      "epoch": 1.1098,
      "grad_norm": 5.623551845550537,
      "learning_rate": 2.6191919191919195e-05,
      "loss": 0.2514,
      "step": 4814
    },
    {
      "epoch": 1.1099,
      "grad_norm": 5.7642669677734375,
      "learning_rate": 2.6186868686868687e-05,
      "loss": 0.6275,
      "step": 4815
    },
    {
      "epoch": 1.11,
      "grad_norm": 5.5274434089660645,
      "learning_rate": 2.6181818181818187e-05,
      "loss": 0.2315,
      "step": 4816
    },
    {
      "epoch": 1.1101,
      "grad_norm": 13.30423641204834,
      "learning_rate": 2.6176767676767676e-05,
      "loss": 0.2937,
      "step": 4817
    },
    {
      "epoch": 1.1102,
      "grad_norm": 6.604697227478027,
      "learning_rate": 2.6171717171717175e-05,
      "loss": 0.8615,
      "step": 4818
    },
    {
      "epoch": 1.1103,
      "grad_norm": 6.533760070800781,
      "learning_rate": 2.6166666666666668e-05,
      "loss": 0.3217,
      "step": 4819
    },
    {
      "epoch": 1.1104,
      "grad_norm": 6.597803592681885,
      "learning_rate": 2.6161616161616164e-05,
      "loss": 0.1937,
      "step": 4820
    },
    {
      "epoch": 1.1105,
      "grad_norm": 80.58687591552734,
      "learning_rate": 2.6156565656565657e-05,
      "loss": 0.1403,
      "step": 4821
    },
    {
      "epoch": 1.1106,
      "grad_norm": 2.405834913253784,
      "learning_rate": 2.6151515151515156e-05,
      "loss": 0.0832,
      "step": 4822
    },
    {
      "epoch": 1.1107,
      "grad_norm": 7.3836493492126465,
      "learning_rate": 2.6146464646464645e-05,
      "loss": 0.3359,
      "step": 4823
    },
    {
      "epoch": 1.1108,
      "grad_norm": 3.2684197425842285,
      "learning_rate": 2.6141414141414145e-05,
      "loss": 0.198,
      "step": 4824
    },
    {
      "epoch": 1.1109,
      "grad_norm": 8.170228958129883,
      "learning_rate": 2.6136363636363637e-05,
      "loss": 0.8129,
      "step": 4825
    },
    {
      "epoch": 1.111,
      "grad_norm": 6.617372035980225,
      "learning_rate": 2.6131313131313133e-05,
      "loss": 0.3387,
      "step": 4826
    },
    {
      "epoch": 1.1111,
      "grad_norm": 3.957838773727417,
      "learning_rate": 2.6126262626262626e-05,
      "loss": 0.0724,
      "step": 4827
    },
    {
      "epoch": 1.1112,
      "grad_norm": 4.184390068054199,
      "learning_rate": 2.6121212121212125e-05,
      "loss": 0.0967,
      "step": 4828
    },
    {
      "epoch": 1.1113,
      "grad_norm": 15.05860710144043,
      "learning_rate": 2.6116161616161618e-05,
      "loss": 0.7083,
      "step": 4829
    },
    {
      "epoch": 1.1114,
      "grad_norm": 1.1087125539779663,
      "learning_rate": 2.6111111111111114e-05,
      "loss": 0.0127,
      "step": 4830
    },
    {
      "epoch": 1.1115,
      "grad_norm": 7.469234943389893,
      "learning_rate": 2.6106060606060606e-05,
      "loss": 0.1348,
      "step": 4831
    },
    {
      "epoch": 1.1116,
      "grad_norm": 8.293919563293457,
      "learning_rate": 2.6101010101010102e-05,
      "loss": 0.3397,
      "step": 4832
    },
    {
      "epoch": 1.1117,
      "grad_norm": 3.0663836002349854,
      "learning_rate": 2.6095959595959595e-05,
      "loss": 0.1147,
      "step": 4833
    },
    {
      "epoch": 1.1118,
      "grad_norm": 25.67450523376465,
      "learning_rate": 2.6090909090909094e-05,
      "loss": 0.0656,
      "step": 4834
    },
    {
      "epoch": 1.1118999999999999,
      "grad_norm": 6.326839923858643,
      "learning_rate": 2.6085858585858587e-05,
      "loss": 0.3224,
      "step": 4835
    },
    {
      "epoch": 1.112,
      "grad_norm": 5.610405921936035,
      "learning_rate": 2.6080808080808083e-05,
      "loss": 0.1104,
      "step": 4836
    },
    {
      "epoch": 1.1121,
      "grad_norm": 11.232041358947754,
      "learning_rate": 2.6075757575757575e-05,
      "loss": 1.187,
      "step": 4837
    },
    {
      "epoch": 1.1122,
      "grad_norm": 3.0309979915618896,
      "learning_rate": 2.6070707070707075e-05,
      "loss": 0.0409,
      "step": 4838
    },
    {
      "epoch": 1.1123,
      "grad_norm": 13.82136058807373,
      "learning_rate": 2.6065656565656564e-05,
      "loss": 0.3398,
      "step": 4839
    },
    {
      "epoch": 1.1124,
      "grad_norm": 8.583460807800293,
      "learning_rate": 2.6060606060606063e-05,
      "loss": 0.1714,
      "step": 4840
    },
    {
      "epoch": 1.1125,
      "grad_norm": 7.136071681976318,
      "learning_rate": 2.6055555555555556e-05,
      "loss": 0.1006,
      "step": 4841
    },
    {
      "epoch": 1.1126,
      "grad_norm": 10.51154899597168,
      "learning_rate": 2.6050505050505052e-05,
      "loss": 0.3807,
      "step": 4842
    },
    {
      "epoch": 1.1127,
      "grad_norm": 4.129927158355713,
      "learning_rate": 2.6045454545454545e-05,
      "loss": 0.2068,
      "step": 4843
    },
    {
      "epoch": 1.1128,
      "grad_norm": 5.6799774169921875,
      "learning_rate": 2.6040404040404044e-05,
      "loss": 0.2536,
      "step": 4844
    },
    {
      "epoch": 1.1129,
      "grad_norm": 23.56550407409668,
      "learning_rate": 2.6035353535353537e-05,
      "loss": 0.6298,
      "step": 4845
    },
    {
      "epoch": 1.113,
      "grad_norm": 1.3590842485427856,
      "learning_rate": 2.6030303030303033e-05,
      "loss": 0.032,
      "step": 4846
    },
    {
      "epoch": 1.1131,
      "grad_norm": 3.2809948921203613,
      "learning_rate": 2.6025252525252525e-05,
      "loss": 0.0691,
      "step": 4847
    },
    {
      "epoch": 1.1132,
      "grad_norm": 10.237061500549316,
      "learning_rate": 2.602020202020202e-05,
      "loss": 0.362,
      "step": 4848
    },
    {
      "epoch": 1.1133,
      "grad_norm": 5.329871654510498,
      "learning_rate": 2.6015151515151514e-05,
      "loss": 0.4845,
      "step": 4849
    },
    {
      "epoch": 1.1134,
      "grad_norm": 6.846738338470459,
      "learning_rate": 2.6010101010101013e-05,
      "loss": 0.7439,
      "step": 4850
    },
    {
      "epoch": 1.1135,
      "grad_norm": 3.4157395362854004,
      "learning_rate": 2.6005050505050506e-05,
      "loss": 0.5414,
      "step": 4851
    },
    {
      "epoch": 1.1136,
      "grad_norm": 9.709443092346191,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.3022,
      "step": 4852
    },
    {
      "epoch": 1.1137,
      "grad_norm": 8.684234619140625,
      "learning_rate": 2.5994949494949494e-05,
      "loss": 0.9255,
      "step": 4853
    },
    {
      "epoch": 1.1138,
      "grad_norm": 0.7861385941505432,
      "learning_rate": 2.598989898989899e-05,
      "loss": 0.004,
      "step": 4854
    },
    {
      "epoch": 1.1139000000000001,
      "grad_norm": 6.736790657043457,
      "learning_rate": 2.5984848484848483e-05,
      "loss": 0.4334,
      "step": 4855
    },
    {
      "epoch": 1.114,
      "grad_norm": 5.790373802185059,
      "learning_rate": 2.5979797979797982e-05,
      "loss": 0.4273,
      "step": 4856
    },
    {
      "epoch": 1.1141,
      "grad_norm": 6.266722202301025,
      "learning_rate": 2.5974747474747475e-05,
      "loss": 0.597,
      "step": 4857
    },
    {
      "epoch": 1.1142,
      "grad_norm": 3.040034770965576,
      "learning_rate": 2.596969696969697e-05,
      "loss": 0.0584,
      "step": 4858
    },
    {
      "epoch": 1.1143,
      "grad_norm": 4.486883640289307,
      "learning_rate": 2.5964646464646464e-05,
      "loss": 0.2806,
      "step": 4859
    },
    {
      "epoch": 1.1144,
      "grad_norm": 6.032585620880127,
      "learning_rate": 2.5959595959595963e-05,
      "loss": 0.7273,
      "step": 4860
    },
    {
      "epoch": 1.1145,
      "grad_norm": 27.91804313659668,
      "learning_rate": 2.5954545454545452e-05,
      "loss": 0.3758,
      "step": 4861
    },
    {
      "epoch": 1.1146,
      "grad_norm": 78.80984497070312,
      "learning_rate": 2.594949494949495e-05,
      "loss": 0.4291,
      "step": 4862
    },
    {
      "epoch": 1.1147,
      "grad_norm": 2.597867965698242,
      "learning_rate": 2.5944444444444444e-05,
      "loss": 0.4962,
      "step": 4863
    },
    {
      "epoch": 1.1148,
      "grad_norm": 18.38748550415039,
      "learning_rate": 2.593939393939394e-05,
      "loss": 0.3567,
      "step": 4864
    },
    {
      "epoch": 1.1149,
      "grad_norm": 7.5966691970825195,
      "learning_rate": 2.5934343434343433e-05,
      "loss": 0.9976,
      "step": 4865
    },
    {
      "epoch": 1.115,
      "grad_norm": 9.883023262023926,
      "learning_rate": 2.5929292929292932e-05,
      "loss": 1.0375,
      "step": 4866
    },
    {
      "epoch": 1.1151,
      "grad_norm": 6.940661430358887,
      "learning_rate": 2.5924242424242425e-05,
      "loss": 0.1057,
      "step": 4867
    },
    {
      "epoch": 1.1152,
      "grad_norm": 3.0792598724365234,
      "learning_rate": 2.591919191919192e-05,
      "loss": 0.5029,
      "step": 4868
    },
    {
      "epoch": 1.1153,
      "grad_norm": 1.8872848749160767,
      "learning_rate": 2.5914141414141413e-05,
      "loss": 0.0556,
      "step": 4869
    },
    {
      "epoch": 1.1154,
      "grad_norm": 1.1752320528030396,
      "learning_rate": 2.590909090909091e-05,
      "loss": 0.0341,
      "step": 4870
    },
    {
      "epoch": 1.1155,
      "grad_norm": 7.149911880493164,
      "learning_rate": 2.5904040404040402e-05,
      "loss": 0.811,
      "step": 4871
    },
    {
      "epoch": 1.1156,
      "grad_norm": 17.154647827148438,
      "learning_rate": 2.58989898989899e-05,
      "loss": 0.2184,
      "step": 4872
    },
    {
      "epoch": 1.1157,
      "grad_norm": 11.731772422790527,
      "learning_rate": 2.5893939393939394e-05,
      "loss": 0.3362,
      "step": 4873
    },
    {
      "epoch": 1.1158,
      "grad_norm": 2.880307674407959,
      "learning_rate": 2.588888888888889e-05,
      "loss": 0.0507,
      "step": 4874
    },
    {
      "epoch": 1.1159,
      "grad_norm": 8.645869255065918,
      "learning_rate": 2.5883838383838382e-05,
      "loss": 0.0648,
      "step": 4875
    },
    {
      "epoch": 1.116,
      "grad_norm": 1.21147620677948,
      "learning_rate": 2.587878787878788e-05,
      "loss": 0.0116,
      "step": 4876
    },
    {
      "epoch": 1.1161,
      "grad_norm": 3.922422409057617,
      "learning_rate": 2.587373737373737e-05,
      "loss": 0.1388,
      "step": 4877
    },
    {
      "epoch": 1.1162,
      "grad_norm": 14.7849702835083,
      "learning_rate": 2.586868686868687e-05,
      "loss": 0.446,
      "step": 4878
    },
    {
      "epoch": 1.1163,
      "grad_norm": 9.427644729614258,
      "learning_rate": 2.5863636363636363e-05,
      "loss": 0.0953,
      "step": 4879
    },
    {
      "epoch": 1.1164,
      "grad_norm": 1.289626121520996,
      "learning_rate": 2.585858585858586e-05,
      "loss": 0.0153,
      "step": 4880
    },
    {
      "epoch": 1.1165,
      "grad_norm": 6.786375999450684,
      "learning_rate": 2.585353535353536e-05,
      "loss": 0.3607,
      "step": 4881
    },
    {
      "epoch": 1.1166,
      "grad_norm": 4.983189105987549,
      "learning_rate": 2.584848484848485e-05,
      "loss": 0.2616,
      "step": 4882
    },
    {
      "epoch": 1.1167,
      "grad_norm": 5.19439697265625,
      "learning_rate": 2.5843434343434347e-05,
      "loss": 0.2994,
      "step": 4883
    },
    {
      "epoch": 1.1168,
      "grad_norm": 7.1995978355407715,
      "learning_rate": 2.583838383838384e-05,
      "loss": 0.3521,
      "step": 4884
    },
    {
      "epoch": 1.1169,
      "grad_norm": 1.9861009120941162,
      "learning_rate": 2.5833333333333336e-05,
      "loss": 0.0242,
      "step": 4885
    },
    {
      "epoch": 1.117,
      "grad_norm": 3.6369173526763916,
      "learning_rate": 2.5828282828282828e-05,
      "loss": 0.1032,
      "step": 4886
    },
    {
      "epoch": 1.1171,
      "grad_norm": 3.2232959270477295,
      "learning_rate": 2.5823232323232328e-05,
      "loss": 0.1026,
      "step": 4887
    },
    {
      "epoch": 1.1172,
      "grad_norm": 3.8063466548919678,
      "learning_rate": 2.581818181818182e-05,
      "loss": 0.1803,
      "step": 4888
    },
    {
      "epoch": 1.1173,
      "grad_norm": 6.456389427185059,
      "learning_rate": 2.5813131313131316e-05,
      "loss": 0.1653,
      "step": 4889
    },
    {
      "epoch": 1.1174,
      "grad_norm": 4.19140625,
      "learning_rate": 2.580808080808081e-05,
      "loss": 0.0277,
      "step": 4890
    },
    {
      "epoch": 1.1175,
      "grad_norm": 0.5179581046104431,
      "learning_rate": 2.5803030303030308e-05,
      "loss": 0.0245,
      "step": 4891
    },
    {
      "epoch": 1.1176,
      "grad_norm": 3.1179449558258057,
      "learning_rate": 2.5797979797979797e-05,
      "loss": 0.5206,
      "step": 4892
    },
    {
      "epoch": 1.1177,
      "grad_norm": 6.480026721954346,
      "learning_rate": 2.5792929292929297e-05,
      "loss": 0.0329,
      "step": 4893
    },
    {
      "epoch": 1.1178,
      "grad_norm": 3.9032251834869385,
      "learning_rate": 2.578787878787879e-05,
      "loss": 0.1498,
      "step": 4894
    },
    {
      "epoch": 1.1179000000000001,
      "grad_norm": 2.1656672954559326,
      "learning_rate": 2.5782828282828285e-05,
      "loss": 0.0305,
      "step": 4895
    },
    {
      "epoch": 1.1179999999999999,
      "grad_norm": 7.67807674407959,
      "learning_rate": 2.5777777777777778e-05,
      "loss": 0.3815,
      "step": 4896
    },
    {
      "epoch": 1.1181,
      "grad_norm": 0.28315383195877075,
      "learning_rate": 2.5772727272727277e-05,
      "loss": 0.0024,
      "step": 4897
    },
    {
      "epoch": 1.1182,
      "grad_norm": 8.036892890930176,
      "learning_rate": 2.576767676767677e-05,
      "loss": 1.3327,
      "step": 4898
    },
    {
      "epoch": 1.1183,
      "grad_norm": 3.4367799758911133,
      "learning_rate": 2.5762626262626266e-05,
      "loss": 0.2876,
      "step": 4899
    },
    {
      "epoch": 1.1184,
      "grad_norm": 3.144611120223999,
      "learning_rate": 2.575757575757576e-05,
      "loss": 0.2884,
      "step": 4900
    },
    {
      "epoch": 1.1185,
      "grad_norm": 6.751226902008057,
      "learning_rate": 2.5752525252525254e-05,
      "loss": 0.3584,
      "step": 4901
    },
    {
      "epoch": 1.1186,
      "grad_norm": 3.5650811195373535,
      "learning_rate": 2.5747474747474747e-05,
      "loss": 0.1074,
      "step": 4902
    },
    {
      "epoch": 1.1187,
      "grad_norm": 11.509328842163086,
      "learning_rate": 2.5742424242424246e-05,
      "loss": 0.6827,
      "step": 4903
    },
    {
      "epoch": 1.1188,
      "grad_norm": 2.2421247959136963,
      "learning_rate": 2.573737373737374e-05,
      "loss": 0.0632,
      "step": 4904
    },
    {
      "epoch": 1.1189,
      "grad_norm": 3.9575626850128174,
      "learning_rate": 2.5732323232323235e-05,
      "loss": 0.323,
      "step": 4905
    },
    {
      "epoch": 1.119,
      "grad_norm": 2.2622361183166504,
      "learning_rate": 2.5727272727272728e-05,
      "loss": 0.0432,
      "step": 4906
    },
    {
      "epoch": 1.1191,
      "grad_norm": 2.928767204284668,
      "learning_rate": 2.5722222222222224e-05,
      "loss": 0.1235,
      "step": 4907
    },
    {
      "epoch": 1.1192,
      "grad_norm": 2.581092357635498,
      "learning_rate": 2.5717171717171716e-05,
      "loss": 0.0876,
      "step": 4908
    },
    {
      "epoch": 1.1193,
      "grad_norm": 8.120458602905273,
      "learning_rate": 2.5712121212121216e-05,
      "loss": 0.7068,
      "step": 4909
    },
    {
      "epoch": 1.1194,
      "grad_norm": 3.054490804672241,
      "learning_rate": 2.5707070707070708e-05,
      "loss": 0.2465,
      "step": 4910
    },
    {
      "epoch": 1.1195,
      "grad_norm": 2.1114089488983154,
      "learning_rate": 2.5702020202020204e-05,
      "loss": 0.0452,
      "step": 4911
    },
    {
      "epoch": 1.1196,
      "grad_norm": 0.8970494866371155,
      "learning_rate": 2.5696969696969697e-05,
      "loss": 0.4462,
      "step": 4912
    },
    {
      "epoch": 1.1197,
      "grad_norm": 2.2392661571502686,
      "learning_rate": 2.5691919191919196e-05,
      "loss": 0.0496,
      "step": 4913
    },
    {
      "epoch": 1.1198,
      "grad_norm": 3.7471375465393066,
      "learning_rate": 2.5686868686868685e-05,
      "loss": 0.1499,
      "step": 4914
    },
    {
      "epoch": 1.1199,
      "grad_norm": 4.873441696166992,
      "learning_rate": 2.5681818181818185e-05,
      "loss": 0.6988,
      "step": 4915
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5218933820724487,
      "learning_rate": 2.5676767676767677e-05,
      "loss": 0.4387,
      "step": 4916
    },
    {
      "epoch": 1.1201,
      "grad_norm": 4.340986251831055,
      "learning_rate": 2.5671717171717173e-05,
      "loss": 0.1562,
      "step": 4917
    },
    {
      "epoch": 1.1202,
      "grad_norm": 6.8937788009643555,
      "learning_rate": 2.5666666666666666e-05,
      "loss": 0.2523,
      "step": 4918
    },
    {
      "epoch": 1.1203,
      "grad_norm": 3.0403380393981934,
      "learning_rate": 2.5661616161616165e-05,
      "loss": 0.0837,
      "step": 4919
    },
    {
      "epoch": 1.1204,
      "grad_norm": 1.4949485063552856,
      "learning_rate": 2.5656565656565658e-05,
      "loss": 0.0585,
      "step": 4920
    },
    {
      "epoch": 1.1205,
      "grad_norm": 1.857343077659607,
      "learning_rate": 2.5651515151515154e-05,
      "loss": 0.0347,
      "step": 4921
    },
    {
      "epoch": 1.1206,
      "grad_norm": 4.027801990509033,
      "learning_rate": 2.5646464646464647e-05,
      "loss": 0.317,
      "step": 4922
    },
    {
      "epoch": 1.1207,
      "grad_norm": 2.9687421321868896,
      "learning_rate": 2.5641414141414143e-05,
      "loss": 0.0412,
      "step": 4923
    },
    {
      "epoch": 1.1208,
      "grad_norm": 2.5494611263275146,
      "learning_rate": 2.5636363636363635e-05,
      "loss": 0.4693,
      "step": 4924
    },
    {
      "epoch": 1.1209,
      "grad_norm": 6.343966960906982,
      "learning_rate": 2.5631313131313135e-05,
      "loss": 0.1848,
      "step": 4925
    },
    {
      "epoch": 1.121,
      "grad_norm": 9.350842475891113,
      "learning_rate": 2.5626262626262627e-05,
      "loss": 0.8181,
      "step": 4926
    },
    {
      "epoch": 1.1211,
      "grad_norm": 6.8393874168396,
      "learning_rate": 2.5621212121212123e-05,
      "loss": 0.7414,
      "step": 4927
    },
    {
      "epoch": 1.1212,
      "grad_norm": 3.546954393386841,
      "learning_rate": 2.5616161616161616e-05,
      "loss": 0.2759,
      "step": 4928
    },
    {
      "epoch": 1.1213,
      "grad_norm": 7.666682243347168,
      "learning_rate": 2.5611111111111115e-05,
      "loss": 0.3109,
      "step": 4929
    },
    {
      "epoch": 1.1214,
      "grad_norm": 5.086737632751465,
      "learning_rate": 2.5606060606060604e-05,
      "loss": 0.3783,
      "step": 4930
    },
    {
      "epoch": 1.1215,
      "grad_norm": 1.9803892374038696,
      "learning_rate": 2.5601010101010104e-05,
      "loss": 0.1073,
      "step": 4931
    },
    {
      "epoch": 1.1216,
      "grad_norm": 20.457365036010742,
      "learning_rate": 2.5595959595959596e-05,
      "loss": 0.4938,
      "step": 4932
    },
    {
      "epoch": 1.1217,
      "grad_norm": 4.281153678894043,
      "learning_rate": 2.5590909090909092e-05,
      "loss": 0.3412,
      "step": 4933
    },
    {
      "epoch": 1.1218,
      "grad_norm": 2.2190403938293457,
      "learning_rate": 2.5585858585858585e-05,
      "loss": 0.5418,
      "step": 4934
    },
    {
      "epoch": 1.1219,
      "grad_norm": 3.8737800121307373,
      "learning_rate": 2.5580808080808084e-05,
      "loss": 0.6347,
      "step": 4935
    },
    {
      "epoch": 1.1219999999999999,
      "grad_norm": 5.756459712982178,
      "learning_rate": 2.5575757575757573e-05,
      "loss": 0.3688,
      "step": 4936
    },
    {
      "epoch": 1.1221,
      "grad_norm": 8.387085914611816,
      "learning_rate": 2.5570707070707073e-05,
      "loss": 0.3308,
      "step": 4937
    },
    {
      "epoch": 1.1222,
      "grad_norm": 3.46882963180542,
      "learning_rate": 2.5565656565656565e-05,
      "loss": 0.2208,
      "step": 4938
    },
    {
      "epoch": 1.1223,
      "grad_norm": 2.132046937942505,
      "learning_rate": 2.556060606060606e-05,
      "loss": 0.116,
      "step": 4939
    },
    {
      "epoch": 1.1224,
      "grad_norm": 3.9232912063598633,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 0.4713,
      "step": 4940
    },
    {
      "epoch": 1.1225,
      "grad_norm": 4.006885528564453,
      "learning_rate": 2.5550505050505053e-05,
      "loss": 0.6288,
      "step": 4941
    },
    {
      "epoch": 1.1226,
      "grad_norm": 22.058732986450195,
      "learning_rate": 2.5545454545454546e-05,
      "loss": 0.628,
      "step": 4942
    },
    {
      "epoch": 1.1227,
      "grad_norm": 6.4080071449279785,
      "learning_rate": 2.5540404040404042e-05,
      "loss": 0.1624,
      "step": 4943
    },
    {
      "epoch": 1.1228,
      "grad_norm": 3.057769775390625,
      "learning_rate": 2.5535353535353535e-05,
      "loss": 0.1084,
      "step": 4944
    },
    {
      "epoch": 1.1229,
      "grad_norm": 4.3078999519348145,
      "learning_rate": 2.553030303030303e-05,
      "loss": 0.3014,
      "step": 4945
    },
    {
      "epoch": 1.123,
      "grad_norm": 11.249069213867188,
      "learning_rate": 2.5525252525252523e-05,
      "loss": 0.4589,
      "step": 4946
    },
    {
      "epoch": 1.1231,
      "grad_norm": 3.5336227416992188,
      "learning_rate": 2.5520202020202023e-05,
      "loss": 0.1474,
      "step": 4947
    },
    {
      "epoch": 1.1232,
      "grad_norm": 2.556990146636963,
      "learning_rate": 2.5515151515151515e-05,
      "loss": 0.4992,
      "step": 4948
    },
    {
      "epoch": 1.1233,
      "grad_norm": 2.2366690635681152,
      "learning_rate": 2.551010101010101e-05,
      "loss": 0.4534,
      "step": 4949
    },
    {
      "epoch": 1.1234,
      "grad_norm": 11.544108390808105,
      "learning_rate": 2.5505050505050504e-05,
      "loss": 0.3832,
      "step": 4950
    },
    {
      "epoch": 1.1235,
      "grad_norm": 24.855031967163086,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.3832,
      "step": 4951
    },
    {
      "epoch": 1.1236,
      "grad_norm": 32.15934371948242,
      "learning_rate": 2.5494949494949492e-05,
      "loss": 0.4397,
      "step": 4952
    },
    {
      "epoch": 1.1237,
      "grad_norm": 3.459784507751465,
      "learning_rate": 2.5489898989898992e-05,
      "loss": 0.7039,
      "step": 4953
    },
    {
      "epoch": 1.1238,
      "grad_norm": 4.929975986480713,
      "learning_rate": 2.5484848484848484e-05,
      "loss": 0.1943,
      "step": 4954
    },
    {
      "epoch": 1.1239,
      "grad_norm": 2.344959020614624,
      "learning_rate": 2.547979797979798e-05,
      "loss": 0.4702,
      "step": 4955
    },
    {
      "epoch": 1.124,
      "grad_norm": 9.411661148071289,
      "learning_rate": 2.5474747474747473e-05,
      "loss": 0.1312,
      "step": 4956
    },
    {
      "epoch": 1.1241,
      "grad_norm": 4.192747116088867,
      "learning_rate": 2.5469696969696972e-05,
      "loss": 0.1845,
      "step": 4957
    },
    {
      "epoch": 1.1242,
      "grad_norm": 2.1852784156799316,
      "learning_rate": 2.5464646464646465e-05,
      "loss": 0.0588,
      "step": 4958
    },
    {
      "epoch": 1.1243,
      "grad_norm": 46.07902145385742,
      "learning_rate": 2.545959595959596e-05,
      "loss": 0.2124,
      "step": 4959
    },
    {
      "epoch": 1.1244,
      "grad_norm": 2.111412525177002,
      "learning_rate": 2.5454545454545454e-05,
      "loss": 0.4964,
      "step": 4960
    },
    {
      "epoch": 1.1245,
      "grad_norm": 3.24001145362854,
      "learning_rate": 2.544949494949495e-05,
      "loss": 0.0662,
      "step": 4961
    },
    {
      "epoch": 1.1246,
      "grad_norm": 6.520898818969727,
      "learning_rate": 2.5444444444444442e-05,
      "loss": 0.681,
      "step": 4962
    },
    {
      "epoch": 1.1247,
      "grad_norm": 2.28692889213562,
      "learning_rate": 2.543939393939394e-05,
      "loss": 0.2733,
      "step": 4963
    },
    {
      "epoch": 1.1248,
      "grad_norm": 5.756214141845703,
      "learning_rate": 2.5434343434343434e-05,
      "loss": 0.045,
      "step": 4964
    },
    {
      "epoch": 1.1249,
      "grad_norm": 3.3112523555755615,
      "learning_rate": 2.542929292929293e-05,
      "loss": 0.0912,
      "step": 4965
    },
    {
      "epoch": 1.125,
      "grad_norm": 5.8007988929748535,
      "learning_rate": 2.542424242424243e-05,
      "loss": 0.2718,
      "step": 4966
    },
    {
      "epoch": 1.1251,
      "grad_norm": 4.652823448181152,
      "learning_rate": 2.541919191919192e-05,
      "loss": 0.5475,
      "step": 4967
    },
    {
      "epoch": 1.1252,
      "grad_norm": 4.597703456878662,
      "learning_rate": 2.5414141414141418e-05,
      "loss": 0.3075,
      "step": 4968
    },
    {
      "epoch": 1.1253,
      "grad_norm": 13.99496841430664,
      "learning_rate": 2.540909090909091e-05,
      "loss": 0.932,
      "step": 4969
    },
    {
      "epoch": 1.1254,
      "grad_norm": 5.948421478271484,
      "learning_rate": 2.5404040404040407e-05,
      "loss": 0.324,
      "step": 4970
    },
    {
      "epoch": 1.1255,
      "grad_norm": 6.277963638305664,
      "learning_rate": 2.53989898989899e-05,
      "loss": 0.5308,
      "step": 4971
    },
    {
      "epoch": 1.1256,
      "grad_norm": 12.014594078063965,
      "learning_rate": 2.53939393939394e-05,
      "loss": 0.1808,
      "step": 4972
    },
    {
      "epoch": 1.1257,
      "grad_norm": 7.302523136138916,
      "learning_rate": 2.538888888888889e-05,
      "loss": 0.7423,
      "step": 4973
    },
    {
      "epoch": 1.1258,
      "grad_norm": 11.325374603271484,
      "learning_rate": 2.5383838383838387e-05,
      "loss": 0.485,
      "step": 4974
    },
    {
      "epoch": 1.1259000000000001,
      "grad_norm": 4.833463668823242,
      "learning_rate": 2.537878787878788e-05,
      "loss": 0.262,
      "step": 4975
    },
    {
      "epoch": 1.126,
      "grad_norm": 3.2082290649414062,
      "learning_rate": 2.5373737373737376e-05,
      "loss": 0.1366,
      "step": 4976
    },
    {
      "epoch": 1.1261,
      "grad_norm": 2.570699453353882,
      "learning_rate": 2.536868686868687e-05,
      "loss": 0.2009,
      "step": 4977
    },
    {
      "epoch": 1.1262,
      "grad_norm": 3.703838348388672,
      "learning_rate": 2.5363636363636368e-05,
      "loss": 0.26,
      "step": 4978
    },
    {
      "epoch": 1.1263,
      "grad_norm": 15.684362411499023,
      "learning_rate": 2.535858585858586e-05,
      "loss": 0.2603,
      "step": 4979
    },
    {
      "epoch": 1.1264,
      "grad_norm": 7.1570963859558105,
      "learning_rate": 2.5353535353535356e-05,
      "loss": 0.4287,
      "step": 4980
    },
    {
      "epoch": 1.1265,
      "grad_norm": 21.678112030029297,
      "learning_rate": 2.534848484848485e-05,
      "loss": 0.3086,
      "step": 4981
    },
    {
      "epoch": 1.1266,
      "grad_norm": 4.650572299957275,
      "learning_rate": 2.534343434343435e-05,
      "loss": 0.2815,
      "step": 4982
    },
    {
      "epoch": 1.1267,
      "grad_norm": 5.775352954864502,
      "learning_rate": 2.5338383838383838e-05,
      "loss": 0.4828,
      "step": 4983
    },
    {
      "epoch": 1.1268,
      "grad_norm": 10.37144947052002,
      "learning_rate": 2.5333333333333337e-05,
      "loss": 0.5569,
      "step": 4984
    },
    {
      "epoch": 1.1269,
      "grad_norm": 7.749471187591553,
      "learning_rate": 2.532828282828283e-05,
      "loss": 0.2102,
      "step": 4985
    },
    {
      "epoch": 1.127,
      "grad_norm": 7.013164520263672,
      "learning_rate": 2.5323232323232326e-05,
      "loss": 0.4896,
      "step": 4986
    },
    {
      "epoch": 1.1271,
      "grad_norm": 2.0308539867401123,
      "learning_rate": 2.5318181818181818e-05,
      "loss": 0.0891,
      "step": 4987
    },
    {
      "epoch": 1.1272,
      "grad_norm": 5.981311798095703,
      "learning_rate": 2.5313131313131318e-05,
      "loss": 0.1718,
      "step": 4988
    },
    {
      "epoch": 1.1273,
      "grad_norm": 3.379598379135132,
      "learning_rate": 2.530808080808081e-05,
      "loss": 0.0255,
      "step": 4989
    },
    {
      "epoch": 1.1274,
      "grad_norm": 10.206424713134766,
      "learning_rate": 2.5303030303030306e-05,
      "loss": 0.4606,
      "step": 4990
    },
    {
      "epoch": 1.1275,
      "grad_norm": 5.210668563842773,
      "learning_rate": 2.52979797979798e-05,
      "loss": 0.0835,
      "step": 4991
    },
    {
      "epoch": 1.1276,
      "grad_norm": 3.070016860961914,
      "learning_rate": 2.5292929292929295e-05,
      "loss": 0.0552,
      "step": 4992
    },
    {
      "epoch": 1.1277,
      "grad_norm": 3.9337081909179688,
      "learning_rate": 2.5287878787878787e-05,
      "loss": 0.1509,
      "step": 4993
    },
    {
      "epoch": 1.1278,
      "grad_norm": 28.500621795654297,
      "learning_rate": 2.5282828282828287e-05,
      "loss": 0.3701,
      "step": 4994
    },
    {
      "epoch": 1.1279,
      "grad_norm": 8.127056121826172,
      "learning_rate": 2.527777777777778e-05,
      "loss": 0.4091,
      "step": 4995
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 2.7708423137664795,
      "learning_rate": 2.5272727272727275e-05,
      "loss": 0.5563,
      "step": 4996
    },
    {
      "epoch": 1.1280999999999999,
      "grad_norm": 2.903247833251953,
      "learning_rate": 2.5267676767676768e-05,
      "loss": 0.0642,
      "step": 4997
    },
    {
      "epoch": 1.1282,
      "grad_norm": 4.5608673095703125,
      "learning_rate": 2.5262626262626264e-05,
      "loss": 0.1189,
      "step": 4998
    },
    {
      "epoch": 1.1283,
      "grad_norm": 3.982322931289673,
      "learning_rate": 2.5257575757575757e-05,
      "loss": 0.2297,
      "step": 4999
    },
    {
      "epoch": 1.1284,
      "grad_norm": 3.8852264881134033,
      "learning_rate": 2.5252525252525256e-05,
      "loss": 0.0751,
      "step": 5000
    },
    {
      "epoch": 1.1285,
      "grad_norm": 6.85060977935791,
      "learning_rate": 2.524747474747475e-05,
      "loss": 0.2058,
      "step": 5001
    },
    {
      "epoch": 1.1286,
      "grad_norm": 4.887478828430176,
      "learning_rate": 2.5242424242424244e-05,
      "loss": 0.6114,
      "step": 5002
    },
    {
      "epoch": 1.1287,
      "grad_norm": 10.284256935119629,
      "learning_rate": 2.5237373737373737e-05,
      "loss": 0.1954,
      "step": 5003
    },
    {
      "epoch": 1.1288,
      "grad_norm": 2.982182741165161,
      "learning_rate": 2.5232323232323236e-05,
      "loss": 0.0513,
      "step": 5004
    },
    {
      "epoch": 1.1289,
      "grad_norm": 15.975808143615723,
      "learning_rate": 2.5227272727272726e-05,
      "loss": 0.6992,
      "step": 5005
    },
    {
      "epoch": 1.129,
      "grad_norm": 5.659376621246338,
      "learning_rate": 2.5222222222222225e-05,
      "loss": 0.0364,
      "step": 5006
    },
    {
      "epoch": 1.1291,
      "grad_norm": 4.606686592102051,
      "learning_rate": 2.5217171717171718e-05,
      "loss": 0.1064,
      "step": 5007
    },
    {
      "epoch": 1.1292,
      "grad_norm": 7.917055130004883,
      "learning_rate": 2.5212121212121214e-05,
      "loss": 0.2445,
      "step": 5008
    },
    {
      "epoch": 1.1293,
      "grad_norm": 4.255456447601318,
      "learning_rate": 2.5207070707070706e-05,
      "loss": 0.226,
      "step": 5009
    },
    {
      "epoch": 1.1294,
      "grad_norm": 5.0229878425598145,
      "learning_rate": 2.5202020202020206e-05,
      "loss": 0.0509,
      "step": 5010
    },
    {
      "epoch": 1.1295,
      "grad_norm": 4.4284563064575195,
      "learning_rate": 2.5196969696969698e-05,
      "loss": 0.2434,
      "step": 5011
    },
    {
      "epoch": 1.1296,
      "grad_norm": 1.1362754106521606,
      "learning_rate": 2.5191919191919194e-05,
      "loss": 0.0366,
      "step": 5012
    },
    {
      "epoch": 1.1297,
      "grad_norm": 3.865320920944214,
      "learning_rate": 2.5186868686868687e-05,
      "loss": 0.2205,
      "step": 5013
    },
    {
      "epoch": 1.1298,
      "grad_norm": 4.959005832672119,
      "learning_rate": 2.5181818181818183e-05,
      "loss": 0.5335,
      "step": 5014
    },
    {
      "epoch": 1.1299,
      "grad_norm": 6.487668037414551,
      "learning_rate": 2.5176767676767675e-05,
      "loss": 0.3451,
      "step": 5015
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.364931106567383,
      "learning_rate": 2.5171717171717175e-05,
      "loss": 0.142,
      "step": 5016
    },
    {
      "epoch": 1.1301,
      "grad_norm": 3.489450454711914,
      "learning_rate": 2.5166666666666667e-05,
      "loss": 0.0741,
      "step": 5017
    },
    {
      "epoch": 1.1302,
      "grad_norm": 21.393836975097656,
      "learning_rate": 2.5161616161616163e-05,
      "loss": 0.0272,
      "step": 5018
    },
    {
      "epoch": 1.1303,
      "grad_norm": 10.235710144042969,
      "learning_rate": 2.5156565656565656e-05,
      "loss": 0.2212,
      "step": 5019
    },
    {
      "epoch": 1.1304,
      "grad_norm": 2.6421713829040527,
      "learning_rate": 2.5151515151515155e-05,
      "loss": 0.1271,
      "step": 5020
    },
    {
      "epoch": 1.1305,
      "grad_norm": 2.77447247505188,
      "learning_rate": 2.5146464646464645e-05,
      "loss": 0.1074,
      "step": 5021
    },
    {
      "epoch": 1.1306,
      "grad_norm": 4.288763999938965,
      "learning_rate": 2.5141414141414144e-05,
      "loss": 0.2895,
      "step": 5022
    },
    {
      "epoch": 1.1307,
      "grad_norm": 7.74495792388916,
      "learning_rate": 2.5136363636363637e-05,
      "loss": 0.1243,
      "step": 5023
    },
    {
      "epoch": 1.1308,
      "grad_norm": 7.9153056144714355,
      "learning_rate": 2.5131313131313133e-05,
      "loss": 0.1714,
      "step": 5024
    },
    {
      "epoch": 1.1309,
      "grad_norm": 3.358354091644287,
      "learning_rate": 2.5126262626262625e-05,
      "loss": 0.1259,
      "step": 5025
    },
    {
      "epoch": 1.131,
      "grad_norm": 3.8471410274505615,
      "learning_rate": 2.5121212121212125e-05,
      "loss": 0.3749,
      "step": 5026
    },
    {
      "epoch": 1.1311,
      "grad_norm": 1.4472359418869019,
      "learning_rate": 2.5116161616161614e-05,
      "loss": 0.0402,
      "step": 5027
    },
    {
      "epoch": 1.1312,
      "grad_norm": 5.275332927703857,
      "learning_rate": 2.5111111111111113e-05,
      "loss": 0.1741,
      "step": 5028
    },
    {
      "epoch": 1.1313,
      "grad_norm": 4.901299953460693,
      "learning_rate": 2.5106060606060606e-05,
      "loss": 0.1937,
      "step": 5029
    },
    {
      "epoch": 1.1314,
      "grad_norm": 4.704351902008057,
      "learning_rate": 2.5101010101010102e-05,
      "loss": 0.1319,
      "step": 5030
    },
    {
      "epoch": 1.1315,
      "grad_norm": 8.742284774780273,
      "learning_rate": 2.5095959595959594e-05,
      "loss": 0.6408,
      "step": 5031
    },
    {
      "epoch": 1.1316,
      "grad_norm": 13.831192016601562,
      "learning_rate": 2.5090909090909094e-05,
      "loss": 0.4384,
      "step": 5032
    },
    {
      "epoch": 1.1317,
      "grad_norm": 3.149827003479004,
      "learning_rate": 2.5085858585858586e-05,
      "loss": 0.0994,
      "step": 5033
    },
    {
      "epoch": 1.1318,
      "grad_norm": 3.236286163330078,
      "learning_rate": 2.5080808080808082e-05,
      "loss": 0.6166,
      "step": 5034
    },
    {
      "epoch": 1.1319,
      "grad_norm": 20.231304168701172,
      "learning_rate": 2.5075757575757575e-05,
      "loss": 0.1269,
      "step": 5035
    },
    {
      "epoch": 1.1320000000000001,
      "grad_norm": 3.517427444458008,
      "learning_rate": 2.507070707070707e-05,
      "loss": 0.1419,
      "step": 5036
    },
    {
      "epoch": 1.1320999999999999,
      "grad_norm": 4.259015083312988,
      "learning_rate": 2.5065656565656563e-05,
      "loss": 0.098,
      "step": 5037
    },
    {
      "epoch": 1.1322,
      "grad_norm": 2.0287158489227295,
      "learning_rate": 2.5060606060606063e-05,
      "loss": 0.2161,
      "step": 5038
    },
    {
      "epoch": 1.1323,
      "grad_norm": 2.66994309425354,
      "learning_rate": 2.5055555555555555e-05,
      "loss": 0.2354,
      "step": 5039
    },
    {
      "epoch": 1.1324,
      "grad_norm": 1.658425211906433,
      "learning_rate": 2.505050505050505e-05,
      "loss": 0.0325,
      "step": 5040
    },
    {
      "epoch": 1.1325,
      "grad_norm": 6.73345947265625,
      "learning_rate": 2.5045454545454544e-05,
      "loss": 0.2063,
      "step": 5041
    },
    {
      "epoch": 1.1326,
      "grad_norm": 6.834265232086182,
      "learning_rate": 2.5040404040404043e-05,
      "loss": 0.2126,
      "step": 5042
    },
    {
      "epoch": 1.1327,
      "grad_norm": 3.9969441890716553,
      "learning_rate": 2.5035353535353533e-05,
      "loss": 0.1599,
      "step": 5043
    },
    {
      "epoch": 1.1328,
      "grad_norm": 9.276053428649902,
      "learning_rate": 2.5030303030303032e-05,
      "loss": 0.4916,
      "step": 5044
    },
    {
      "epoch": 1.1329,
      "grad_norm": 10.434085845947266,
      "learning_rate": 2.5025252525252525e-05,
      "loss": 0.2796,
      "step": 5045
    },
    {
      "epoch": 1.133,
      "grad_norm": 14.766300201416016,
      "learning_rate": 2.502020202020202e-05,
      "loss": 0.1072,
      "step": 5046
    },
    {
      "epoch": 1.1331,
      "grad_norm": 3.844647169113159,
      "learning_rate": 2.5015151515151513e-05,
      "loss": 0.627,
      "step": 5047
    },
    {
      "epoch": 1.1332,
      "grad_norm": 5.9739251136779785,
      "learning_rate": 2.5010101010101013e-05,
      "loss": 0.3337,
      "step": 5048
    },
    {
      "epoch": 1.1333,
      "grad_norm": 2.9449262619018555,
      "learning_rate": 2.5005050505050502e-05,
      "loss": 0.5113,
      "step": 5049
    },
    {
      "epoch": 1.1334,
      "grad_norm": 7.065964698791504,
      "learning_rate": 2.5e-05,
      "loss": 0.5388,
      "step": 5050
    },
    {
      "epoch": 1.1335,
      "grad_norm": 3.935030221939087,
      "learning_rate": 2.4994949494949497e-05,
      "loss": 0.3085,
      "step": 5051
    },
    {
      "epoch": 1.1336,
      "grad_norm": 4.480363368988037,
      "learning_rate": 2.498989898989899e-05,
      "loss": 0.2086,
      "step": 5052
    },
    {
      "epoch": 1.1337,
      "grad_norm": 3.020570755004883,
      "learning_rate": 2.4984848484848486e-05,
      "loss": 0.2441,
      "step": 5053
    },
    {
      "epoch": 1.1338,
      "grad_norm": 7.114346981048584,
      "learning_rate": 2.4979797979797982e-05,
      "loss": 0.2064,
      "step": 5054
    },
    {
      "epoch": 1.1339,
      "grad_norm": 7.7801513671875,
      "learning_rate": 2.4974747474747474e-05,
      "loss": 0.2786,
      "step": 5055
    },
    {
      "epoch": 1.134,
      "grad_norm": 5.9330220222473145,
      "learning_rate": 2.496969696969697e-05,
      "loss": 0.1333,
      "step": 5056
    },
    {
      "epoch": 1.1341,
      "grad_norm": 4.935941219329834,
      "learning_rate": 2.4964646464646466e-05,
      "loss": 0.6822,
      "step": 5057
    },
    {
      "epoch": 1.1342,
      "grad_norm": 5.406343936920166,
      "learning_rate": 2.495959595959596e-05,
      "loss": 0.4057,
      "step": 5058
    },
    {
      "epoch": 1.1343,
      "grad_norm": 7.261625289916992,
      "learning_rate": 2.4954545454545455e-05,
      "loss": 0.5239,
      "step": 5059
    },
    {
      "epoch": 1.1344,
      "grad_norm": 7.582565784454346,
      "learning_rate": 2.494949494949495e-05,
      "loss": 0.4216,
      "step": 5060
    },
    {
      "epoch": 1.1345,
      "grad_norm": 0.5711351633071899,
      "learning_rate": 2.4944444444444447e-05,
      "loss": 0.0062,
      "step": 5061
    },
    {
      "epoch": 1.1346,
      "grad_norm": 1.7158342599868774,
      "learning_rate": 2.493939393939394e-05,
      "loss": 0.048,
      "step": 5062
    },
    {
      "epoch": 1.1347,
      "grad_norm": 1.479666829109192,
      "learning_rate": 2.4934343434343436e-05,
      "loss": 0.0405,
      "step": 5063
    },
    {
      "epoch": 1.1348,
      "grad_norm": 2.3515005111694336,
      "learning_rate": 2.492929292929293e-05,
      "loss": 0.0681,
      "step": 5064
    },
    {
      "epoch": 1.1349,
      "grad_norm": 6.83437967300415,
      "learning_rate": 2.4924242424242424e-05,
      "loss": 0.1926,
      "step": 5065
    },
    {
      "epoch": 1.135,
      "grad_norm": 2.4407105445861816,
      "learning_rate": 2.491919191919192e-05,
      "loss": 0.0418,
      "step": 5066
    },
    {
      "epoch": 1.1351,
      "grad_norm": 5.778260707855225,
      "learning_rate": 2.4914141414141416e-05,
      "loss": 0.2805,
      "step": 5067
    },
    {
      "epoch": 1.1352,
      "grad_norm": 2.6237294673919678,
      "learning_rate": 2.490909090909091e-05,
      "loss": 0.0991,
      "step": 5068
    },
    {
      "epoch": 1.1353,
      "grad_norm": 6.486207962036133,
      "learning_rate": 2.4904040404040405e-05,
      "loss": 0.1568,
      "step": 5069
    },
    {
      "epoch": 1.1354,
      "grad_norm": 10.345623970031738,
      "learning_rate": 2.48989898989899e-05,
      "loss": 0.0744,
      "step": 5070
    },
    {
      "epoch": 1.1355,
      "grad_norm": 4.065065383911133,
      "learning_rate": 2.4893939393939393e-05,
      "loss": 0.0503,
      "step": 5071
    },
    {
      "epoch": 1.1356,
      "grad_norm": 9.120564460754395,
      "learning_rate": 2.488888888888889e-05,
      "loss": 0.2587,
      "step": 5072
    },
    {
      "epoch": 1.1357,
      "grad_norm": 1.6657394170761108,
      "learning_rate": 2.4883838383838385e-05,
      "loss": 0.0157,
      "step": 5073
    },
    {
      "epoch": 1.1358,
      "grad_norm": 5.452230930328369,
      "learning_rate": 2.4878787878787878e-05,
      "loss": 0.1979,
      "step": 5074
    },
    {
      "epoch": 1.1359,
      "grad_norm": 6.605992317199707,
      "learning_rate": 2.4873737373737374e-05,
      "loss": 0.1596,
      "step": 5075
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.7749288082122803,
      "learning_rate": 2.486868686868687e-05,
      "loss": 0.0068,
      "step": 5076
    },
    {
      "epoch": 1.1360999999999999,
      "grad_norm": 5.173564910888672,
      "learning_rate": 2.4863636363636362e-05,
      "loss": 0.0988,
      "step": 5077
    },
    {
      "epoch": 1.1362,
      "grad_norm": 6.1824445724487305,
      "learning_rate": 2.485858585858586e-05,
      "loss": 0.1905,
      "step": 5078
    },
    {
      "epoch": 1.1363,
      "grad_norm": 8.363350868225098,
      "learning_rate": 2.4853535353535354e-05,
      "loss": 0.3261,
      "step": 5079
    },
    {
      "epoch": 1.1364,
      "grad_norm": 3.0680811405181885,
      "learning_rate": 2.4848484848484847e-05,
      "loss": 0.1172,
      "step": 5080
    },
    {
      "epoch": 1.1365,
      "grad_norm": 4.481549263000488,
      "learning_rate": 2.4843434343434343e-05,
      "loss": 0.0863,
      "step": 5081
    },
    {
      "epoch": 1.1366,
      "grad_norm": 2.6992225646972656,
      "learning_rate": 2.483838383838384e-05,
      "loss": 0.0736,
      "step": 5082
    },
    {
      "epoch": 1.1367,
      "grad_norm": 4.859525203704834,
      "learning_rate": 2.4833333333333335e-05,
      "loss": 0.6619,
      "step": 5083
    },
    {
      "epoch": 1.1368,
      "grad_norm": 1.087936282157898,
      "learning_rate": 2.4828282828282828e-05,
      "loss": 0.0449,
      "step": 5084
    },
    {
      "epoch": 1.1369,
      "grad_norm": 6.408845901489258,
      "learning_rate": 2.4823232323232324e-05,
      "loss": 0.7218,
      "step": 5085
    },
    {
      "epoch": 1.137,
      "grad_norm": 2.5762789249420166,
      "learning_rate": 2.481818181818182e-05,
      "loss": 0.0491,
      "step": 5086
    },
    {
      "epoch": 1.1371,
      "grad_norm": 3.575564384460449,
      "learning_rate": 2.4813131313131312e-05,
      "loss": 0.2813,
      "step": 5087
    },
    {
      "epoch": 1.1372,
      "grad_norm": 6.2168498039245605,
      "learning_rate": 2.4808080808080808e-05,
      "loss": 0.2394,
      "step": 5088
    },
    {
      "epoch": 1.1373,
      "grad_norm": 30.44783592224121,
      "learning_rate": 2.4803030303030304e-05,
      "loss": 1.0284,
      "step": 5089
    },
    {
      "epoch": 1.1374,
      "grad_norm": 9.128175735473633,
      "learning_rate": 2.4797979797979797e-05,
      "loss": 0.3832,
      "step": 5090
    },
    {
      "epoch": 1.1375,
      "grad_norm": 9.312085151672363,
      "learning_rate": 2.4792929292929293e-05,
      "loss": 0.1355,
      "step": 5091
    },
    {
      "epoch": 1.1376,
      "grad_norm": 1.0890462398529053,
      "learning_rate": 2.478787878787879e-05,
      "loss": 0.019,
      "step": 5092
    },
    {
      "epoch": 1.1377,
      "grad_norm": 5.015799045562744,
      "learning_rate": 2.4782828282828285e-05,
      "loss": 0.2364,
      "step": 5093
    },
    {
      "epoch": 1.1378,
      "grad_norm": 8.689985275268555,
      "learning_rate": 2.477777777777778e-05,
      "loss": 0.1592,
      "step": 5094
    },
    {
      "epoch": 1.1379,
      "grad_norm": 4.956414222717285,
      "learning_rate": 2.4772727272727277e-05,
      "loss": 0.1565,
      "step": 5095
    },
    {
      "epoch": 1.138,
      "grad_norm": 2.3741660118103027,
      "learning_rate": 2.476767676767677e-05,
      "loss": 0.101,
      "step": 5096
    },
    {
      "epoch": 1.1381000000000001,
      "grad_norm": 24.736343383789062,
      "learning_rate": 2.4762626262626265e-05,
      "loss": 0.3229,
      "step": 5097
    },
    {
      "epoch": 1.1381999999999999,
      "grad_norm": 4.246650218963623,
      "learning_rate": 2.475757575757576e-05,
      "loss": 0.5642,
      "step": 5098
    },
    {
      "epoch": 1.1383,
      "grad_norm": 3.767301082611084,
      "learning_rate": 2.4752525252525254e-05,
      "loss": 0.0848,
      "step": 5099
    },
    {
      "epoch": 1.1384,
      "grad_norm": 8.241412162780762,
      "learning_rate": 2.474747474747475e-05,
      "loss": 0.3665,
      "step": 5100
    },
    {
      "epoch": 1.1385,
      "grad_norm": 12.84190559387207,
      "learning_rate": 2.4742424242424246e-05,
      "loss": 1.0765,
      "step": 5101
    },
    {
      "epoch": 1.1386,
      "grad_norm": 5.237433910369873,
      "learning_rate": 2.473737373737374e-05,
      "loss": 0.2508,
      "step": 5102
    },
    {
      "epoch": 1.1387,
      "grad_norm": 16.881364822387695,
      "learning_rate": 2.4732323232323234e-05,
      "loss": 0.8466,
      "step": 5103
    },
    {
      "epoch": 1.1388,
      "grad_norm": 8.21771240234375,
      "learning_rate": 2.472727272727273e-05,
      "loss": 0.1341,
      "step": 5104
    },
    {
      "epoch": 1.1389,
      "grad_norm": 2.7339038848876953,
      "learning_rate": 2.4722222222222223e-05,
      "loss": 0.0756,
      "step": 5105
    },
    {
      "epoch": 1.139,
      "grad_norm": 6.809495449066162,
      "learning_rate": 2.471717171717172e-05,
      "loss": 0.2029,
      "step": 5106
    },
    {
      "epoch": 1.1391,
      "grad_norm": 4.392782211303711,
      "learning_rate": 2.4712121212121215e-05,
      "loss": 0.1117,
      "step": 5107
    },
    {
      "epoch": 1.1392,
      "grad_norm": 5.213833808898926,
      "learning_rate": 2.4707070707070708e-05,
      "loss": 0.5574,
      "step": 5108
    },
    {
      "epoch": 1.1393,
      "grad_norm": 6.049360275268555,
      "learning_rate": 2.4702020202020204e-05,
      "loss": 0.3562,
      "step": 5109
    },
    {
      "epoch": 1.1394,
      "grad_norm": 7.898448467254639,
      "learning_rate": 2.46969696969697e-05,
      "loss": 0.3169,
      "step": 5110
    },
    {
      "epoch": 1.1395,
      "grad_norm": 9.742411613464355,
      "learning_rate": 2.4691919191919192e-05,
      "loss": 1.1158,
      "step": 5111
    },
    {
      "epoch": 1.1396,
      "grad_norm": 8.04572868347168,
      "learning_rate": 2.4686868686868688e-05,
      "loss": 0.317,
      "step": 5112
    },
    {
      "epoch": 1.1397,
      "grad_norm": 6.730393886566162,
      "learning_rate": 2.4681818181818184e-05,
      "loss": 0.7049,
      "step": 5113
    },
    {
      "epoch": 1.1398,
      "grad_norm": 9.147459983825684,
      "learning_rate": 2.467676767676768e-05,
      "loss": 0.5323,
      "step": 5114
    },
    {
      "epoch": 1.1399,
      "grad_norm": 7.00915002822876,
      "learning_rate": 2.4671717171717173e-05,
      "loss": 0.2899,
      "step": 5115
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 7.0939621925354,
      "learning_rate": 2.466666666666667e-05,
      "loss": 0.3898,
      "step": 5116
    },
    {
      "epoch": 1.1401,
      "grad_norm": 4.501914978027344,
      "learning_rate": 2.4661616161616165e-05,
      "loss": 0.177,
      "step": 5117
    },
    {
      "epoch": 1.1402,
      "grad_norm": 1.7601025104522705,
      "learning_rate": 2.4656565656565657e-05,
      "loss": 0.0594,
      "step": 5118
    },
    {
      "epoch": 1.1403,
      "grad_norm": 1.6716821193695068,
      "learning_rate": 2.4651515151515153e-05,
      "loss": 0.076,
      "step": 5119
    },
    {
      "epoch": 1.1404,
      "grad_norm": 11.822834968566895,
      "learning_rate": 2.464646464646465e-05,
      "loss": 0.1906,
      "step": 5120
    },
    {
      "epoch": 1.1405,
      "grad_norm": 0.6237279772758484,
      "learning_rate": 2.4641414141414142e-05,
      "loss": 0.4353,
      "step": 5121
    },
    {
      "epoch": 1.1406,
      "grad_norm": 2.572549343109131,
      "learning_rate": 2.4636363636363638e-05,
      "loss": 0.2298,
      "step": 5122
    },
    {
      "epoch": 1.1407,
      "grad_norm": 7.295194625854492,
      "learning_rate": 2.4631313131313134e-05,
      "loss": 0.4108,
      "step": 5123
    },
    {
      "epoch": 1.1408,
      "grad_norm": 3.6432459354400635,
      "learning_rate": 2.4626262626262627e-05,
      "loss": 0.0377,
      "step": 5124
    },
    {
      "epoch": 1.1409,
      "grad_norm": 8.90928840637207,
      "learning_rate": 2.4621212121212123e-05,
      "loss": 0.6109,
      "step": 5125
    },
    {
      "epoch": 1.141,
      "grad_norm": 4.57892370223999,
      "learning_rate": 2.461616161616162e-05,
      "loss": 0.0797,
      "step": 5126
    },
    {
      "epoch": 1.1411,
      "grad_norm": 3.5240414142608643,
      "learning_rate": 2.461111111111111e-05,
      "loss": 0.6616,
      "step": 5127
    },
    {
      "epoch": 1.1412,
      "grad_norm": 1.9887019395828247,
      "learning_rate": 2.4606060606060607e-05,
      "loss": 0.0172,
      "step": 5128
    },
    {
      "epoch": 1.1413,
      "grad_norm": 2.979999542236328,
      "learning_rate": 2.4601010101010103e-05,
      "loss": 0.6316,
      "step": 5129
    },
    {
      "epoch": 1.1414,
      "grad_norm": 6.017578601837158,
      "learning_rate": 2.4595959595959596e-05,
      "loss": 0.2269,
      "step": 5130
    },
    {
      "epoch": 1.1415,
      "grad_norm": 6.946823596954346,
      "learning_rate": 2.4590909090909092e-05,
      "loss": 0.3478,
      "step": 5131
    },
    {
      "epoch": 1.1416,
      "grad_norm": 5.1888508796691895,
      "learning_rate": 2.4585858585858588e-05,
      "loss": 0.1332,
      "step": 5132
    },
    {
      "epoch": 1.1417,
      "grad_norm": 2.5507843494415283,
      "learning_rate": 2.4580808080808084e-05,
      "loss": 0.2313,
      "step": 5133
    },
    {
      "epoch": 1.1418,
      "grad_norm": 7.404869079589844,
      "learning_rate": 2.4575757575757576e-05,
      "loss": 1.2686,
      "step": 5134
    },
    {
      "epoch": 1.1419,
      "grad_norm": 4.202610492706299,
      "learning_rate": 2.4570707070707072e-05,
      "loss": 0.3064,
      "step": 5135
    },
    {
      "epoch": 1.142,
      "grad_norm": 2.384225845336914,
      "learning_rate": 2.4565656565656568e-05,
      "loss": 0.0588,
      "step": 5136
    },
    {
      "epoch": 1.1421000000000001,
      "grad_norm": 12.26535415649414,
      "learning_rate": 2.456060606060606e-05,
      "loss": 0.4602,
      "step": 5137
    },
    {
      "epoch": 1.1421999999999999,
      "grad_norm": 6.928979396820068,
      "learning_rate": 2.4555555555555557e-05,
      "loss": 0.1841,
      "step": 5138
    },
    {
      "epoch": 1.1423,
      "grad_norm": 8.87103271484375,
      "learning_rate": 2.4550505050505053e-05,
      "loss": 0.1998,
      "step": 5139
    },
    {
      "epoch": 1.1424,
      "grad_norm": 8.2178316116333,
      "learning_rate": 2.4545454545454545e-05,
      "loss": 0.2214,
      "step": 5140
    },
    {
      "epoch": 1.1425,
      "grad_norm": 3.187429666519165,
      "learning_rate": 2.454040404040404e-05,
      "loss": 0.7625,
      "step": 5141
    },
    {
      "epoch": 1.1426,
      "grad_norm": 18.774372100830078,
      "learning_rate": 2.4535353535353537e-05,
      "loss": 0.3074,
      "step": 5142
    },
    {
      "epoch": 1.1427,
      "grad_norm": 11.63023567199707,
      "learning_rate": 2.453030303030303e-05,
      "loss": 0.6705,
      "step": 5143
    },
    {
      "epoch": 1.1428,
      "grad_norm": 7.292832851409912,
      "learning_rate": 2.4525252525252526e-05,
      "loss": 0.3201,
      "step": 5144
    },
    {
      "epoch": 1.1429,
      "grad_norm": 6.725473880767822,
      "learning_rate": 2.4520202020202022e-05,
      "loss": 0.3749,
      "step": 5145
    },
    {
      "epoch": 1.143,
      "grad_norm": 4.819268226623535,
      "learning_rate": 2.4515151515151515e-05,
      "loss": 0.3963,
      "step": 5146
    },
    {
      "epoch": 1.1431,
      "grad_norm": 4.708950996398926,
      "learning_rate": 2.451010101010101e-05,
      "loss": 0.0457,
      "step": 5147
    },
    {
      "epoch": 1.1432,
      "grad_norm": 6.024604797363281,
      "learning_rate": 2.4505050505050507e-05,
      "loss": 0.1972,
      "step": 5148
    },
    {
      "epoch": 1.1433,
      "grad_norm": 5.073880195617676,
      "learning_rate": 2.45e-05,
      "loss": 0.137,
      "step": 5149
    },
    {
      "epoch": 1.1434,
      "grad_norm": 3.590162992477417,
      "learning_rate": 2.4494949494949495e-05,
      "loss": 0.0339,
      "step": 5150
    },
    {
      "epoch": 1.1435,
      "grad_norm": 7.886445045471191,
      "learning_rate": 2.448989898989899e-05,
      "loss": 0.3665,
      "step": 5151
    },
    {
      "epoch": 1.1436,
      "grad_norm": 10.706446647644043,
      "learning_rate": 2.4484848484848484e-05,
      "loss": 0.2093,
      "step": 5152
    },
    {
      "epoch": 1.1437,
      "grad_norm": 9.439184188842773,
      "learning_rate": 2.447979797979798e-05,
      "loss": 0.1261,
      "step": 5153
    },
    {
      "epoch": 1.1438,
      "grad_norm": 2.0380170345306396,
      "learning_rate": 2.4474747474747476e-05,
      "loss": 0.0355,
      "step": 5154
    },
    {
      "epoch": 1.1439,
      "grad_norm": 8.679521560668945,
      "learning_rate": 2.4469696969696972e-05,
      "loss": 0.2716,
      "step": 5155
    },
    {
      "epoch": 1.144,
      "grad_norm": 1.8479934930801392,
      "learning_rate": 2.4464646464646464e-05,
      "loss": 0.0325,
      "step": 5156
    },
    {
      "epoch": 1.1441,
      "grad_norm": 18.066265106201172,
      "learning_rate": 2.445959595959596e-05,
      "loss": 0.7386,
      "step": 5157
    },
    {
      "epoch": 1.1442,
      "grad_norm": 1.9490203857421875,
      "learning_rate": 2.4454545454545456e-05,
      "loss": 0.022,
      "step": 5158
    },
    {
      "epoch": 1.1443,
      "grad_norm": 2.3748996257781982,
      "learning_rate": 2.444949494949495e-05,
      "loss": 0.1146,
      "step": 5159
    },
    {
      "epoch": 1.1444,
      "grad_norm": 1.6488122940063477,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 0.4852,
      "step": 5160
    },
    {
      "epoch": 1.1445,
      "grad_norm": 5.698278427124023,
      "learning_rate": 2.443939393939394e-05,
      "loss": 0.3732,
      "step": 5161
    },
    {
      "epoch": 1.1446,
      "grad_norm": 2.2053136825561523,
      "learning_rate": 2.4434343434343434e-05,
      "loss": 0.0857,
      "step": 5162
    },
    {
      "epoch": 1.1447,
      "grad_norm": 5.791157245635986,
      "learning_rate": 2.442929292929293e-05,
      "loss": 0.382,
      "step": 5163
    },
    {
      "epoch": 1.1448,
      "grad_norm": 6.724499702453613,
      "learning_rate": 2.4424242424242426e-05,
      "loss": 0.1997,
      "step": 5164
    },
    {
      "epoch": 1.1449,
      "grad_norm": 7.312654972076416,
      "learning_rate": 2.4419191919191918e-05,
      "loss": 0.389,
      "step": 5165
    },
    {
      "epoch": 1.145,
      "grad_norm": 3.3563106060028076,
      "learning_rate": 2.4414141414141414e-05,
      "loss": 0.0967,
      "step": 5166
    },
    {
      "epoch": 1.1451,
      "grad_norm": 11.796554565429688,
      "learning_rate": 2.440909090909091e-05,
      "loss": 0.1822,
      "step": 5167
    },
    {
      "epoch": 1.1452,
      "grad_norm": 26.823192596435547,
      "learning_rate": 2.4404040404040403e-05,
      "loss": 0.5217,
      "step": 5168
    },
    {
      "epoch": 1.1453,
      "grad_norm": 55.24634552001953,
      "learning_rate": 2.43989898989899e-05,
      "loss": 0.3949,
      "step": 5169
    },
    {
      "epoch": 1.1454,
      "grad_norm": 1.7380661964416504,
      "learning_rate": 2.4393939393939395e-05,
      "loss": 0.0285,
      "step": 5170
    },
    {
      "epoch": 1.1455,
      "grad_norm": 4.179076671600342,
      "learning_rate": 2.4388888888888887e-05,
      "loss": 0.2005,
      "step": 5171
    },
    {
      "epoch": 1.1456,
      "grad_norm": 47.27809143066406,
      "learning_rate": 2.4383838383838383e-05,
      "loss": 0.5704,
      "step": 5172
    },
    {
      "epoch": 1.1457,
      "grad_norm": 2.5830516815185547,
      "learning_rate": 2.437878787878788e-05,
      "loss": 0.0111,
      "step": 5173
    },
    {
      "epoch": 1.1458,
      "grad_norm": 10.173704147338867,
      "learning_rate": 2.4373737373737375e-05,
      "loss": 0.4034,
      "step": 5174
    },
    {
      "epoch": 1.1459,
      "grad_norm": 6.81486701965332,
      "learning_rate": 2.4368686868686868e-05,
      "loss": 0.6153,
      "step": 5175
    },
    {
      "epoch": 1.146,
      "grad_norm": 1.9611924886703491,
      "learning_rate": 2.4363636363636364e-05,
      "loss": 0.469,
      "step": 5176
    },
    {
      "epoch": 1.1461000000000001,
      "grad_norm": 7.785665512084961,
      "learning_rate": 2.435858585858586e-05,
      "loss": 0.356,
      "step": 5177
    },
    {
      "epoch": 1.1461999999999999,
      "grad_norm": 5.901285171508789,
      "learning_rate": 2.4353535353535352e-05,
      "loss": 0.5724,
      "step": 5178
    },
    {
      "epoch": 1.1463,
      "grad_norm": 0.90590500831604,
      "learning_rate": 2.4348484848484852e-05,
      "loss": 0.0127,
      "step": 5179
    },
    {
      "epoch": 1.1464,
      "grad_norm": 1.1326513290405273,
      "learning_rate": 2.4343434343434344e-05,
      "loss": 0.0072,
      "step": 5180
    },
    {
      "epoch": 1.1465,
      "grad_norm": 4.550037860870361,
      "learning_rate": 2.433838383838384e-05,
      "loss": 0.1378,
      "step": 5181
    },
    {
      "epoch": 1.1466,
      "grad_norm": 1.7685221433639526,
      "learning_rate": 2.4333333333333336e-05,
      "loss": 0.0223,
      "step": 5182
    },
    {
      "epoch": 1.1467,
      "grad_norm": 10.933194160461426,
      "learning_rate": 2.432828282828283e-05,
      "loss": 0.3575,
      "step": 5183
    },
    {
      "epoch": 1.1468,
      "grad_norm": 1.9419296979904175,
      "learning_rate": 2.4323232323232325e-05,
      "loss": 0.0709,
      "step": 5184
    },
    {
      "epoch": 1.1469,
      "grad_norm": 6.7305989265441895,
      "learning_rate": 2.431818181818182e-05,
      "loss": 0.3337,
      "step": 5185
    },
    {
      "epoch": 1.147,
      "grad_norm": 3.120828866958618,
      "learning_rate": 2.4313131313131317e-05,
      "loss": 0.556,
      "step": 5186
    },
    {
      "epoch": 1.1471,
      "grad_norm": 8.667496681213379,
      "learning_rate": 2.430808080808081e-05,
      "loss": 0.9295,
      "step": 5187
    },
    {
      "epoch": 1.1472,
      "grad_norm": 10.851874351501465,
      "learning_rate": 2.4303030303030306e-05,
      "loss": 0.7186,
      "step": 5188
    },
    {
      "epoch": 1.1473,
      "grad_norm": 2.9446308612823486,
      "learning_rate": 2.42979797979798e-05,
      "loss": 0.0891,
      "step": 5189
    },
    {
      "epoch": 1.1474,
      "grad_norm": 5.9243292808532715,
      "learning_rate": 2.4292929292929294e-05,
      "loss": 0.2945,
      "step": 5190
    },
    {
      "epoch": 1.1475,
      "grad_norm": 3.1467697620391846,
      "learning_rate": 2.428787878787879e-05,
      "loss": 0.2508,
      "step": 5191
    },
    {
      "epoch": 1.1476,
      "grad_norm": 5.0739970207214355,
      "learning_rate": 2.4282828282828286e-05,
      "loss": 0.6468,
      "step": 5192
    },
    {
      "epoch": 1.1477,
      "grad_norm": 5.6685943603515625,
      "learning_rate": 2.427777777777778e-05,
      "loss": 0.3184,
      "step": 5193
    },
    {
      "epoch": 1.1478,
      "grad_norm": 9.129612922668457,
      "learning_rate": 2.4272727272727275e-05,
      "loss": 0.1112,
      "step": 5194
    },
    {
      "epoch": 1.1479,
      "grad_norm": 2.5199522972106934,
      "learning_rate": 2.426767676767677e-05,
      "loss": 0.1888,
      "step": 5195
    },
    {
      "epoch": 1.148,
      "grad_norm": 3.56650447845459,
      "learning_rate": 2.4262626262626263e-05,
      "loss": 0.1062,
      "step": 5196
    },
    {
      "epoch": 1.1481,
      "grad_norm": 2.7913901805877686,
      "learning_rate": 2.425757575757576e-05,
      "loss": 0.4772,
      "step": 5197
    },
    {
      "epoch": 1.1482,
      "grad_norm": 3.1831717491149902,
      "learning_rate": 2.4252525252525255e-05,
      "loss": 0.1392,
      "step": 5198
    },
    {
      "epoch": 1.1482999999999999,
      "grad_norm": 2.3358304500579834,
      "learning_rate": 2.4247474747474748e-05,
      "loss": 0.215,
      "step": 5199
    },
    {
      "epoch": 1.1484,
      "grad_norm": 1.249155044555664,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.031,
      "step": 5200
    },
    {
      "epoch": 1.1485,
      "grad_norm": 7.929420471191406,
      "learning_rate": 2.423737373737374e-05,
      "loss": 0.7508,
      "step": 5201
    },
    {
      "epoch": 1.1486,
      "grad_norm": 3.5554301738739014,
      "learning_rate": 2.4232323232323232e-05,
      "loss": 0.0894,
      "step": 5202
    },
    {
      "epoch": 1.1487,
      "grad_norm": 2.489854097366333,
      "learning_rate": 2.422727272727273e-05,
      "loss": 0.4649,
      "step": 5203
    },
    {
      "epoch": 1.1488,
      "grad_norm": 13.807524681091309,
      "learning_rate": 2.4222222222222224e-05,
      "loss": 0.2234,
      "step": 5204
    },
    {
      "epoch": 1.1489,
      "grad_norm": 3.314765453338623,
      "learning_rate": 2.421717171717172e-05,
      "loss": 0.1081,
      "step": 5205
    },
    {
      "epoch": 1.149,
      "grad_norm": 4.247476100921631,
      "learning_rate": 2.4212121212121213e-05,
      "loss": 0.1229,
      "step": 5206
    },
    {
      "epoch": 1.1491,
      "grad_norm": 11.194107055664062,
      "learning_rate": 2.420707070707071e-05,
      "loss": 0.3206,
      "step": 5207
    },
    {
      "epoch": 1.1492,
      "grad_norm": 4.201623439788818,
      "learning_rate": 2.4202020202020205e-05,
      "loss": 0.6767,
      "step": 5208
    },
    {
      "epoch": 1.1493,
      "grad_norm": 0.7707706093788147,
      "learning_rate": 2.4196969696969698e-05,
      "loss": 0.0159,
      "step": 5209
    },
    {
      "epoch": 1.1494,
      "grad_norm": 2.3471336364746094,
      "learning_rate": 2.4191919191919194e-05,
      "loss": 0.0705,
      "step": 5210
    },
    {
      "epoch": 1.1495,
      "grad_norm": 5.240692615509033,
      "learning_rate": 2.418686868686869e-05,
      "loss": 0.1643,
      "step": 5211
    },
    {
      "epoch": 1.1496,
      "grad_norm": 6.1733198165893555,
      "learning_rate": 2.4181818181818182e-05,
      "loss": 0.6649,
      "step": 5212
    },
    {
      "epoch": 1.1497,
      "grad_norm": 4.671651840209961,
      "learning_rate": 2.4176767676767678e-05,
      "loss": 0.2844,
      "step": 5213
    },
    {
      "epoch": 1.1498,
      "grad_norm": 3.409233331680298,
      "learning_rate": 2.4171717171717174e-05,
      "loss": 0.1222,
      "step": 5214
    },
    {
      "epoch": 1.1499,
      "grad_norm": 9.58279037475586,
      "learning_rate": 2.4166666666666667e-05,
      "loss": 0.4862,
      "step": 5215
    },
    {
      "epoch": 1.15,
      "grad_norm": 3.670637607574463,
      "learning_rate": 2.4161616161616163e-05,
      "loss": 0.3011,
      "step": 5216
    },
    {
      "epoch": 1.1501000000000001,
      "grad_norm": 6.137842178344727,
      "learning_rate": 2.415656565656566e-05,
      "loss": 0.046,
      "step": 5217
    },
    {
      "epoch": 1.1502,
      "grad_norm": 3.137018918991089,
      "learning_rate": 2.415151515151515e-05,
      "loss": 0.2422,
      "step": 5218
    },
    {
      "epoch": 1.1503,
      "grad_norm": 4.660213470458984,
      "learning_rate": 2.4146464646464647e-05,
      "loss": 0.0426,
      "step": 5219
    },
    {
      "epoch": 1.1504,
      "grad_norm": 4.130661964416504,
      "learning_rate": 2.4141414141414143e-05,
      "loss": 0.3359,
      "step": 5220
    },
    {
      "epoch": 1.1505,
      "grad_norm": 5.184135437011719,
      "learning_rate": 2.4136363636363636e-05,
      "loss": 0.977,
      "step": 5221
    },
    {
      "epoch": 1.1506,
      "grad_norm": 14.1240816116333,
      "learning_rate": 2.4131313131313132e-05,
      "loss": 0.2475,
      "step": 5222
    },
    {
      "epoch": 1.1507,
      "grad_norm": 7.138504981994629,
      "learning_rate": 2.4126262626262628e-05,
      "loss": 0.2351,
      "step": 5223
    },
    {
      "epoch": 1.1508,
      "grad_norm": 4.010661602020264,
      "learning_rate": 2.4121212121212124e-05,
      "loss": 0.108,
      "step": 5224
    },
    {
      "epoch": 1.1509,
      "grad_norm": 10.049962043762207,
      "learning_rate": 2.4116161616161617e-05,
      "loss": 0.3106,
      "step": 5225
    },
    {
      "epoch": 1.151,
      "grad_norm": 6.473389148712158,
      "learning_rate": 2.4111111111111113e-05,
      "loss": 0.1252,
      "step": 5226
    },
    {
      "epoch": 1.1511,
      "grad_norm": 11.022449493408203,
      "learning_rate": 2.410606060606061e-05,
      "loss": 0.4525,
      "step": 5227
    },
    {
      "epoch": 1.1512,
      "grad_norm": 3.2878849506378174,
      "learning_rate": 2.41010101010101e-05,
      "loss": 0.5476,
      "step": 5228
    },
    {
      "epoch": 1.1513,
      "grad_norm": 6.946929454803467,
      "learning_rate": 2.4095959595959597e-05,
      "loss": 0.2229,
      "step": 5229
    },
    {
      "epoch": 1.1514,
      "grad_norm": 20.631765365600586,
      "learning_rate": 2.4090909090909093e-05,
      "loss": 0.6334,
      "step": 5230
    },
    {
      "epoch": 1.1515,
      "grad_norm": 4.71771240234375,
      "learning_rate": 2.4085858585858586e-05,
      "loss": 0.1573,
      "step": 5231
    },
    {
      "epoch": 1.1516,
      "grad_norm": 2.6184275150299072,
      "learning_rate": 2.4080808080808082e-05,
      "loss": 0.0952,
      "step": 5232
    },
    {
      "epoch": 1.1517,
      "grad_norm": 4.262170791625977,
      "learning_rate": 2.4075757575757578e-05,
      "loss": 0.2583,
      "step": 5233
    },
    {
      "epoch": 1.1518,
      "grad_norm": 6.774460315704346,
      "learning_rate": 2.407070707070707e-05,
      "loss": 0.2147,
      "step": 5234
    },
    {
      "epoch": 1.1519,
      "grad_norm": 7.36597204208374,
      "learning_rate": 2.4065656565656566e-05,
      "loss": 0.1537,
      "step": 5235
    },
    {
      "epoch": 1.152,
      "grad_norm": 14.243470191955566,
      "learning_rate": 2.4060606060606062e-05,
      "loss": 0.1424,
      "step": 5236
    },
    {
      "epoch": 1.1521,
      "grad_norm": 4.005297660827637,
      "learning_rate": 2.4055555555555555e-05,
      "loss": 0.2966,
      "step": 5237
    },
    {
      "epoch": 1.1522000000000001,
      "grad_norm": 2.223299026489258,
      "learning_rate": 2.405050505050505e-05,
      "loss": 0.0605,
      "step": 5238
    },
    {
      "epoch": 1.1522999999999999,
      "grad_norm": 1.3762108087539673,
      "learning_rate": 2.4045454545454547e-05,
      "loss": 0.0233,
      "step": 5239
    },
    {
      "epoch": 1.1524,
      "grad_norm": 8.128499031066895,
      "learning_rate": 2.404040404040404e-05,
      "loss": 0.4402,
      "step": 5240
    },
    {
      "epoch": 1.1525,
      "grad_norm": 1.6135501861572266,
      "learning_rate": 2.4035353535353535e-05,
      "loss": 0.4673,
      "step": 5241
    },
    {
      "epoch": 1.1526,
      "grad_norm": 2.9544830322265625,
      "learning_rate": 2.403030303030303e-05,
      "loss": 0.0748,
      "step": 5242
    },
    {
      "epoch": 1.1527,
      "grad_norm": 9.781081199645996,
      "learning_rate": 2.4025252525252524e-05,
      "loss": 0.1759,
      "step": 5243
    },
    {
      "epoch": 1.1528,
      "grad_norm": 2.3765740394592285,
      "learning_rate": 2.402020202020202e-05,
      "loss": 0.0707,
      "step": 5244
    },
    {
      "epoch": 1.1529,
      "grad_norm": 3.3569350242614746,
      "learning_rate": 2.4015151515151516e-05,
      "loss": 0.583,
      "step": 5245
    },
    {
      "epoch": 1.153,
      "grad_norm": 6.0660080909729,
      "learning_rate": 2.4010101010101012e-05,
      "loss": 0.1554,
      "step": 5246
    },
    {
      "epoch": 1.1531,
      "grad_norm": 4.827608585357666,
      "learning_rate": 2.4005050505050505e-05,
      "loss": 0.1213,
      "step": 5247
    },
    {
      "epoch": 1.1532,
      "grad_norm": 4.74304723739624,
      "learning_rate": 2.4e-05,
      "loss": 0.2281,
      "step": 5248
    },
    {
      "epoch": 1.1533,
      "grad_norm": 11.809650421142578,
      "learning_rate": 2.3994949494949497e-05,
      "loss": 0.3967,
      "step": 5249
    },
    {
      "epoch": 1.1534,
      "grad_norm": 4.990000247955322,
      "learning_rate": 2.398989898989899e-05,
      "loss": 0.2733,
      "step": 5250
    },
    {
      "epoch": 1.1535,
      "grad_norm": 4.5520124435424805,
      "learning_rate": 2.3984848484848485e-05,
      "loss": 0.5123,
      "step": 5251
    },
    {
      "epoch": 1.1536,
      "grad_norm": 2.254462957382202,
      "learning_rate": 2.397979797979798e-05,
      "loss": 0.1235,
      "step": 5252
    },
    {
      "epoch": 1.1537,
      "grad_norm": 4.8699445724487305,
      "learning_rate": 2.3974747474747474e-05,
      "loss": 0.1783,
      "step": 5253
    },
    {
      "epoch": 1.1538,
      "grad_norm": 6.0304741859436035,
      "learning_rate": 2.396969696969697e-05,
      "loss": 0.0811,
      "step": 5254
    },
    {
      "epoch": 1.1539,
      "grad_norm": 16.323041915893555,
      "learning_rate": 2.3964646464646466e-05,
      "loss": 0.1528,
      "step": 5255
    },
    {
      "epoch": 1.154,
      "grad_norm": 3.062220573425293,
      "learning_rate": 2.395959595959596e-05,
      "loss": 0.0887,
      "step": 5256
    },
    {
      "epoch": 1.1541,
      "grad_norm": 12.347646713256836,
      "learning_rate": 2.3954545454545454e-05,
      "loss": 0.1992,
      "step": 5257
    },
    {
      "epoch": 1.1542,
      "grad_norm": 8.35369873046875,
      "learning_rate": 2.394949494949495e-05,
      "loss": 0.197,
      "step": 5258
    },
    {
      "epoch": 1.1543,
      "grad_norm": 4.315428256988525,
      "learning_rate": 2.3944444444444443e-05,
      "loss": 0.4539,
      "step": 5259
    },
    {
      "epoch": 1.1544,
      "grad_norm": 5.6281609535217285,
      "learning_rate": 2.393939393939394e-05,
      "loss": 0.242,
      "step": 5260
    },
    {
      "epoch": 1.1545,
      "grad_norm": 1.7478257417678833,
      "learning_rate": 2.3934343434343435e-05,
      "loss": 0.0246,
      "step": 5261
    },
    {
      "epoch": 1.1546,
      "grad_norm": 41.474666595458984,
      "learning_rate": 2.3929292929292928e-05,
      "loss": 0.0815,
      "step": 5262
    },
    {
      "epoch": 1.1547,
      "grad_norm": 6.210231304168701,
      "learning_rate": 2.3924242424242424e-05,
      "loss": 0.304,
      "step": 5263
    },
    {
      "epoch": 1.1548,
      "grad_norm": 1.092751145362854,
      "learning_rate": 2.3919191919191923e-05,
      "loss": 0.0049,
      "step": 5264
    },
    {
      "epoch": 1.1549,
      "grad_norm": 2.826411247253418,
      "learning_rate": 2.3914141414141416e-05,
      "loss": 0.0566,
      "step": 5265
    },
    {
      "epoch": 1.155,
      "grad_norm": 5.6577935218811035,
      "learning_rate": 2.390909090909091e-05,
      "loss": 0.2108,
      "step": 5266
    },
    {
      "epoch": 1.1551,
      "grad_norm": 4.539240837097168,
      "learning_rate": 2.3904040404040407e-05,
      "loss": 0.7805,
      "step": 5267
    },
    {
      "epoch": 1.1552,
      "grad_norm": 14.55349063873291,
      "learning_rate": 2.38989898989899e-05,
      "loss": 0.4076,
      "step": 5268
    },
    {
      "epoch": 1.1553,
      "grad_norm": 3.2956793308258057,
      "learning_rate": 2.3893939393939396e-05,
      "loss": 0.0932,
      "step": 5269
    },
    {
      "epoch": 1.1554,
      "grad_norm": 4.394725799560547,
      "learning_rate": 2.3888888888888892e-05,
      "loss": 0.218,
      "step": 5270
    },
    {
      "epoch": 1.1555,
      "grad_norm": 4.696659564971924,
      "learning_rate": 2.3883838383838385e-05,
      "loss": 0.3778,
      "step": 5271
    },
    {
      "epoch": 1.1556,
      "grad_norm": 20.893762588500977,
      "learning_rate": 2.387878787878788e-05,
      "loss": 0.1243,
      "step": 5272
    },
    {
      "epoch": 1.1557,
      "grad_norm": 0.44036075472831726,
      "learning_rate": 2.3873737373737377e-05,
      "loss": 0.0032,
      "step": 5273
    },
    {
      "epoch": 1.1558,
      "grad_norm": 16.204391479492188,
      "learning_rate": 2.386868686868687e-05,
      "loss": 0.3107,
      "step": 5274
    },
    {
      "epoch": 1.1559,
      "grad_norm": 4.413903713226318,
      "learning_rate": 2.3863636363636365e-05,
      "loss": 0.1366,
      "step": 5275
    },
    {
      "epoch": 1.156,
      "grad_norm": 2.995917558670044,
      "learning_rate": 2.385858585858586e-05,
      "loss": 0.0619,
      "step": 5276
    },
    {
      "epoch": 1.1561,
      "grad_norm": 5.88496732711792,
      "learning_rate": 2.3853535353535357e-05,
      "loss": 0.1457,
      "step": 5277
    },
    {
      "epoch": 1.1562000000000001,
      "grad_norm": 3.537994623184204,
      "learning_rate": 2.384848484848485e-05,
      "loss": 0.1168,
      "step": 5278
    },
    {
      "epoch": 1.1562999999999999,
      "grad_norm": 3.795682907104492,
      "learning_rate": 2.3843434343434346e-05,
      "loss": 0.2297,
      "step": 5279
    },
    {
      "epoch": 1.1564,
      "grad_norm": 2.757481575012207,
      "learning_rate": 2.3838383838383842e-05,
      "loss": 0.5096,
      "step": 5280
    },
    {
      "epoch": 1.1565,
      "grad_norm": 13.258453369140625,
      "learning_rate": 2.3833333333333334e-05,
      "loss": 0.3241,
      "step": 5281
    },
    {
      "epoch": 1.1566,
      "grad_norm": 4.227352619171143,
      "learning_rate": 2.382828282828283e-05,
      "loss": 0.7971,
      "step": 5282
    },
    {
      "epoch": 1.1567,
      "grad_norm": 9.656052589416504,
      "learning_rate": 2.3823232323232326e-05,
      "loss": 0.0568,
      "step": 5283
    },
    {
      "epoch": 1.1568,
      "grad_norm": 10.709249496459961,
      "learning_rate": 2.381818181818182e-05,
      "loss": 0.4221,
      "step": 5284
    },
    {
      "epoch": 1.1569,
      "grad_norm": 4.735139846801758,
      "learning_rate": 2.3813131313131315e-05,
      "loss": 0.6971,
      "step": 5285
    },
    {
      "epoch": 1.157,
      "grad_norm": 3.1903538703918457,
      "learning_rate": 2.380808080808081e-05,
      "loss": 0.254,
      "step": 5286
    },
    {
      "epoch": 1.1571,
      "grad_norm": 6.971375465393066,
      "learning_rate": 2.3803030303030304e-05,
      "loss": 0.3501,
      "step": 5287
    },
    {
      "epoch": 1.1572,
      "grad_norm": 2.771714925765991,
      "learning_rate": 2.37979797979798e-05,
      "loss": 0.1011,
      "step": 5288
    },
    {
      "epoch": 1.1573,
      "grad_norm": 5.7561936378479,
      "learning_rate": 2.3792929292929296e-05,
      "loss": 0.2189,
      "step": 5289
    },
    {
      "epoch": 1.1574,
      "grad_norm": 2.4800174236297607,
      "learning_rate": 2.3787878787878788e-05,
      "loss": 0.0899,
      "step": 5290
    },
    {
      "epoch": 1.1575,
      "grad_norm": 4.941097736358643,
      "learning_rate": 2.3782828282828284e-05,
      "loss": 0.4245,
      "step": 5291
    },
    {
      "epoch": 1.1576,
      "grad_norm": 5.176090717315674,
      "learning_rate": 2.377777777777778e-05,
      "loss": 0.3429,
      "step": 5292
    },
    {
      "epoch": 1.1577,
      "grad_norm": 1.7095109224319458,
      "learning_rate": 2.3772727272727273e-05,
      "loss": 0.056,
      "step": 5293
    },
    {
      "epoch": 1.1578,
      "grad_norm": 3.8754773139953613,
      "learning_rate": 2.376767676767677e-05,
      "loss": 0.574,
      "step": 5294
    },
    {
      "epoch": 1.1579,
      "grad_norm": 5.26104211807251,
      "learning_rate": 2.3762626262626265e-05,
      "loss": 0.1496,
      "step": 5295
    },
    {
      "epoch": 1.158,
      "grad_norm": 3.8937747478485107,
      "learning_rate": 2.375757575757576e-05,
      "loss": 0.0539,
      "step": 5296
    },
    {
      "epoch": 1.1581,
      "grad_norm": 3.838218927383423,
      "learning_rate": 2.3752525252525253e-05,
      "loss": 0.2234,
      "step": 5297
    },
    {
      "epoch": 1.1582,
      "grad_norm": 10.07823657989502,
      "learning_rate": 2.374747474747475e-05,
      "loss": 0.1478,
      "step": 5298
    },
    {
      "epoch": 1.1583,
      "grad_norm": 4.381494522094727,
      "learning_rate": 2.3742424242424245e-05,
      "loss": 0.0387,
      "step": 5299
    },
    {
      "epoch": 1.1584,
      "grad_norm": 7.616401672363281,
      "learning_rate": 2.3737373737373738e-05,
      "loss": 0.3125,
      "step": 5300
    },
    {
      "epoch": 1.1585,
      "grad_norm": 7.331244945526123,
      "learning_rate": 2.3732323232323234e-05,
      "loss": 0.1556,
      "step": 5301
    },
    {
      "epoch": 1.1586,
      "grad_norm": 4.962352752685547,
      "learning_rate": 2.372727272727273e-05,
      "loss": 0.1716,
      "step": 5302
    },
    {
      "epoch": 1.1587,
      "grad_norm": 11.828078269958496,
      "learning_rate": 2.3722222222222222e-05,
      "loss": 0.1753,
      "step": 5303
    },
    {
      "epoch": 1.1588,
      "grad_norm": 3.595783233642578,
      "learning_rate": 2.371717171717172e-05,
      "loss": 0.1566,
      "step": 5304
    },
    {
      "epoch": 1.1589,
      "grad_norm": 5.139683723449707,
      "learning_rate": 2.3712121212121214e-05,
      "loss": 0.6825,
      "step": 5305
    },
    {
      "epoch": 1.159,
      "grad_norm": 5.4909844398498535,
      "learning_rate": 2.3707070707070707e-05,
      "loss": 0.4816,
      "step": 5306
    },
    {
      "epoch": 1.1591,
      "grad_norm": 4.600339889526367,
      "learning_rate": 2.3702020202020203e-05,
      "loss": 0.4046,
      "step": 5307
    },
    {
      "epoch": 1.1592,
      "grad_norm": 10.772997856140137,
      "learning_rate": 2.36969696969697e-05,
      "loss": 0.1228,
      "step": 5308
    },
    {
      "epoch": 1.1593,
      "grad_norm": 6.302374839782715,
      "learning_rate": 2.369191919191919e-05,
      "loss": 0.2567,
      "step": 5309
    },
    {
      "epoch": 1.1594,
      "grad_norm": 2.345188617706299,
      "learning_rate": 2.3686868686868688e-05,
      "loss": 0.094,
      "step": 5310
    },
    {
      "epoch": 1.1595,
      "grad_norm": 3.076612949371338,
      "learning_rate": 2.3681818181818184e-05,
      "loss": 0.5471,
      "step": 5311
    },
    {
      "epoch": 1.1596,
      "grad_norm": 6.294826507568359,
      "learning_rate": 2.3676767676767676e-05,
      "loss": 0.7787,
      "step": 5312
    },
    {
      "epoch": 1.1597,
      "grad_norm": 4.78568696975708,
      "learning_rate": 2.3671717171717172e-05,
      "loss": 0.6033,
      "step": 5313
    },
    {
      "epoch": 1.1598,
      "grad_norm": 5.388808727264404,
      "learning_rate": 2.3666666666666668e-05,
      "loss": 0.2376,
      "step": 5314
    },
    {
      "epoch": 1.1599,
      "grad_norm": 4.406245708465576,
      "learning_rate": 2.366161616161616e-05,
      "loss": 0.1973,
      "step": 5315
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.034773349761963,
      "learning_rate": 2.3656565656565657e-05,
      "loss": 0.2689,
      "step": 5316
    },
    {
      "epoch": 1.1601,
      "grad_norm": 5.808801174163818,
      "learning_rate": 2.3651515151515153e-05,
      "loss": 0.3645,
      "step": 5317
    },
    {
      "epoch": 1.1602000000000001,
      "grad_norm": 10.806351661682129,
      "learning_rate": 2.364646464646465e-05,
      "loss": 0.2371,
      "step": 5318
    },
    {
      "epoch": 1.1602999999999999,
      "grad_norm": 8.635770797729492,
      "learning_rate": 2.364141414141414e-05,
      "loss": 0.1752,
      "step": 5319
    },
    {
      "epoch": 1.1604,
      "grad_norm": 26.449724197387695,
      "learning_rate": 2.3636363636363637e-05,
      "loss": 0.7175,
      "step": 5320
    },
    {
      "epoch": 1.1605,
      "grad_norm": 4.2079010009765625,
      "learning_rate": 2.3631313131313133e-05,
      "loss": 0.2086,
      "step": 5321
    },
    {
      "epoch": 1.1606,
      "grad_norm": 5.816718578338623,
      "learning_rate": 2.3626262626262626e-05,
      "loss": 1.0541,
      "step": 5322
    },
    {
      "epoch": 1.1607,
      "grad_norm": 10.270827293395996,
      "learning_rate": 2.3621212121212122e-05,
      "loss": 0.2063,
      "step": 5323
    },
    {
      "epoch": 1.1608,
      "grad_norm": 1.5510354042053223,
      "learning_rate": 2.3616161616161618e-05,
      "loss": 0.9637,
      "step": 5324
    },
    {
      "epoch": 1.1609,
      "grad_norm": 6.6787028312683105,
      "learning_rate": 2.361111111111111e-05,
      "loss": 0.2495,
      "step": 5325
    },
    {
      "epoch": 1.161,
      "grad_norm": 5.056324005126953,
      "learning_rate": 2.3606060606060607e-05,
      "loss": 0.1594,
      "step": 5326
    },
    {
      "epoch": 1.1611,
      "grad_norm": 3.3840296268463135,
      "learning_rate": 2.3601010101010103e-05,
      "loss": 0.2471,
      "step": 5327
    },
    {
      "epoch": 1.1612,
      "grad_norm": 11.509418487548828,
      "learning_rate": 2.3595959595959595e-05,
      "loss": 0.37,
      "step": 5328
    },
    {
      "epoch": 1.1613,
      "grad_norm": 8.596117973327637,
      "learning_rate": 2.359090909090909e-05,
      "loss": 0.5526,
      "step": 5329
    },
    {
      "epoch": 1.1614,
      "grad_norm": 6.410821914672852,
      "learning_rate": 2.3585858585858587e-05,
      "loss": 0.1156,
      "step": 5330
    },
    {
      "epoch": 1.1615,
      "grad_norm": 6.783490180969238,
      "learning_rate": 2.358080808080808e-05,
      "loss": 0.0497,
      "step": 5331
    },
    {
      "epoch": 1.1616,
      "grad_norm": 4.427325248718262,
      "learning_rate": 2.3575757575757576e-05,
      "loss": 0.1206,
      "step": 5332
    },
    {
      "epoch": 1.1617,
      "grad_norm": 6.727798938751221,
      "learning_rate": 2.3570707070707072e-05,
      "loss": 0.2468,
      "step": 5333
    },
    {
      "epoch": 1.1618,
      "grad_norm": 3.4442567825317383,
      "learning_rate": 2.3565656565656564e-05,
      "loss": 0.085,
      "step": 5334
    },
    {
      "epoch": 1.1619,
      "grad_norm": 3.1344985961914062,
      "learning_rate": 2.356060606060606e-05,
      "loss": 0.0611,
      "step": 5335
    },
    {
      "epoch": 1.162,
      "grad_norm": 1.370388388633728,
      "learning_rate": 2.3555555555555556e-05,
      "loss": 0.4833,
      "step": 5336
    },
    {
      "epoch": 1.1621,
      "grad_norm": 4.768537521362305,
      "learning_rate": 2.3550505050505052e-05,
      "loss": 0.632,
      "step": 5337
    },
    {
      "epoch": 1.1622,
      "grad_norm": 4.08383846282959,
      "learning_rate": 2.3545454545454545e-05,
      "loss": 0.3141,
      "step": 5338
    },
    {
      "epoch": 1.1623,
      "grad_norm": 13.136862754821777,
      "learning_rate": 2.354040404040404e-05,
      "loss": 0.3015,
      "step": 5339
    },
    {
      "epoch": 1.1623999999999999,
      "grad_norm": 8.634408950805664,
      "learning_rate": 2.3535353535353537e-05,
      "loss": 0.1713,
      "step": 5340
    },
    {
      "epoch": 1.1625,
      "grad_norm": 6.727282524108887,
      "learning_rate": 2.353030303030303e-05,
      "loss": 0.1192,
      "step": 5341
    },
    {
      "epoch": 1.1626,
      "grad_norm": 3.2574355602264404,
      "learning_rate": 2.3525252525252525e-05,
      "loss": 0.1828,
      "step": 5342
    },
    {
      "epoch": 1.1627,
      "grad_norm": 4.057934284210205,
      "learning_rate": 2.352020202020202e-05,
      "loss": 0.0579,
      "step": 5343
    },
    {
      "epoch": 1.1628,
      "grad_norm": 2.168262243270874,
      "learning_rate": 2.3515151515151514e-05,
      "loss": 0.1076,
      "step": 5344
    },
    {
      "epoch": 1.1629,
      "grad_norm": 3.380601644515991,
      "learning_rate": 2.351010101010101e-05,
      "loss": 0.1413,
      "step": 5345
    },
    {
      "epoch": 1.163,
      "grad_norm": 8.271151542663574,
      "learning_rate": 2.3505050505050506e-05,
      "loss": 0.1079,
      "step": 5346
    },
    {
      "epoch": 1.1631,
      "grad_norm": 10.102361679077148,
      "learning_rate": 2.35e-05,
      "loss": 0.2739,
      "step": 5347
    },
    {
      "epoch": 1.1632,
      "grad_norm": 8.928216934204102,
      "learning_rate": 2.3494949494949495e-05,
      "loss": 0.3228,
      "step": 5348
    },
    {
      "epoch": 1.1633,
      "grad_norm": 32.35628128051758,
      "learning_rate": 2.3489898989898994e-05,
      "loss": 1.7361,
      "step": 5349
    },
    {
      "epoch": 1.1634,
      "grad_norm": 3.8595707416534424,
      "learning_rate": 2.3484848484848487e-05,
      "loss": 0.2138,
      "step": 5350
    },
    {
      "epoch": 1.1635,
      "grad_norm": 2.455465078353882,
      "learning_rate": 2.3479797979797983e-05,
      "loss": 0.0717,
      "step": 5351
    },
    {
      "epoch": 1.1636,
      "grad_norm": 4.244874000549316,
      "learning_rate": 2.347474747474748e-05,
      "loss": 0.2081,
      "step": 5352
    },
    {
      "epoch": 1.1637,
      "grad_norm": 2.151803493499756,
      "learning_rate": 2.346969696969697e-05,
      "loss": 0.2501,
      "step": 5353
    },
    {
      "epoch": 1.1638,
      "grad_norm": 5.137261390686035,
      "learning_rate": 2.3464646464646467e-05,
      "loss": 0.0753,
      "step": 5354
    },
    {
      "epoch": 1.1639,
      "grad_norm": 12.602594375610352,
      "learning_rate": 2.3459595959595963e-05,
      "loss": 0.3394,
      "step": 5355
    },
    {
      "epoch": 1.164,
      "grad_norm": 5.388693809509277,
      "learning_rate": 2.3454545454545456e-05,
      "loss": 0.8203,
      "step": 5356
    },
    {
      "epoch": 1.1641,
      "grad_norm": 13.166933059692383,
      "learning_rate": 2.3449494949494952e-05,
      "loss": 0.8292,
      "step": 5357
    },
    {
      "epoch": 1.1642000000000001,
      "grad_norm": 3.669250011444092,
      "learning_rate": 2.3444444444444448e-05,
      "loss": 0.2109,
      "step": 5358
    },
    {
      "epoch": 1.1643,
      "grad_norm": 1.986613392829895,
      "learning_rate": 2.343939393939394e-05,
      "loss": 0.0885,
      "step": 5359
    },
    {
      "epoch": 1.1644,
      "grad_norm": 8.097289085388184,
      "learning_rate": 2.3434343434343436e-05,
      "loss": 0.3605,
      "step": 5360
    },
    {
      "epoch": 1.1645,
      "grad_norm": 13.94277572631836,
      "learning_rate": 2.3429292929292932e-05,
      "loss": 0.9781,
      "step": 5361
    },
    {
      "epoch": 1.1646,
      "grad_norm": 3.906566619873047,
      "learning_rate": 2.3424242424242425e-05,
      "loss": 0.3154,
      "step": 5362
    },
    {
      "epoch": 1.1647,
      "grad_norm": 7.5940070152282715,
      "learning_rate": 2.341919191919192e-05,
      "loss": 0.3329,
      "step": 5363
    },
    {
      "epoch": 1.1648,
      "grad_norm": 7.718069076538086,
      "learning_rate": 2.3414141414141417e-05,
      "loss": 0.5288,
      "step": 5364
    },
    {
      "epoch": 1.1649,
      "grad_norm": 9.655661582946777,
      "learning_rate": 2.340909090909091e-05,
      "loss": 0.2834,
      "step": 5365
    },
    {
      "epoch": 1.165,
      "grad_norm": 6.0250444412231445,
      "learning_rate": 2.3404040404040405e-05,
      "loss": 0.1684,
      "step": 5366
    },
    {
      "epoch": 1.1651,
      "grad_norm": 6.718356609344482,
      "learning_rate": 2.33989898989899e-05,
      "loss": 0.2847,
      "step": 5367
    },
    {
      "epoch": 1.1652,
      "grad_norm": 2.370281457901001,
      "learning_rate": 2.3393939393939397e-05,
      "loss": 0.1046,
      "step": 5368
    },
    {
      "epoch": 1.1653,
      "grad_norm": 3.4982709884643555,
      "learning_rate": 2.338888888888889e-05,
      "loss": 0.1428,
      "step": 5369
    },
    {
      "epoch": 1.1654,
      "grad_norm": 2.6514737606048584,
      "learning_rate": 2.3383838383838386e-05,
      "loss": 0.5175,
      "step": 5370
    },
    {
      "epoch": 1.1655,
      "grad_norm": 2.1568329334259033,
      "learning_rate": 2.3378787878787882e-05,
      "loss": 1.0108,
      "step": 5371
    },
    {
      "epoch": 1.1656,
      "grad_norm": 0.9871082901954651,
      "learning_rate": 2.3373737373737375e-05,
      "loss": 0.0247,
      "step": 5372
    },
    {
      "epoch": 1.1657,
      "grad_norm": 2.560000419616699,
      "learning_rate": 2.336868686868687e-05,
      "loss": 0.4888,
      "step": 5373
    },
    {
      "epoch": 1.1658,
      "grad_norm": 6.758738994598389,
      "learning_rate": 2.3363636363636367e-05,
      "loss": 0.3199,
      "step": 5374
    },
    {
      "epoch": 1.1659,
      "grad_norm": 1.9432401657104492,
      "learning_rate": 2.335858585858586e-05,
      "loss": 0.022,
      "step": 5375
    },
    {
      "epoch": 1.166,
      "grad_norm": 5.728640556335449,
      "learning_rate": 2.3353535353535355e-05,
      "loss": 0.376,
      "step": 5376
    },
    {
      "epoch": 1.1661,
      "grad_norm": 2.869727373123169,
      "learning_rate": 2.334848484848485e-05,
      "loss": 0.0184,
      "step": 5377
    },
    {
      "epoch": 1.1662,
      "grad_norm": 5.214388847351074,
      "learning_rate": 2.3343434343434344e-05,
      "loss": 0.251,
      "step": 5378
    },
    {
      "epoch": 1.1663000000000001,
      "grad_norm": 4.516401290893555,
      "learning_rate": 2.333838383838384e-05,
      "loss": 0.3935,
      "step": 5379
    },
    {
      "epoch": 1.1663999999999999,
      "grad_norm": 3.8566176891326904,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.332,
      "step": 5380
    },
    {
      "epoch": 1.1665,
      "grad_norm": 4.395230770111084,
      "learning_rate": 2.332828282828283e-05,
      "loss": 0.1833,
      "step": 5381
    },
    {
      "epoch": 1.1666,
      "grad_norm": 1.417683720588684,
      "learning_rate": 2.3323232323232324e-05,
      "loss": 0.0296,
      "step": 5382
    },
    {
      "epoch": 1.1667,
      "grad_norm": 2.1594722270965576,
      "learning_rate": 2.331818181818182e-05,
      "loss": 0.2197,
      "step": 5383
    },
    {
      "epoch": 1.1668,
      "grad_norm": 3.264962911605835,
      "learning_rate": 2.3313131313131313e-05,
      "loss": 0.1659,
      "step": 5384
    },
    {
      "epoch": 1.1669,
      "grad_norm": 4.204865455627441,
      "learning_rate": 2.330808080808081e-05,
      "loss": 0.1622,
      "step": 5385
    },
    {
      "epoch": 1.167,
      "grad_norm": 7.329836845397949,
      "learning_rate": 2.3303030303030305e-05,
      "loss": 0.1972,
      "step": 5386
    },
    {
      "epoch": 1.1671,
      "grad_norm": 2.688175678253174,
      "learning_rate": 2.3297979797979798e-05,
      "loss": 0.0409,
      "step": 5387
    },
    {
      "epoch": 1.1672,
      "grad_norm": 9.045695304870605,
      "learning_rate": 2.3292929292929294e-05,
      "loss": 0.1632,
      "step": 5388
    },
    {
      "epoch": 1.1673,
      "grad_norm": 17.948612213134766,
      "learning_rate": 2.328787878787879e-05,
      "loss": 0.5006,
      "step": 5389
    },
    {
      "epoch": 1.1674,
      "grad_norm": 18.713857650756836,
      "learning_rate": 2.3282828282828286e-05,
      "loss": 0.2429,
      "step": 5390
    },
    {
      "epoch": 1.1675,
      "grad_norm": 6.667062282562256,
      "learning_rate": 2.3277777777777778e-05,
      "loss": 0.7532,
      "step": 5391
    },
    {
      "epoch": 1.1676,
      "grad_norm": 8.586874008178711,
      "learning_rate": 2.3272727272727274e-05,
      "loss": 0.5949,
      "step": 5392
    },
    {
      "epoch": 1.1677,
      "grad_norm": 4.331235408782959,
      "learning_rate": 2.326767676767677e-05,
      "loss": 0.1312,
      "step": 5393
    },
    {
      "epoch": 1.1678,
      "grad_norm": 2.7389066219329834,
      "learning_rate": 2.3262626262626263e-05,
      "loss": 0.0583,
      "step": 5394
    },
    {
      "epoch": 1.1679,
      "grad_norm": 1.749775767326355,
      "learning_rate": 2.325757575757576e-05,
      "loss": 0.0428,
      "step": 5395
    },
    {
      "epoch": 1.168,
      "grad_norm": 2.4712088108062744,
      "learning_rate": 2.3252525252525255e-05,
      "loss": 0.0088,
      "step": 5396
    },
    {
      "epoch": 1.1681,
      "grad_norm": 0.654240071773529,
      "learning_rate": 2.3247474747474747e-05,
      "loss": 0.0169,
      "step": 5397
    },
    {
      "epoch": 1.1682,
      "grad_norm": 6.585247993469238,
      "learning_rate": 2.3242424242424243e-05,
      "loss": 0.1122,
      "step": 5398
    },
    {
      "epoch": 1.1683,
      "grad_norm": 33.13890838623047,
      "learning_rate": 2.323737373737374e-05,
      "loss": 0.0966,
      "step": 5399
    },
    {
      "epoch": 1.1684,
      "grad_norm": 4.010470867156982,
      "learning_rate": 2.3232323232323232e-05,
      "loss": 0.1247,
      "step": 5400
    },
    {
      "epoch": 1.1685,
      "grad_norm": 1.7978235483169556,
      "learning_rate": 2.3227272727272728e-05,
      "loss": 0.0721,
      "step": 5401
    },
    {
      "epoch": 1.1686,
      "grad_norm": 2.6459147930145264,
      "learning_rate": 2.3222222222222224e-05,
      "loss": 0.1766,
      "step": 5402
    },
    {
      "epoch": 1.1687,
      "grad_norm": 6.2986650466918945,
      "learning_rate": 2.3217171717171716e-05,
      "loss": 0.3685,
      "step": 5403
    },
    {
      "epoch": 1.1688,
      "grad_norm": 4.914694786071777,
      "learning_rate": 2.3212121212121212e-05,
      "loss": 0.7443,
      "step": 5404
    },
    {
      "epoch": 1.1689,
      "grad_norm": 0.574228823184967,
      "learning_rate": 2.320707070707071e-05,
      "loss": 0.0067,
      "step": 5405
    },
    {
      "epoch": 1.169,
      "grad_norm": 3.6442182064056396,
      "learning_rate": 2.32020202020202e-05,
      "loss": 0.0399,
      "step": 5406
    },
    {
      "epoch": 1.1691,
      "grad_norm": 4.31878137588501,
      "learning_rate": 2.3196969696969697e-05,
      "loss": 0.3163,
      "step": 5407
    },
    {
      "epoch": 1.1692,
      "grad_norm": 3.072453498840332,
      "learning_rate": 2.3191919191919193e-05,
      "loss": 0.4562,
      "step": 5408
    },
    {
      "epoch": 1.1693,
      "grad_norm": 8.810712814331055,
      "learning_rate": 2.318686868686869e-05,
      "loss": 0.5596,
      "step": 5409
    },
    {
      "epoch": 1.1694,
      "grad_norm": 8.599708557128906,
      "learning_rate": 2.318181818181818e-05,
      "loss": 0.1238,
      "step": 5410
    },
    {
      "epoch": 1.1695,
      "grad_norm": 3.0722153186798096,
      "learning_rate": 2.3176767676767678e-05,
      "loss": 0.1614,
      "step": 5411
    },
    {
      "epoch": 1.1696,
      "grad_norm": 2.4556937217712402,
      "learning_rate": 2.3171717171717174e-05,
      "loss": 0.0511,
      "step": 5412
    },
    {
      "epoch": 1.1697,
      "grad_norm": 7.893933296203613,
      "learning_rate": 2.3166666666666666e-05,
      "loss": 0.1215,
      "step": 5413
    },
    {
      "epoch": 1.1698,
      "grad_norm": 10.920034408569336,
      "learning_rate": 2.3161616161616162e-05,
      "loss": 0.0793,
      "step": 5414
    },
    {
      "epoch": 1.1699,
      "grad_norm": 3.8610475063323975,
      "learning_rate": 2.3156565656565658e-05,
      "loss": 0.1933,
      "step": 5415
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.0225279331207275,
      "learning_rate": 2.315151515151515e-05,
      "loss": 0.264,
      "step": 5416
    },
    {
      "epoch": 1.1701,
      "grad_norm": 5.012161731719971,
      "learning_rate": 2.3146464646464647e-05,
      "loss": 0.6908,
      "step": 5417
    },
    {
      "epoch": 1.1702,
      "grad_norm": 2.44518780708313,
      "learning_rate": 2.3141414141414143e-05,
      "loss": 0.0712,
      "step": 5418
    },
    {
      "epoch": 1.1703000000000001,
      "grad_norm": 7.288547992706299,
      "learning_rate": 2.3136363636363635e-05,
      "loss": 0.0602,
      "step": 5419
    },
    {
      "epoch": 1.1703999999999999,
      "grad_norm": 16.546045303344727,
      "learning_rate": 2.313131313131313e-05,
      "loss": 0.0665,
      "step": 5420
    },
    {
      "epoch": 1.1705,
      "grad_norm": 2.527491569519043,
      "learning_rate": 2.3126262626262627e-05,
      "loss": 0.1995,
      "step": 5421
    },
    {
      "epoch": 1.1706,
      "grad_norm": 4.24088191986084,
      "learning_rate": 2.312121212121212e-05,
      "loss": 0.0938,
      "step": 5422
    },
    {
      "epoch": 1.1707,
      "grad_norm": 3.7560975551605225,
      "learning_rate": 2.3116161616161616e-05,
      "loss": 0.159,
      "step": 5423
    },
    {
      "epoch": 1.1708,
      "grad_norm": 2.8262593746185303,
      "learning_rate": 2.3111111111111112e-05,
      "loss": 0.0887,
      "step": 5424
    },
    {
      "epoch": 1.1709,
      "grad_norm": 1.992082118988037,
      "learning_rate": 2.3106060606060605e-05,
      "loss": 0.013,
      "step": 5425
    },
    {
      "epoch": 1.171,
      "grad_norm": 3.1185615062713623,
      "learning_rate": 2.31010101010101e-05,
      "loss": 0.3067,
      "step": 5426
    },
    {
      "epoch": 1.1711,
      "grad_norm": 3.674757957458496,
      "learning_rate": 2.3095959595959597e-05,
      "loss": 0.18,
      "step": 5427
    },
    {
      "epoch": 1.1712,
      "grad_norm": 2.025519847869873,
      "learning_rate": 2.309090909090909e-05,
      "loss": 0.2735,
      "step": 5428
    },
    {
      "epoch": 1.1713,
      "grad_norm": 2.2047574520111084,
      "learning_rate": 2.3085858585858585e-05,
      "loss": 0.0616,
      "step": 5429
    },
    {
      "epoch": 1.1714,
      "grad_norm": 3.4750728607177734,
      "learning_rate": 2.308080808080808e-05,
      "loss": 0.1405,
      "step": 5430
    },
    {
      "epoch": 1.1715,
      "grad_norm": 4.061604022979736,
      "learning_rate": 2.3075757575757577e-05,
      "loss": 0.1448,
      "step": 5431
    },
    {
      "epoch": 1.1716,
      "grad_norm": 2.9295260906219482,
      "learning_rate": 2.307070707070707e-05,
      "loss": 0.0754,
      "step": 5432
    },
    {
      "epoch": 1.1717,
      "grad_norm": 3.7035200595855713,
      "learning_rate": 2.3065656565656566e-05,
      "loss": 0.25,
      "step": 5433
    },
    {
      "epoch": 1.1718,
      "grad_norm": 4.871304035186768,
      "learning_rate": 2.306060606060606e-05,
      "loss": 0.3257,
      "step": 5434
    },
    {
      "epoch": 1.1719,
      "grad_norm": 1.7785111665725708,
      "learning_rate": 2.3055555555555558e-05,
      "loss": 0.486,
      "step": 5435
    },
    {
      "epoch": 1.172,
      "grad_norm": 3.046701669692993,
      "learning_rate": 2.3050505050505054e-05,
      "loss": 0.186,
      "step": 5436
    },
    {
      "epoch": 1.1721,
      "grad_norm": 6.4331159591674805,
      "learning_rate": 2.3045454545454546e-05,
      "loss": 0.7452,
      "step": 5437
    },
    {
      "epoch": 1.1722,
      "grad_norm": 12.576933860778809,
      "learning_rate": 2.3040404040404042e-05,
      "loss": 0.3964,
      "step": 5438
    },
    {
      "epoch": 1.1723,
      "grad_norm": 7.708917140960693,
      "learning_rate": 2.3035353535353538e-05,
      "loss": 0.0856,
      "step": 5439
    },
    {
      "epoch": 1.1724,
      "grad_norm": 3.0938475131988525,
      "learning_rate": 2.3030303030303034e-05,
      "loss": 0.1098,
      "step": 5440
    },
    {
      "epoch": 1.1724999999999999,
      "grad_norm": 7.1778435707092285,
      "learning_rate": 2.3025252525252527e-05,
      "loss": 0.4507,
      "step": 5441
    },
    {
      "epoch": 1.1726,
      "grad_norm": 8.137772560119629,
      "learning_rate": 2.3020202020202023e-05,
      "loss": 0.0986,
      "step": 5442
    },
    {
      "epoch": 1.1727,
      "grad_norm": 1.0203931331634521,
      "learning_rate": 2.301515151515152e-05,
      "loss": 0.0062,
      "step": 5443
    },
    {
      "epoch": 1.1728,
      "grad_norm": 9.610995292663574,
      "learning_rate": 2.301010101010101e-05,
      "loss": 0.1359,
      "step": 5444
    },
    {
      "epoch": 1.1729,
      "grad_norm": 4.88545036315918,
      "learning_rate": 2.3005050505050507e-05,
      "loss": 0.0264,
      "step": 5445
    },
    {
      "epoch": 1.173,
      "grad_norm": 2.633821487426758,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.2456,
      "step": 5446
    },
    {
      "epoch": 1.1731,
      "grad_norm": 4.653584003448486,
      "learning_rate": 2.2994949494949496e-05,
      "loss": 0.6193,
      "step": 5447
    },
    {
      "epoch": 1.1732,
      "grad_norm": 5.493785858154297,
      "learning_rate": 2.2989898989898992e-05,
      "loss": 0.2535,
      "step": 5448
    },
    {
      "epoch": 1.1733,
      "grad_norm": 3.475191116333008,
      "learning_rate": 2.2984848484848488e-05,
      "loss": 0.116,
      "step": 5449
    },
    {
      "epoch": 1.1734,
      "grad_norm": 5.158650875091553,
      "learning_rate": 2.297979797979798e-05,
      "loss": 0.2916,
      "step": 5450
    },
    {
      "epoch": 1.1735,
      "grad_norm": 8.625284194946289,
      "learning_rate": 2.2974747474747477e-05,
      "loss": 0.1545,
      "step": 5451
    },
    {
      "epoch": 1.1736,
      "grad_norm": 2.747210741043091,
      "learning_rate": 2.2969696969696973e-05,
      "loss": 0.6906,
      "step": 5452
    },
    {
      "epoch": 1.1737,
      "grad_norm": 7.855458736419678,
      "learning_rate": 2.2964646464646465e-05,
      "loss": 0.1924,
      "step": 5453
    },
    {
      "epoch": 1.1738,
      "grad_norm": 5.391030788421631,
      "learning_rate": 2.295959595959596e-05,
      "loss": 0.1354,
      "step": 5454
    },
    {
      "epoch": 1.1739,
      "grad_norm": 10.776924133300781,
      "learning_rate": 2.2954545454545457e-05,
      "loss": 0.2392,
      "step": 5455
    },
    {
      "epoch": 1.174,
      "grad_norm": 5.133842468261719,
      "learning_rate": 2.294949494949495e-05,
      "loss": 0.3066,
      "step": 5456
    },
    {
      "epoch": 1.1741,
      "grad_norm": 2.461775302886963,
      "learning_rate": 2.2944444444444446e-05,
      "loss": 0.1195,
      "step": 5457
    },
    {
      "epoch": 1.1742,
      "grad_norm": 1.926068902015686,
      "learning_rate": 2.2939393939393942e-05,
      "loss": 0.0397,
      "step": 5458
    },
    {
      "epoch": 1.1743000000000001,
      "grad_norm": 1.241310954093933,
      "learning_rate": 2.2934343434343434e-05,
      "loss": 0.0151,
      "step": 5459
    },
    {
      "epoch": 1.1743999999999999,
      "grad_norm": 1.0120223760604858,
      "learning_rate": 2.292929292929293e-05,
      "loss": 0.0168,
      "step": 5460
    },
    {
      "epoch": 1.1745,
      "grad_norm": 2.2319416999816895,
      "learning_rate": 2.2924242424242426e-05,
      "loss": 0.0242,
      "step": 5461
    },
    {
      "epoch": 1.1746,
      "grad_norm": 5.323766708374023,
      "learning_rate": 2.2919191919191922e-05,
      "loss": 0.6046,
      "step": 5462
    },
    {
      "epoch": 1.1747,
      "grad_norm": 2.4194130897521973,
      "learning_rate": 2.2914141414141415e-05,
      "loss": 0.088,
      "step": 5463
    },
    {
      "epoch": 1.1748,
      "grad_norm": 5.68917989730835,
      "learning_rate": 2.290909090909091e-05,
      "loss": 0.4306,
      "step": 5464
    },
    {
      "epoch": 1.1749,
      "grad_norm": 4.140972137451172,
      "learning_rate": 2.2904040404040407e-05,
      "loss": 0.0623,
      "step": 5465
    },
    {
      "epoch": 1.175,
      "grad_norm": 2.1993496417999268,
      "learning_rate": 2.28989898989899e-05,
      "loss": 0.0541,
      "step": 5466
    },
    {
      "epoch": 1.1751,
      "grad_norm": 7.270163059234619,
      "learning_rate": 2.2893939393939395e-05,
      "loss": 0.8447,
      "step": 5467
    },
    {
      "epoch": 1.1752,
      "grad_norm": 12.875811576843262,
      "learning_rate": 2.288888888888889e-05,
      "loss": 0.2,
      "step": 5468
    },
    {
      "epoch": 1.1753,
      "grad_norm": 1.255020022392273,
      "learning_rate": 2.2883838383838384e-05,
      "loss": 0.0276,
      "step": 5469
    },
    {
      "epoch": 1.1754,
      "grad_norm": 5.501597881317139,
      "learning_rate": 2.287878787878788e-05,
      "loss": 0.1273,
      "step": 5470
    },
    {
      "epoch": 1.1755,
      "grad_norm": 1.8658955097198486,
      "learning_rate": 2.2873737373737376e-05,
      "loss": 0.0332,
      "step": 5471
    },
    {
      "epoch": 1.1756,
      "grad_norm": 2.394376039505005,
      "learning_rate": 2.286868686868687e-05,
      "loss": 0.0734,
      "step": 5472
    },
    {
      "epoch": 1.1757,
      "grad_norm": 11.77702522277832,
      "learning_rate": 2.2863636363636365e-05,
      "loss": 1.328,
      "step": 5473
    },
    {
      "epoch": 1.1758,
      "grad_norm": 6.465887069702148,
      "learning_rate": 2.285858585858586e-05,
      "loss": 0.207,
      "step": 5474
    },
    {
      "epoch": 1.1759,
      "grad_norm": 2.3937857151031494,
      "learning_rate": 2.2853535353535353e-05,
      "loss": 0.0346,
      "step": 5475
    },
    {
      "epoch": 1.176,
      "grad_norm": 2.519479751586914,
      "learning_rate": 2.284848484848485e-05,
      "loss": 0.0766,
      "step": 5476
    },
    {
      "epoch": 1.1761,
      "grad_norm": 3.2292253971099854,
      "learning_rate": 2.2843434343434345e-05,
      "loss": 0.5435,
      "step": 5477
    },
    {
      "epoch": 1.1762,
      "grad_norm": 3.1198854446411133,
      "learning_rate": 2.2838383838383838e-05,
      "loss": 0.2529,
      "step": 5478
    },
    {
      "epoch": 1.1763,
      "grad_norm": 4.567643642425537,
      "learning_rate": 2.2833333333333334e-05,
      "loss": 0.0574,
      "step": 5479
    },
    {
      "epoch": 1.1764000000000001,
      "grad_norm": 1.9573894739151,
      "learning_rate": 2.282828282828283e-05,
      "loss": 0.0481,
      "step": 5480
    },
    {
      "epoch": 1.1764999999999999,
      "grad_norm": 2.960827350616455,
      "learning_rate": 2.2823232323232326e-05,
      "loss": 0.0747,
      "step": 5481
    },
    {
      "epoch": 1.1766,
      "grad_norm": 7.577508926391602,
      "learning_rate": 2.281818181818182e-05,
      "loss": 0.2276,
      "step": 5482
    },
    {
      "epoch": 1.1767,
      "grad_norm": 4.093561172485352,
      "learning_rate": 2.2813131313131314e-05,
      "loss": 0.3195,
      "step": 5483
    },
    {
      "epoch": 1.1768,
      "grad_norm": 1.9666657447814941,
      "learning_rate": 2.280808080808081e-05,
      "loss": 0.0481,
      "step": 5484
    },
    {
      "epoch": 1.1769,
      "grad_norm": 4.963584899902344,
      "learning_rate": 2.2803030303030303e-05,
      "loss": 0.2046,
      "step": 5485
    },
    {
      "epoch": 1.177,
      "grad_norm": 3.2093238830566406,
      "learning_rate": 2.27979797979798e-05,
      "loss": 0.4857,
      "step": 5486
    },
    {
      "epoch": 1.1771,
      "grad_norm": 0.1973022222518921,
      "learning_rate": 2.2792929292929295e-05,
      "loss": 0.0022,
      "step": 5487
    },
    {
      "epoch": 1.1772,
      "grad_norm": 13.212974548339844,
      "learning_rate": 2.2787878787878788e-05,
      "loss": 0.353,
      "step": 5488
    },
    {
      "epoch": 1.1773,
      "grad_norm": 5.174311637878418,
      "learning_rate": 2.2782828282828284e-05,
      "loss": 0.1311,
      "step": 5489
    },
    {
      "epoch": 1.1774,
      "grad_norm": 1.8702869415283203,
      "learning_rate": 2.277777777777778e-05,
      "loss": 0.019,
      "step": 5490
    },
    {
      "epoch": 1.1775,
      "grad_norm": 5.010149002075195,
      "learning_rate": 2.2772727272727272e-05,
      "loss": 1.0401,
      "step": 5491
    },
    {
      "epoch": 1.1776,
      "grad_norm": 1.856319546699524,
      "learning_rate": 2.2767676767676768e-05,
      "loss": 0.0351,
      "step": 5492
    },
    {
      "epoch": 1.1777,
      "grad_norm": 1.0499123334884644,
      "learning_rate": 2.2762626262626264e-05,
      "loss": 0.0111,
      "step": 5493
    },
    {
      "epoch": 1.1778,
      "grad_norm": 5.576166152954102,
      "learning_rate": 2.2757575757575757e-05,
      "loss": 0.86,
      "step": 5494
    },
    {
      "epoch": 1.1779,
      "grad_norm": 8.634578704833984,
      "learning_rate": 2.2752525252525253e-05,
      "loss": 0.3005,
      "step": 5495
    },
    {
      "epoch": 1.178,
      "grad_norm": 0.8643753528594971,
      "learning_rate": 2.274747474747475e-05,
      "loss": 0.0126,
      "step": 5496
    },
    {
      "epoch": 1.1781,
      "grad_norm": 5.7434401512146,
      "learning_rate": 2.274242424242424e-05,
      "loss": 0.0625,
      "step": 5497
    },
    {
      "epoch": 1.1782,
      "grad_norm": 8.718451499938965,
      "learning_rate": 2.2737373737373737e-05,
      "loss": 0.1208,
      "step": 5498
    },
    {
      "epoch": 1.1783,
      "grad_norm": 6.814505577087402,
      "learning_rate": 2.2732323232323233e-05,
      "loss": 0.3438,
      "step": 5499
    },
    {
      "epoch": 1.1784,
      "grad_norm": 5.126078128814697,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.0799,
      "step": 5500
    },
    {
      "epoch": 1.1785,
      "grad_norm": 6.136263847351074,
      "learning_rate": 2.2722222222222222e-05,
      "loss": 0.5165,
      "step": 5501
    },
    {
      "epoch": 1.1786,
      "grad_norm": 2.398171901702881,
      "learning_rate": 2.2717171717171718e-05,
      "loss": 0.0328,
      "step": 5502
    },
    {
      "epoch": 1.1787,
      "grad_norm": 1.5658929347991943,
      "learning_rate": 2.2712121212121214e-05,
      "loss": 0.9339,
      "step": 5503
    },
    {
      "epoch": 1.1788,
      "grad_norm": 4.710813999176025,
      "learning_rate": 2.2707070707070706e-05,
      "loss": 0.2491,
      "step": 5504
    },
    {
      "epoch": 1.1789,
      "grad_norm": 9.957258224487305,
      "learning_rate": 2.2702020202020202e-05,
      "loss": 0.8066,
      "step": 5505
    },
    {
      "epoch": 1.179,
      "grad_norm": 0.49851125478744507,
      "learning_rate": 2.26969696969697e-05,
      "loss": 0.006,
      "step": 5506
    },
    {
      "epoch": 1.1791,
      "grad_norm": 4.852683067321777,
      "learning_rate": 2.269191919191919e-05,
      "loss": 0.0986,
      "step": 5507
    },
    {
      "epoch": 1.1792,
      "grad_norm": 3.2113983631134033,
      "learning_rate": 2.2686868686868687e-05,
      "loss": 0.0386,
      "step": 5508
    },
    {
      "epoch": 1.1793,
      "grad_norm": 5.252051830291748,
      "learning_rate": 2.2681818181818183e-05,
      "loss": 0.1711,
      "step": 5509
    },
    {
      "epoch": 1.1794,
      "grad_norm": 29.281391143798828,
      "learning_rate": 2.2676767676767676e-05,
      "loss": 0.0741,
      "step": 5510
    },
    {
      "epoch": 1.1795,
      "grad_norm": 8.433541297912598,
      "learning_rate": 2.267171717171717e-05,
      "loss": 0.5923,
      "step": 5511
    },
    {
      "epoch": 1.1796,
      "grad_norm": 19.092605590820312,
      "learning_rate": 2.2666666666666668e-05,
      "loss": 0.5439,
      "step": 5512
    },
    {
      "epoch": 1.1797,
      "grad_norm": 2.097367763519287,
      "learning_rate": 2.266161616161616e-05,
      "loss": 0.0824,
      "step": 5513
    },
    {
      "epoch": 1.1798,
      "grad_norm": 1.1659408807754517,
      "learning_rate": 2.2656565656565656e-05,
      "loss": 0.0189,
      "step": 5514
    },
    {
      "epoch": 1.1799,
      "grad_norm": 3.2752835750579834,
      "learning_rate": 2.2651515151515152e-05,
      "loss": 0.1086,
      "step": 5515
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.7794580459594727,
      "learning_rate": 2.2646464646464645e-05,
      "loss": 0.1685,
      "step": 5516
    },
    {
      "epoch": 1.1801,
      "grad_norm": 16.491968154907227,
      "learning_rate": 2.264141414141414e-05,
      "loss": 0.3153,
      "step": 5517
    },
    {
      "epoch": 1.1802,
      "grad_norm": 1.0184228420257568,
      "learning_rate": 2.2636363636363637e-05,
      "loss": 0.0574,
      "step": 5518
    },
    {
      "epoch": 1.1803,
      "grad_norm": 5.214111804962158,
      "learning_rate": 2.263131313131313e-05,
      "loss": 0.5894,
      "step": 5519
    },
    {
      "epoch": 1.1804000000000001,
      "grad_norm": 4.049544811248779,
      "learning_rate": 2.262626262626263e-05,
      "loss": 0.1255,
      "step": 5520
    },
    {
      "epoch": 1.1804999999999999,
      "grad_norm": 3.0616867542266846,
      "learning_rate": 2.2621212121212125e-05,
      "loss": 0.2859,
      "step": 5521
    },
    {
      "epoch": 1.1806,
      "grad_norm": 2.177603006362915,
      "learning_rate": 2.2616161616161617e-05,
      "loss": 0.0675,
      "step": 5522
    },
    {
      "epoch": 1.1807,
      "grad_norm": 11.9568452835083,
      "learning_rate": 2.2611111111111113e-05,
      "loss": 0.3336,
      "step": 5523
    },
    {
      "epoch": 1.1808,
      "grad_norm": 1.3090922832489014,
      "learning_rate": 2.260606060606061e-05,
      "loss": 0.0664,
      "step": 5524
    },
    {
      "epoch": 1.1809,
      "grad_norm": 5.717720031738281,
      "learning_rate": 2.2601010101010102e-05,
      "loss": 0.7016,
      "step": 5525
    },
    {
      "epoch": 1.181,
      "grad_norm": 3.2413828372955322,
      "learning_rate": 2.2595959595959598e-05,
      "loss": 0.3083,
      "step": 5526
    },
    {
      "epoch": 1.1811,
      "grad_norm": 6.059289932250977,
      "learning_rate": 2.2590909090909094e-05,
      "loss": 0.3128,
      "step": 5527
    },
    {
      "epoch": 1.1812,
      "grad_norm": 11.106884956359863,
      "learning_rate": 2.2585858585858587e-05,
      "loss": 0.086,
      "step": 5528
    },
    {
      "epoch": 1.1813,
      "grad_norm": 4.468175411224365,
      "learning_rate": 2.2580808080808083e-05,
      "loss": 0.1213,
      "step": 5529
    },
    {
      "epoch": 1.1814,
      "grad_norm": 3.266491413116455,
      "learning_rate": 2.257575757575758e-05,
      "loss": 0.4777,
      "step": 5530
    },
    {
      "epoch": 1.1815,
      "grad_norm": 6.205183982849121,
      "learning_rate": 2.2570707070707074e-05,
      "loss": 0.3234,
      "step": 5531
    },
    {
      "epoch": 1.1816,
      "grad_norm": 7.715623378753662,
      "learning_rate": 2.2565656565656567e-05,
      "loss": 0.6719,
      "step": 5532
    },
    {
      "epoch": 1.1817,
      "grad_norm": 3.5136470794677734,
      "learning_rate": 2.2560606060606063e-05,
      "loss": 0.5406,
      "step": 5533
    },
    {
      "epoch": 1.1818,
      "grad_norm": 8.165285110473633,
      "learning_rate": 2.255555555555556e-05,
      "loss": 0.4795,
      "step": 5534
    },
    {
      "epoch": 1.1819,
      "grad_norm": 3.123204231262207,
      "learning_rate": 2.255050505050505e-05,
      "loss": 0.5787,
      "step": 5535
    },
    {
      "epoch": 1.182,
      "grad_norm": 10.053882598876953,
      "learning_rate": 2.2545454545454548e-05,
      "loss": 0.3575,
      "step": 5536
    },
    {
      "epoch": 1.1821,
      "grad_norm": 4.1254777908325195,
      "learning_rate": 2.2540404040404044e-05,
      "loss": 0.1806,
      "step": 5537
    },
    {
      "epoch": 1.1822,
      "grad_norm": 1.818729043006897,
      "learning_rate": 2.2535353535353536e-05,
      "loss": 0.0549,
      "step": 5538
    },
    {
      "epoch": 1.1823,
      "grad_norm": 6.363375186920166,
      "learning_rate": 2.2530303030303032e-05,
      "loss": 0.3413,
      "step": 5539
    },
    {
      "epoch": 1.1824,
      "grad_norm": 2.4641072750091553,
      "learning_rate": 2.2525252525252528e-05,
      "loss": 0.0721,
      "step": 5540
    },
    {
      "epoch": 1.1825,
      "grad_norm": 1.3191641569137573,
      "learning_rate": 2.252020202020202e-05,
      "loss": 0.0324,
      "step": 5541
    },
    {
      "epoch": 1.1826,
      "grad_norm": 5.008520603179932,
      "learning_rate": 2.2515151515151517e-05,
      "loss": 0.1261,
      "step": 5542
    },
    {
      "epoch": 1.1827,
      "grad_norm": 4.215297698974609,
      "learning_rate": 2.2510101010101013e-05,
      "loss": 0.4975,
      "step": 5543
    },
    {
      "epoch": 1.1828,
      "grad_norm": 5.413934230804443,
      "learning_rate": 2.2505050505050505e-05,
      "loss": 0.645,
      "step": 5544
    },
    {
      "epoch": 1.1829,
      "grad_norm": 1.8610676527023315,
      "learning_rate": 2.25e-05,
      "loss": 0.0485,
      "step": 5545
    },
    {
      "epoch": 1.183,
      "grad_norm": 0.4696890711784363,
      "learning_rate": 2.2494949494949497e-05,
      "loss": 0.0094,
      "step": 5546
    },
    {
      "epoch": 1.1831,
      "grad_norm": 8.307943344116211,
      "learning_rate": 2.248989898989899e-05,
      "loss": 0.1372,
      "step": 5547
    },
    {
      "epoch": 1.1832,
      "grad_norm": 4.2809343338012695,
      "learning_rate": 2.2484848484848486e-05,
      "loss": 0.3542,
      "step": 5548
    },
    {
      "epoch": 1.1833,
      "grad_norm": 2.512059211730957,
      "learning_rate": 2.2479797979797982e-05,
      "loss": 0.5191,
      "step": 5549
    },
    {
      "epoch": 1.1834,
      "grad_norm": 0.375430166721344,
      "learning_rate": 2.2474747474747475e-05,
      "loss": 0.0059,
      "step": 5550
    },
    {
      "epoch": 1.1835,
      "grad_norm": 0.4713049530982971,
      "learning_rate": 2.246969696969697e-05,
      "loss": 0.4464,
      "step": 5551
    },
    {
      "epoch": 1.1836,
      "grad_norm": 5.760817527770996,
      "learning_rate": 2.2464646464646467e-05,
      "loss": 0.2365,
      "step": 5552
    },
    {
      "epoch": 1.1837,
      "grad_norm": 3.353541612625122,
      "learning_rate": 2.2459595959595963e-05,
      "loss": 0.0754,
      "step": 5553
    },
    {
      "epoch": 1.1838,
      "grad_norm": 9.352603912353516,
      "learning_rate": 2.2454545454545455e-05,
      "loss": 0.1579,
      "step": 5554
    },
    {
      "epoch": 1.1839,
      "grad_norm": 2.2642011642456055,
      "learning_rate": 2.244949494949495e-05,
      "loss": 0.0597,
      "step": 5555
    },
    {
      "epoch": 1.184,
      "grad_norm": 3.60965633392334,
      "learning_rate": 2.2444444444444447e-05,
      "loss": 0.17,
      "step": 5556
    },
    {
      "epoch": 1.1841,
      "grad_norm": 5.372970104217529,
      "learning_rate": 2.243939393939394e-05,
      "loss": 0.2655,
      "step": 5557
    },
    {
      "epoch": 1.1842,
      "grad_norm": 4.373719215393066,
      "learning_rate": 2.2434343434343436e-05,
      "loss": 0.3766,
      "step": 5558
    },
    {
      "epoch": 1.1843,
      "grad_norm": 6.113659858703613,
      "learning_rate": 2.2429292929292932e-05,
      "loss": 0.3179,
      "step": 5559
    },
    {
      "epoch": 1.1844000000000001,
      "grad_norm": 4.93377161026001,
      "learning_rate": 2.2424242424242424e-05,
      "loss": 0.3479,
      "step": 5560
    },
    {
      "epoch": 1.1844999999999999,
      "grad_norm": 4.553586006164551,
      "learning_rate": 2.241919191919192e-05,
      "loss": 1.0237,
      "step": 5561
    },
    {
      "epoch": 1.1846,
      "grad_norm": 4.02022647857666,
      "learning_rate": 2.2414141414141416e-05,
      "loss": 0.0629,
      "step": 5562
    },
    {
      "epoch": 1.1847,
      "grad_norm": 14.138870239257812,
      "learning_rate": 2.240909090909091e-05,
      "loss": 0.7419,
      "step": 5563
    },
    {
      "epoch": 1.1848,
      "grad_norm": 3.1265485286712646,
      "learning_rate": 2.2404040404040405e-05,
      "loss": 0.271,
      "step": 5564
    },
    {
      "epoch": 1.1849,
      "grad_norm": 4.313718318939209,
      "learning_rate": 2.23989898989899e-05,
      "loss": 0.2847,
      "step": 5565
    },
    {
      "epoch": 1.185,
      "grad_norm": 5.99876070022583,
      "learning_rate": 2.2393939393939393e-05,
      "loss": 0.4652,
      "step": 5566
    },
    {
      "epoch": 1.1851,
      "grad_norm": 8.74134635925293,
      "learning_rate": 2.238888888888889e-05,
      "loss": 0.0647,
      "step": 5567
    },
    {
      "epoch": 1.1852,
      "grad_norm": 8.175097465515137,
      "learning_rate": 2.2383838383838385e-05,
      "loss": 0.2391,
      "step": 5568
    },
    {
      "epoch": 1.1853,
      "grad_norm": 4.0233259201049805,
      "learning_rate": 2.2378787878787878e-05,
      "loss": 0.0922,
      "step": 5569
    },
    {
      "epoch": 1.1854,
      "grad_norm": 4.03978157043457,
      "learning_rate": 2.2373737373737374e-05,
      "loss": 0.098,
      "step": 5570
    },
    {
      "epoch": 1.1855,
      "grad_norm": 5.57703161239624,
      "learning_rate": 2.236868686868687e-05,
      "loss": 0.2224,
      "step": 5571
    },
    {
      "epoch": 1.1856,
      "grad_norm": 4.531424522399902,
      "learning_rate": 2.2363636363636366e-05,
      "loss": 0.0446,
      "step": 5572
    },
    {
      "epoch": 1.1857,
      "grad_norm": 2.027592182159424,
      "learning_rate": 2.235858585858586e-05,
      "loss": 0.1051,
      "step": 5573
    },
    {
      "epoch": 1.1858,
      "grad_norm": 16.454866409301758,
      "learning_rate": 2.2353535353535355e-05,
      "loss": 0.3851,
      "step": 5574
    },
    {
      "epoch": 1.1859,
      "grad_norm": 3.510420799255371,
      "learning_rate": 2.234848484848485e-05,
      "loss": 0.9928,
      "step": 5575
    },
    {
      "epoch": 1.186,
      "grad_norm": 5.827929496765137,
      "learning_rate": 2.2343434343434343e-05,
      "loss": 0.0908,
      "step": 5576
    },
    {
      "epoch": 1.1861,
      "grad_norm": 3.196235418319702,
      "learning_rate": 2.233838383838384e-05,
      "loss": 0.1205,
      "step": 5577
    },
    {
      "epoch": 1.1862,
      "grad_norm": 9.307881355285645,
      "learning_rate": 2.2333333333333335e-05,
      "loss": 0.6306,
      "step": 5578
    },
    {
      "epoch": 1.1863,
      "grad_norm": 4.342923641204834,
      "learning_rate": 2.2328282828282828e-05,
      "loss": 0.2541,
      "step": 5579
    },
    {
      "epoch": 1.1864,
      "grad_norm": 3.8906960487365723,
      "learning_rate": 2.2323232323232324e-05,
      "loss": 0.2162,
      "step": 5580
    },
    {
      "epoch": 1.1865,
      "grad_norm": 3.82487154006958,
      "learning_rate": 2.231818181818182e-05,
      "loss": 0.5716,
      "step": 5581
    },
    {
      "epoch": 1.1865999999999999,
      "grad_norm": 3.168623208999634,
      "learning_rate": 2.2313131313131312e-05,
      "loss": 0.1661,
      "step": 5582
    },
    {
      "epoch": 1.1867,
      "grad_norm": 3.624419927597046,
      "learning_rate": 2.230808080808081e-05,
      "loss": 0.1233,
      "step": 5583
    },
    {
      "epoch": 1.1868,
      "grad_norm": 4.739067077636719,
      "learning_rate": 2.2303030303030304e-05,
      "loss": 0.0914,
      "step": 5584
    },
    {
      "epoch": 1.1869,
      "grad_norm": 15.362131118774414,
      "learning_rate": 2.2297979797979797e-05,
      "loss": 0.1213,
      "step": 5585
    },
    {
      "epoch": 1.187,
      "grad_norm": 3.5571491718292236,
      "learning_rate": 2.2292929292929293e-05,
      "loss": 0.0753,
      "step": 5586
    },
    {
      "epoch": 1.1871,
      "grad_norm": 4.905526638031006,
      "learning_rate": 2.228787878787879e-05,
      "loss": 0.3406,
      "step": 5587
    },
    {
      "epoch": 1.1872,
      "grad_norm": 2.714531660079956,
      "learning_rate": 2.228282828282828e-05,
      "loss": 0.6785,
      "step": 5588
    },
    {
      "epoch": 1.1873,
      "grad_norm": 6.412725448608398,
      "learning_rate": 2.2277777777777778e-05,
      "loss": 0.2035,
      "step": 5589
    },
    {
      "epoch": 1.1874,
      "grad_norm": 6.4831438064575195,
      "learning_rate": 2.2272727272727274e-05,
      "loss": 0.0849,
      "step": 5590
    },
    {
      "epoch": 1.1875,
      "grad_norm": 5.74017333984375,
      "learning_rate": 2.2267676767676766e-05,
      "loss": 0.1335,
      "step": 5591
    },
    {
      "epoch": 1.1876,
      "grad_norm": 3.1012794971466064,
      "learning_rate": 2.2262626262626262e-05,
      "loss": 0.1429,
      "step": 5592
    },
    {
      "epoch": 1.1877,
      "grad_norm": 3.0264296531677246,
      "learning_rate": 2.2257575757575758e-05,
      "loss": 0.0839,
      "step": 5593
    },
    {
      "epoch": 1.1878,
      "grad_norm": 2.632240056991577,
      "learning_rate": 2.2252525252525254e-05,
      "loss": 0.2923,
      "step": 5594
    },
    {
      "epoch": 1.1879,
      "grad_norm": 3.0837647914886475,
      "learning_rate": 2.2247474747474747e-05,
      "loss": 0.0675,
      "step": 5595
    },
    {
      "epoch": 1.188,
      "grad_norm": 4.4922285079956055,
      "learning_rate": 2.2242424242424243e-05,
      "loss": 0.18,
      "step": 5596
    },
    {
      "epoch": 1.1881,
      "grad_norm": 9.81891918182373,
      "learning_rate": 2.223737373737374e-05,
      "loss": 0.6841,
      "step": 5597
    },
    {
      "epoch": 1.1882,
      "grad_norm": 4.090487957000732,
      "learning_rate": 2.223232323232323e-05,
      "loss": 1.0245,
      "step": 5598
    },
    {
      "epoch": 1.1883,
      "grad_norm": 2.0478343963623047,
      "learning_rate": 2.2227272727272727e-05,
      "loss": 0.1067,
      "step": 5599
    },
    {
      "epoch": 1.1884000000000001,
      "grad_norm": 3.3991332054138184,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 1.0274,
      "step": 5600
    },
    {
      "epoch": 1.1885,
      "grad_norm": 5.977842330932617,
      "learning_rate": 2.2217171717171716e-05,
      "loss": 0.2708,
      "step": 5601
    },
    {
      "epoch": 1.1886,
      "grad_norm": 216.8823699951172,
      "learning_rate": 2.2212121212121212e-05,
      "loss": 0.1737,
      "step": 5602
    },
    {
      "epoch": 1.1887,
      "grad_norm": 2.101463794708252,
      "learning_rate": 2.2207070707070708e-05,
      "loss": 0.5033,
      "step": 5603
    },
    {
      "epoch": 1.1888,
      "grad_norm": 6.372531890869141,
      "learning_rate": 2.22020202020202e-05,
      "loss": 0.14,
      "step": 5604
    },
    {
      "epoch": 1.1889,
      "grad_norm": 1.2594448328018188,
      "learning_rate": 2.21969696969697e-05,
      "loss": 0.0324,
      "step": 5605
    },
    {
      "epoch": 1.189,
      "grad_norm": 3.6798808574676514,
      "learning_rate": 2.2191919191919196e-05,
      "loss": 0.6834,
      "step": 5606
    },
    {
      "epoch": 1.1891,
      "grad_norm": 7.009540557861328,
      "learning_rate": 2.218686868686869e-05,
      "loss": 0.1281,
      "step": 5607
    },
    {
      "epoch": 1.1892,
      "grad_norm": 0.8522012233734131,
      "learning_rate": 2.2181818181818184e-05,
      "loss": 0.0183,
      "step": 5608
    },
    {
      "epoch": 1.1893,
      "grad_norm": 6.9969563484191895,
      "learning_rate": 2.217676767676768e-05,
      "loss": 0.6495,
      "step": 5609
    },
    {
      "epoch": 1.1894,
      "grad_norm": 2.0828959941864014,
      "learning_rate": 2.2171717171717173e-05,
      "loss": 0.0505,
      "step": 5610
    },
    {
      "epoch": 1.1895,
      "grad_norm": 2.1739695072174072,
      "learning_rate": 2.216666666666667e-05,
      "loss": 0.2467,
      "step": 5611
    },
    {
      "epoch": 1.1896,
      "grad_norm": 1.6530499458312988,
      "learning_rate": 2.2161616161616165e-05,
      "loss": 0.9333,
      "step": 5612
    },
    {
      "epoch": 1.1897,
      "grad_norm": 8.889896392822266,
      "learning_rate": 2.2156565656565658e-05,
      "loss": 0.8195,
      "step": 5613
    },
    {
      "epoch": 1.1898,
      "grad_norm": 4.742764472961426,
      "learning_rate": 2.2151515151515154e-05,
      "loss": 0.1607,
      "step": 5614
    },
    {
      "epoch": 1.1899,
      "grad_norm": 2.6573805809020996,
      "learning_rate": 2.214646464646465e-05,
      "loss": 0.0854,
      "step": 5615
    },
    {
      "epoch": 1.19,
      "grad_norm": 5.308388710021973,
      "learning_rate": 2.2141414141414142e-05,
      "loss": 0.1959,
      "step": 5616
    },
    {
      "epoch": 1.1901,
      "grad_norm": 2.6553144454956055,
      "learning_rate": 2.2136363636363638e-05,
      "loss": 0.2238,
      "step": 5617
    },
    {
      "epoch": 1.1902,
      "grad_norm": 3.644357681274414,
      "learning_rate": 2.2131313131313134e-05,
      "loss": 0.3008,
      "step": 5618
    },
    {
      "epoch": 1.1903,
      "grad_norm": 1.5676982402801514,
      "learning_rate": 2.2126262626262627e-05,
      "loss": 0.0903,
      "step": 5619
    },
    {
      "epoch": 1.1904,
      "grad_norm": 4.264885902404785,
      "learning_rate": 2.2121212121212123e-05,
      "loss": 0.2701,
      "step": 5620
    },
    {
      "epoch": 1.1905000000000001,
      "grad_norm": 0.7077184319496155,
      "learning_rate": 2.211616161616162e-05,
      "loss": 0.0095,
      "step": 5621
    },
    {
      "epoch": 1.1905999999999999,
      "grad_norm": 3.0698862075805664,
      "learning_rate": 2.211111111111111e-05,
      "loss": 0.2892,
      "step": 5622
    },
    {
      "epoch": 1.1907,
      "grad_norm": 2.7282440662384033,
      "learning_rate": 2.2106060606060607e-05,
      "loss": 0.0605,
      "step": 5623
    },
    {
      "epoch": 1.1908,
      "grad_norm": 4.026416778564453,
      "learning_rate": 2.2101010101010103e-05,
      "loss": 0.1074,
      "step": 5624
    },
    {
      "epoch": 1.1909,
      "grad_norm": 1.633631706237793,
      "learning_rate": 2.20959595959596e-05,
      "loss": 0.4676,
      "step": 5625
    },
    {
      "epoch": 1.191,
      "grad_norm": 4.117895126342773,
      "learning_rate": 2.2090909090909092e-05,
      "loss": 0.531,
      "step": 5626
    },
    {
      "epoch": 1.1911,
      "grad_norm": 4.2167792320251465,
      "learning_rate": 2.2085858585858588e-05,
      "loss": 0.132,
      "step": 5627
    },
    {
      "epoch": 1.1912,
      "grad_norm": 8.46526050567627,
      "learning_rate": 2.2080808080808084e-05,
      "loss": 0.2065,
      "step": 5628
    },
    {
      "epoch": 1.1913,
      "grad_norm": 2.044264316558838,
      "learning_rate": 2.2075757575757577e-05,
      "loss": 0.0464,
      "step": 5629
    },
    {
      "epoch": 1.1914,
      "grad_norm": 5.4067254066467285,
      "learning_rate": 2.2070707070707073e-05,
      "loss": 0.2265,
      "step": 5630
    },
    {
      "epoch": 1.1915,
      "grad_norm": 3.322127103805542,
      "learning_rate": 2.206565656565657e-05,
      "loss": 0.2575,
      "step": 5631
    },
    {
      "epoch": 1.1916,
      "grad_norm": 13.118157386779785,
      "learning_rate": 2.206060606060606e-05,
      "loss": 0.6708,
      "step": 5632
    },
    {
      "epoch": 1.1917,
      "grad_norm": 5.002882957458496,
      "learning_rate": 2.2055555555555557e-05,
      "loss": 0.1251,
      "step": 5633
    },
    {
      "epoch": 1.1918,
      "grad_norm": 4.195161819458008,
      "learning_rate": 2.2050505050505053e-05,
      "loss": 0.2161,
      "step": 5634
    },
    {
      "epoch": 1.1919,
      "grad_norm": 9.178245544433594,
      "learning_rate": 2.2045454545454546e-05,
      "loss": 0.5393,
      "step": 5635
    },
    {
      "epoch": 1.192,
      "grad_norm": 7.977066993713379,
      "learning_rate": 2.204040404040404e-05,
      "loss": 0.3168,
      "step": 5636
    },
    {
      "epoch": 1.1921,
      "grad_norm": 2.037266254425049,
      "learning_rate": 2.2035353535353538e-05,
      "loss": 0.2258,
      "step": 5637
    },
    {
      "epoch": 1.1922,
      "grad_norm": 20.108612060546875,
      "learning_rate": 2.203030303030303e-05,
      "loss": 0.325,
      "step": 5638
    },
    {
      "epoch": 1.1923,
      "grad_norm": 5.589395999908447,
      "learning_rate": 2.2025252525252526e-05,
      "loss": 0.3343,
      "step": 5639
    },
    {
      "epoch": 1.1924,
      "grad_norm": 2.5551164150238037,
      "learning_rate": 2.2020202020202022e-05,
      "loss": 0.0329,
      "step": 5640
    },
    {
      "epoch": 1.1925,
      "grad_norm": 3.6102967262268066,
      "learning_rate": 2.2015151515151515e-05,
      "loss": 0.4856,
      "step": 5641
    },
    {
      "epoch": 1.1926,
      "grad_norm": 1.8273823261260986,
      "learning_rate": 2.201010101010101e-05,
      "loss": 0.0399,
      "step": 5642
    },
    {
      "epoch": 1.1927,
      "grad_norm": 7.6746320724487305,
      "learning_rate": 2.2005050505050507e-05,
      "loss": 0.1785,
      "step": 5643
    },
    {
      "epoch": 1.1928,
      "grad_norm": 1.0512155294418335,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.0292,
      "step": 5644
    },
    {
      "epoch": 1.1929,
      "grad_norm": 9.612520217895508,
      "learning_rate": 2.1994949494949495e-05,
      "loss": 0.4343,
      "step": 5645
    },
    {
      "epoch": 1.193,
      "grad_norm": 1.5963929891586304,
      "learning_rate": 2.198989898989899e-05,
      "loss": 0.0618,
      "step": 5646
    },
    {
      "epoch": 1.1931,
      "grad_norm": 1.8238743543624878,
      "learning_rate": 2.1984848484848487e-05,
      "loss": 0.4716,
      "step": 5647
    },
    {
      "epoch": 1.1932,
      "grad_norm": 6.539017200469971,
      "learning_rate": 2.197979797979798e-05,
      "loss": 0.1281,
      "step": 5648
    },
    {
      "epoch": 1.1933,
      "grad_norm": 5.852899074554443,
      "learning_rate": 2.1974747474747476e-05,
      "loss": 0.3861,
      "step": 5649
    },
    {
      "epoch": 1.1934,
      "grad_norm": 2.2278225421905518,
      "learning_rate": 2.1969696969696972e-05,
      "loss": 0.0406,
      "step": 5650
    },
    {
      "epoch": 1.1935,
      "grad_norm": 7.990543842315674,
      "learning_rate": 2.1964646464646465e-05,
      "loss": 0.2852,
      "step": 5651
    },
    {
      "epoch": 1.1936,
      "grad_norm": 6.505054950714111,
      "learning_rate": 2.195959595959596e-05,
      "loss": 0.2179,
      "step": 5652
    },
    {
      "epoch": 1.1937,
      "grad_norm": 4.239529609680176,
      "learning_rate": 2.1954545454545457e-05,
      "loss": 0.0729,
      "step": 5653
    },
    {
      "epoch": 1.1938,
      "grad_norm": 9.171212196350098,
      "learning_rate": 2.194949494949495e-05,
      "loss": 0.2544,
      "step": 5654
    },
    {
      "epoch": 1.1939,
      "grad_norm": 3.1213865280151367,
      "learning_rate": 2.1944444444444445e-05,
      "loss": 0.1126,
      "step": 5655
    },
    {
      "epoch": 1.194,
      "grad_norm": 1.5524208545684814,
      "learning_rate": 2.193939393939394e-05,
      "loss": 0.0223,
      "step": 5656
    },
    {
      "epoch": 1.1941,
      "grad_norm": 3.087181329727173,
      "learning_rate": 2.1934343434343434e-05,
      "loss": 0.2799,
      "step": 5657
    },
    {
      "epoch": 1.1942,
      "grad_norm": 14.296469688415527,
      "learning_rate": 2.192929292929293e-05,
      "loss": 0.3365,
      "step": 5658
    },
    {
      "epoch": 1.1943,
      "grad_norm": 16.36405372619629,
      "learning_rate": 2.1924242424242426e-05,
      "loss": 0.3026,
      "step": 5659
    },
    {
      "epoch": 1.1944,
      "grad_norm": 3.291428804397583,
      "learning_rate": 2.191919191919192e-05,
      "loss": 0.4893,
      "step": 5660
    },
    {
      "epoch": 1.1945000000000001,
      "grad_norm": 10.810888290405273,
      "learning_rate": 2.1914141414141414e-05,
      "loss": 0.6,
      "step": 5661
    },
    {
      "epoch": 1.1945999999999999,
      "grad_norm": 4.689448833465576,
      "learning_rate": 2.190909090909091e-05,
      "loss": 0.1923,
      "step": 5662
    },
    {
      "epoch": 1.1947,
      "grad_norm": 1.8875006437301636,
      "learning_rate": 2.1904040404040403e-05,
      "loss": 0.0307,
      "step": 5663
    },
    {
      "epoch": 1.1948,
      "grad_norm": 4.1900410652160645,
      "learning_rate": 2.18989898989899e-05,
      "loss": 0.517,
      "step": 5664
    },
    {
      "epoch": 1.1949,
      "grad_norm": 5.840854644775391,
      "learning_rate": 2.1893939393939395e-05,
      "loss": 0.217,
      "step": 5665
    },
    {
      "epoch": 1.195,
      "grad_norm": 10.667098999023438,
      "learning_rate": 2.188888888888889e-05,
      "loss": 0.0306,
      "step": 5666
    },
    {
      "epoch": 1.1951,
      "grad_norm": 1.5408002138137817,
      "learning_rate": 2.1883838383838383e-05,
      "loss": 0.0648,
      "step": 5667
    },
    {
      "epoch": 1.1952,
      "grad_norm": 6.812074661254883,
      "learning_rate": 2.187878787878788e-05,
      "loss": 0.1147,
      "step": 5668
    },
    {
      "epoch": 1.1953,
      "grad_norm": 9.726968765258789,
      "learning_rate": 2.1873737373737375e-05,
      "loss": 0.7857,
      "step": 5669
    },
    {
      "epoch": 1.1954,
      "grad_norm": 3.7576868534088135,
      "learning_rate": 2.1868686868686868e-05,
      "loss": 0.2808,
      "step": 5670
    },
    {
      "epoch": 1.1955,
      "grad_norm": 3.532992124557495,
      "learning_rate": 2.1863636363636364e-05,
      "loss": 0.164,
      "step": 5671
    },
    {
      "epoch": 1.1956,
      "grad_norm": 10.509136199951172,
      "learning_rate": 2.185858585858586e-05,
      "loss": 0.2256,
      "step": 5672
    },
    {
      "epoch": 1.1957,
      "grad_norm": 1.7927557229995728,
      "learning_rate": 2.1853535353535353e-05,
      "loss": 0.0616,
      "step": 5673
    },
    {
      "epoch": 1.1958,
      "grad_norm": 2.677112340927124,
      "learning_rate": 2.184848484848485e-05,
      "loss": 0.0891,
      "step": 5674
    },
    {
      "epoch": 1.1959,
      "grad_norm": 11.57112979888916,
      "learning_rate": 2.1843434343434345e-05,
      "loss": 0.1712,
      "step": 5675
    },
    {
      "epoch": 1.196,
      "grad_norm": 11.668830871582031,
      "learning_rate": 2.1838383838383837e-05,
      "loss": 0.1524,
      "step": 5676
    },
    {
      "epoch": 1.1961,
      "grad_norm": 24.22443389892578,
      "learning_rate": 2.1833333333333333e-05,
      "loss": 0.1644,
      "step": 5677
    },
    {
      "epoch": 1.1962,
      "grad_norm": 10.687227249145508,
      "learning_rate": 2.182828282828283e-05,
      "loss": 0.8965,
      "step": 5678
    },
    {
      "epoch": 1.1963,
      "grad_norm": 3.1234381198883057,
      "learning_rate": 2.1823232323232322e-05,
      "loss": 0.1027,
      "step": 5679
    },
    {
      "epoch": 1.1964,
      "grad_norm": 1.8171344995498657,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 0.0184,
      "step": 5680
    },
    {
      "epoch": 1.1965,
      "grad_norm": 4.630583763122559,
      "learning_rate": 2.1813131313131314e-05,
      "loss": 0.3769,
      "step": 5681
    },
    {
      "epoch": 1.1966,
      "grad_norm": 3.12878680229187,
      "learning_rate": 2.1808080808080806e-05,
      "loss": 0.2539,
      "step": 5682
    },
    {
      "epoch": 1.1967,
      "grad_norm": 11.214635848999023,
      "learning_rate": 2.1803030303030302e-05,
      "loss": 0.2648,
      "step": 5683
    },
    {
      "epoch": 1.1968,
      "grad_norm": 11.309288024902344,
      "learning_rate": 2.17979797979798e-05,
      "loss": 0.0935,
      "step": 5684
    },
    {
      "epoch": 1.1969,
      "grad_norm": 2.4725537300109863,
      "learning_rate": 2.1792929292929294e-05,
      "loss": 0.0764,
      "step": 5685
    },
    {
      "epoch": 1.197,
      "grad_norm": 1.8507133722305298,
      "learning_rate": 2.1787878787878787e-05,
      "loss": 0.1355,
      "step": 5686
    },
    {
      "epoch": 1.1971,
      "grad_norm": 3.0418875217437744,
      "learning_rate": 2.1782828282828283e-05,
      "loss": 0.1498,
      "step": 5687
    },
    {
      "epoch": 1.1972,
      "grad_norm": 3.108677387237549,
      "learning_rate": 2.177777777777778e-05,
      "loss": 0.0609,
      "step": 5688
    },
    {
      "epoch": 1.1973,
      "grad_norm": 3.5423288345336914,
      "learning_rate": 2.177272727272727e-05,
      "loss": 0.0778,
      "step": 5689
    },
    {
      "epoch": 1.1974,
      "grad_norm": 4.017395496368408,
      "learning_rate": 2.1767676767676768e-05,
      "loss": 0.239,
      "step": 5690
    },
    {
      "epoch": 1.1975,
      "grad_norm": 3.436821460723877,
      "learning_rate": 2.1762626262626264e-05,
      "loss": 0.2595,
      "step": 5691
    },
    {
      "epoch": 1.1976,
      "grad_norm": 8.91564655303955,
      "learning_rate": 2.175757575757576e-05,
      "loss": 0.3884,
      "step": 5692
    },
    {
      "epoch": 1.1977,
      "grad_norm": 2.4039409160614014,
      "learning_rate": 2.1752525252525256e-05,
      "loss": 0.0955,
      "step": 5693
    },
    {
      "epoch": 1.1978,
      "grad_norm": 4.528891086578369,
      "learning_rate": 2.1747474747474748e-05,
      "loss": 0.1886,
      "step": 5694
    },
    {
      "epoch": 1.1979,
      "grad_norm": 3.324566125869751,
      "learning_rate": 2.1742424242424244e-05,
      "loss": 0.1255,
      "step": 5695
    },
    {
      "epoch": 1.198,
      "grad_norm": 2.00382399559021,
      "learning_rate": 2.173737373737374e-05,
      "loss": 0.0521,
      "step": 5696
    },
    {
      "epoch": 1.1981,
      "grad_norm": 4.057766914367676,
      "learning_rate": 2.1732323232323236e-05,
      "loss": 0.0348,
      "step": 5697
    },
    {
      "epoch": 1.1982,
      "grad_norm": 3.5031261444091797,
      "learning_rate": 2.172727272727273e-05,
      "loss": 0.0927,
      "step": 5698
    },
    {
      "epoch": 1.1983,
      "grad_norm": 10.86119270324707,
      "learning_rate": 2.1722222222222225e-05,
      "loss": 0.2651,
      "step": 5699
    },
    {
      "epoch": 1.1984,
      "grad_norm": 4.3580217361450195,
      "learning_rate": 2.171717171717172e-05,
      "loss": 0.1059,
      "step": 5700
    },
    {
      "epoch": 1.1985000000000001,
      "grad_norm": 5.393424987792969,
      "learning_rate": 2.1712121212121213e-05,
      "loss": 0.4344,
      "step": 5701
    },
    {
      "epoch": 1.1985999999999999,
      "grad_norm": 1.2285202741622925,
      "learning_rate": 2.170707070707071e-05,
      "loss": 0.0262,
      "step": 5702
    },
    {
      "epoch": 1.1987,
      "grad_norm": 25.418474197387695,
      "learning_rate": 2.1702020202020205e-05,
      "loss": 0.4679,
      "step": 5703
    },
    {
      "epoch": 1.1988,
      "grad_norm": 6.4058332443237305,
      "learning_rate": 2.1696969696969698e-05,
      "loss": 0.129,
      "step": 5704
    },
    {
      "epoch": 1.1989,
      "grad_norm": 1.3331456184387207,
      "learning_rate": 2.1691919191919194e-05,
      "loss": 0.0185,
      "step": 5705
    },
    {
      "epoch": 1.199,
      "grad_norm": 4.051408290863037,
      "learning_rate": 2.168686868686869e-05,
      "loss": 0.1219,
      "step": 5706
    },
    {
      "epoch": 1.1991,
      "grad_norm": 0.9846377968788147,
      "learning_rate": 2.1681818181818182e-05,
      "loss": 0.0166,
      "step": 5707
    },
    {
      "epoch": 1.1992,
      "grad_norm": 4.430857181549072,
      "learning_rate": 2.167676767676768e-05,
      "loss": 0.1512,
      "step": 5708
    },
    {
      "epoch": 1.1993,
      "grad_norm": 5.190492153167725,
      "learning_rate": 2.1671717171717174e-05,
      "loss": 0.1611,
      "step": 5709
    },
    {
      "epoch": 1.1994,
      "grad_norm": 7.633662700653076,
      "learning_rate": 2.1666666666666667e-05,
      "loss": 0.183,
      "step": 5710
    },
    {
      "epoch": 1.1995,
      "grad_norm": 2.329789161682129,
      "learning_rate": 2.1661616161616163e-05,
      "loss": 0.2257,
      "step": 5711
    },
    {
      "epoch": 1.1996,
      "grad_norm": 3.8132338523864746,
      "learning_rate": 2.165656565656566e-05,
      "loss": 0.0878,
      "step": 5712
    },
    {
      "epoch": 1.1997,
      "grad_norm": 2.6754238605499268,
      "learning_rate": 2.165151515151515e-05,
      "loss": 0.085,
      "step": 5713
    },
    {
      "epoch": 1.1998,
      "grad_norm": 1.9146090745925903,
      "learning_rate": 2.1646464646464648e-05,
      "loss": 0.102,
      "step": 5714
    },
    {
      "epoch": 1.1999,
      "grad_norm": 2.9260735511779785,
      "learning_rate": 2.1641414141414144e-05,
      "loss": 0.2213,
      "step": 5715
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.57740306854248,
      "learning_rate": 2.163636363636364e-05,
      "loss": 0.1404,
      "step": 5716
    },
    {
      "epoch": 1.2001,
      "grad_norm": 5.868307113647461,
      "learning_rate": 2.1631313131313132e-05,
      "loss": 0.0704,
      "step": 5717
    },
    {
      "epoch": 1.2002,
      "grad_norm": 11.108161926269531,
      "learning_rate": 2.1626262626262628e-05,
      "loss": 0.4145,
      "step": 5718
    },
    {
      "epoch": 1.2003,
      "grad_norm": 7.3176140785217285,
      "learning_rate": 2.1621212121212124e-05,
      "loss": 0.4665,
      "step": 5719
    },
    {
      "epoch": 1.2004,
      "grad_norm": 5.788232326507568,
      "learning_rate": 2.1616161616161617e-05,
      "loss": 0.2265,
      "step": 5720
    },
    {
      "epoch": 1.2005,
      "grad_norm": 6.699845790863037,
      "learning_rate": 2.1611111111111113e-05,
      "loss": 0.3717,
      "step": 5721
    },
    {
      "epoch": 1.2006000000000001,
      "grad_norm": 6.678947925567627,
      "learning_rate": 2.160606060606061e-05,
      "loss": 0.0748,
      "step": 5722
    },
    {
      "epoch": 1.2006999999999999,
      "grad_norm": 6.084568977355957,
      "learning_rate": 2.16010101010101e-05,
      "loss": 0.552,
      "step": 5723
    },
    {
      "epoch": 1.2008,
      "grad_norm": 9.960003852844238,
      "learning_rate": 2.1595959595959597e-05,
      "loss": 0.5963,
      "step": 5724
    },
    {
      "epoch": 1.2009,
      "grad_norm": 1.2865186929702759,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.0364,
      "step": 5725
    },
    {
      "epoch": 1.201,
      "grad_norm": 1.968029260635376,
      "learning_rate": 2.1585858585858586e-05,
      "loss": 0.0916,
      "step": 5726
    },
    {
      "epoch": 1.2011,
      "grad_norm": 4.343925952911377,
      "learning_rate": 2.1580808080808082e-05,
      "loss": 0.1951,
      "step": 5727
    },
    {
      "epoch": 1.2012,
      "grad_norm": 6.037591934204102,
      "learning_rate": 2.1575757575757578e-05,
      "loss": 0.1941,
      "step": 5728
    },
    {
      "epoch": 1.2013,
      "grad_norm": 0.9707487225532532,
      "learning_rate": 2.157070707070707e-05,
      "loss": 0.0207,
      "step": 5729
    },
    {
      "epoch": 1.2014,
      "grad_norm": 3.5143802165985107,
      "learning_rate": 2.1565656565656567e-05,
      "loss": 0.1569,
      "step": 5730
    },
    {
      "epoch": 1.2015,
      "grad_norm": 4.954933166503906,
      "learning_rate": 2.1560606060606062e-05,
      "loss": 0.1542,
      "step": 5731
    },
    {
      "epoch": 1.2016,
      "grad_norm": 2.6859195232391357,
      "learning_rate": 2.1555555555555555e-05,
      "loss": 0.0481,
      "step": 5732
    },
    {
      "epoch": 1.2017,
      "grad_norm": 7.025749206542969,
      "learning_rate": 2.155050505050505e-05,
      "loss": 0.0967,
      "step": 5733
    },
    {
      "epoch": 1.2018,
      "grad_norm": 1.6635081768035889,
      "learning_rate": 2.1545454545454547e-05,
      "loss": 0.0314,
      "step": 5734
    },
    {
      "epoch": 1.2019,
      "grad_norm": 4.162042140960693,
      "learning_rate": 2.154040404040404e-05,
      "loss": 0.1088,
      "step": 5735
    },
    {
      "epoch": 1.202,
      "grad_norm": 2.2159414291381836,
      "learning_rate": 2.1535353535353536e-05,
      "loss": 0.0553,
      "step": 5736
    },
    {
      "epoch": 1.2021,
      "grad_norm": 3.8033814430236816,
      "learning_rate": 2.153030303030303e-05,
      "loss": 0.1178,
      "step": 5737
    },
    {
      "epoch": 1.2022,
      "grad_norm": 14.179070472717285,
      "learning_rate": 2.1525252525252528e-05,
      "loss": 0.0909,
      "step": 5738
    },
    {
      "epoch": 1.2023,
      "grad_norm": 8.324136734008789,
      "learning_rate": 2.152020202020202e-05,
      "loss": 0.0767,
      "step": 5739
    },
    {
      "epoch": 1.2024,
      "grad_norm": 3.885216236114502,
      "learning_rate": 2.1515151515151516e-05,
      "loss": 0.1327,
      "step": 5740
    },
    {
      "epoch": 1.2025000000000001,
      "grad_norm": 15.047083854675293,
      "learning_rate": 2.1510101010101012e-05,
      "loss": 0.2173,
      "step": 5741
    },
    {
      "epoch": 1.2026,
      "grad_norm": 0.7033824324607849,
      "learning_rate": 2.1505050505050505e-05,
      "loss": 0.0104,
      "step": 5742
    },
    {
      "epoch": 1.2027,
      "grad_norm": 10.086771965026855,
      "learning_rate": 2.15e-05,
      "loss": 0.0719,
      "step": 5743
    },
    {
      "epoch": 1.2028,
      "grad_norm": 3.1360745429992676,
      "learning_rate": 2.1494949494949497e-05,
      "loss": 0.085,
      "step": 5744
    },
    {
      "epoch": 1.2029,
      "grad_norm": 2.2591755390167236,
      "learning_rate": 2.148989898989899e-05,
      "loss": 0.2116,
      "step": 5745
    },
    {
      "epoch": 1.203,
      "grad_norm": 4.2344770431518555,
      "learning_rate": 2.1484848484848485e-05,
      "loss": 0.3663,
      "step": 5746
    },
    {
      "epoch": 1.2031,
      "grad_norm": 4.118664264678955,
      "learning_rate": 2.147979797979798e-05,
      "loss": 0.121,
      "step": 5747
    },
    {
      "epoch": 1.2032,
      "grad_norm": 4.2492451667785645,
      "learning_rate": 2.1474747474747474e-05,
      "loss": 0.2473,
      "step": 5748
    },
    {
      "epoch": 1.2033,
      "grad_norm": 5.178040027618408,
      "learning_rate": 2.146969696969697e-05,
      "loss": 0.2778,
      "step": 5749
    },
    {
      "epoch": 1.2034,
      "grad_norm": 4.755954742431641,
      "learning_rate": 2.1464646464646466e-05,
      "loss": 0.0853,
      "step": 5750
    },
    {
      "epoch": 1.2035,
      "grad_norm": 1.9938116073608398,
      "learning_rate": 2.145959595959596e-05,
      "loss": 0.0345,
      "step": 5751
    },
    {
      "epoch": 1.2036,
      "grad_norm": 6.095796585083008,
      "learning_rate": 2.1454545454545455e-05,
      "loss": 0.1681,
      "step": 5752
    },
    {
      "epoch": 1.2037,
      "grad_norm": 5.324253082275391,
      "learning_rate": 2.144949494949495e-05,
      "loss": 0.1723,
      "step": 5753
    },
    {
      "epoch": 1.2038,
      "grad_norm": 3.6934702396392822,
      "learning_rate": 2.1444444444444443e-05,
      "loss": 0.2306,
      "step": 5754
    },
    {
      "epoch": 1.2039,
      "grad_norm": 3.0361757278442383,
      "learning_rate": 2.143939393939394e-05,
      "loss": 0.0972,
      "step": 5755
    },
    {
      "epoch": 1.204,
      "grad_norm": 6.981258869171143,
      "learning_rate": 2.1434343434343435e-05,
      "loss": 0.362,
      "step": 5756
    },
    {
      "epoch": 1.2041,
      "grad_norm": 5.387760162353516,
      "learning_rate": 2.142929292929293e-05,
      "loss": 1.0622,
      "step": 5757
    },
    {
      "epoch": 1.2042,
      "grad_norm": 3.0940914154052734,
      "learning_rate": 2.1424242424242424e-05,
      "loss": 0.0584,
      "step": 5758
    },
    {
      "epoch": 1.2043,
      "grad_norm": 2.512173652648926,
      "learning_rate": 2.141919191919192e-05,
      "loss": 0.058,
      "step": 5759
    },
    {
      "epoch": 1.2044,
      "grad_norm": 3.3135929107666016,
      "learning_rate": 2.1414141414141416e-05,
      "loss": 0.0809,
      "step": 5760
    },
    {
      "epoch": 1.2045,
      "grad_norm": 1.8544522523880005,
      "learning_rate": 2.140909090909091e-05,
      "loss": 0.0531,
      "step": 5761
    },
    {
      "epoch": 1.2046000000000001,
      "grad_norm": 1.8687036037445068,
      "learning_rate": 2.1404040404040404e-05,
      "loss": 0.0263,
      "step": 5762
    },
    {
      "epoch": 1.2046999999999999,
      "grad_norm": 2.0973334312438965,
      "learning_rate": 2.13989898989899e-05,
      "loss": 0.0991,
      "step": 5763
    },
    {
      "epoch": 1.2048,
      "grad_norm": 2.1435186862945557,
      "learning_rate": 2.1393939393939393e-05,
      "loss": 0.088,
      "step": 5764
    },
    {
      "epoch": 1.2049,
      "grad_norm": 3.968219041824341,
      "learning_rate": 2.138888888888889e-05,
      "loss": 0.1506,
      "step": 5765
    },
    {
      "epoch": 1.205,
      "grad_norm": 1.5658084154129028,
      "learning_rate": 2.1383838383838385e-05,
      "loss": 0.06,
      "step": 5766
    },
    {
      "epoch": 1.2051,
      "grad_norm": 6.766286849975586,
      "learning_rate": 2.1378787878787877e-05,
      "loss": 0.6582,
      "step": 5767
    },
    {
      "epoch": 1.2052,
      "grad_norm": 4.047379016876221,
      "learning_rate": 2.1373737373737373e-05,
      "loss": 0.108,
      "step": 5768
    },
    {
      "epoch": 1.2053,
      "grad_norm": 9.197892189025879,
      "learning_rate": 2.136868686868687e-05,
      "loss": 0.1746,
      "step": 5769
    },
    {
      "epoch": 1.2054,
      "grad_norm": 7.743758678436279,
      "learning_rate": 2.1363636363636362e-05,
      "loss": 0.4063,
      "step": 5770
    },
    {
      "epoch": 1.2055,
      "grad_norm": 5.171073913574219,
      "learning_rate": 2.1358585858585858e-05,
      "loss": 0.1525,
      "step": 5771
    },
    {
      "epoch": 1.2056,
      "grad_norm": 3.6900672912597656,
      "learning_rate": 2.1353535353535354e-05,
      "loss": 0.5993,
      "step": 5772
    },
    {
      "epoch": 1.2057,
      "grad_norm": 2.8603696823120117,
      "learning_rate": 2.1348484848484847e-05,
      "loss": 0.2465,
      "step": 5773
    },
    {
      "epoch": 1.2058,
      "grad_norm": 2.1559362411499023,
      "learning_rate": 2.1343434343434343e-05,
      "loss": 0.2202,
      "step": 5774
    },
    {
      "epoch": 1.2059,
      "grad_norm": 7.212629318237305,
      "learning_rate": 2.133838383838384e-05,
      "loss": 0.1909,
      "step": 5775
    },
    {
      "epoch": 1.206,
      "grad_norm": 2.9808449745178223,
      "learning_rate": 2.1333333333333335e-05,
      "loss": 0.101,
      "step": 5776
    },
    {
      "epoch": 1.2061,
      "grad_norm": 2.027611017227173,
      "learning_rate": 2.132828282828283e-05,
      "loss": 0.0981,
      "step": 5777
    },
    {
      "epoch": 1.2062,
      "grad_norm": 4.641350269317627,
      "learning_rate": 2.1323232323232327e-05,
      "loss": 0.1609,
      "step": 5778
    },
    {
      "epoch": 1.2063,
      "grad_norm": 3.1800355911254883,
      "learning_rate": 2.131818181818182e-05,
      "loss": 0.0763,
      "step": 5779
    },
    {
      "epoch": 1.2064,
      "grad_norm": 2.1619608402252197,
      "learning_rate": 2.1313131313131315e-05,
      "loss": 0.0415,
      "step": 5780
    },
    {
      "epoch": 1.2065,
      "grad_norm": 2.1540133953094482,
      "learning_rate": 2.130808080808081e-05,
      "loss": 0.2658,
      "step": 5781
    },
    {
      "epoch": 1.2066,
      "grad_norm": 2.780517339706421,
      "learning_rate": 2.1303030303030304e-05,
      "loss": 0.2485,
      "step": 5782
    },
    {
      "epoch": 1.2067,
      "grad_norm": 1.758703589439392,
      "learning_rate": 2.12979797979798e-05,
      "loss": 0.05,
      "step": 5783
    },
    {
      "epoch": 1.2068,
      "grad_norm": 7.517246723175049,
      "learning_rate": 2.1292929292929296e-05,
      "loss": 0.3317,
      "step": 5784
    },
    {
      "epoch": 1.2069,
      "grad_norm": 2.8880908489227295,
      "learning_rate": 2.128787878787879e-05,
      "loss": 0.2726,
      "step": 5785
    },
    {
      "epoch": 1.207,
      "grad_norm": 1.1278076171875,
      "learning_rate": 2.1282828282828284e-05,
      "loss": 0.0218,
      "step": 5786
    },
    {
      "epoch": 1.2071,
      "grad_norm": 11.403741836547852,
      "learning_rate": 2.127777777777778e-05,
      "loss": 0.2839,
      "step": 5787
    },
    {
      "epoch": 1.2072,
      "grad_norm": 3.706970453262329,
      "learning_rate": 2.1272727272727276e-05,
      "loss": 0.071,
      "step": 5788
    },
    {
      "epoch": 1.2073,
      "grad_norm": 4.069941997528076,
      "learning_rate": 2.126767676767677e-05,
      "loss": 0.1196,
      "step": 5789
    },
    {
      "epoch": 1.2074,
      "grad_norm": 4.365752220153809,
      "learning_rate": 2.1262626262626265e-05,
      "loss": 0.0953,
      "step": 5790
    },
    {
      "epoch": 1.2075,
      "grad_norm": 2.188363552093506,
      "learning_rate": 2.125757575757576e-05,
      "loss": 0.067,
      "step": 5791
    },
    {
      "epoch": 1.2076,
      "grad_norm": 2.1421658992767334,
      "learning_rate": 2.1252525252525254e-05,
      "loss": 0.0311,
      "step": 5792
    },
    {
      "epoch": 1.2077,
      "grad_norm": 5.134171485900879,
      "learning_rate": 2.124747474747475e-05,
      "loss": 0.112,
      "step": 5793
    },
    {
      "epoch": 1.2078,
      "grad_norm": 3.6131601333618164,
      "learning_rate": 2.1242424242424246e-05,
      "loss": 0.0543,
      "step": 5794
    },
    {
      "epoch": 1.2079,
      "grad_norm": 44.65406036376953,
      "learning_rate": 2.1237373737373738e-05,
      "loss": 0.184,
      "step": 5795
    },
    {
      "epoch": 1.208,
      "grad_norm": 4.162374019622803,
      "learning_rate": 2.1232323232323234e-05,
      "loss": 0.1463,
      "step": 5796
    },
    {
      "epoch": 1.2081,
      "grad_norm": 3.2122714519500732,
      "learning_rate": 2.122727272727273e-05,
      "loss": 0.0869,
      "step": 5797
    },
    {
      "epoch": 1.2082,
      "grad_norm": 5.420196533203125,
      "learning_rate": 2.1222222222222223e-05,
      "loss": 0.1383,
      "step": 5798
    },
    {
      "epoch": 1.2083,
      "grad_norm": 3.876117467880249,
      "learning_rate": 2.121717171717172e-05,
      "loss": 0.3075,
      "step": 5799
    },
    {
      "epoch": 1.2084,
      "grad_norm": 5.109684944152832,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.0971,
      "step": 5800
    },
    {
      "epoch": 1.2085,
      "grad_norm": 4.134227275848389,
      "learning_rate": 2.1207070707070707e-05,
      "loss": 0.1172,
      "step": 5801
    },
    {
      "epoch": 1.2086000000000001,
      "grad_norm": 3.1018683910369873,
      "learning_rate": 2.1202020202020203e-05,
      "loss": 0.061,
      "step": 5802
    },
    {
      "epoch": 1.2086999999999999,
      "grad_norm": 7.769604206085205,
      "learning_rate": 2.11969696969697e-05,
      "loss": 0.2643,
      "step": 5803
    },
    {
      "epoch": 1.2088,
      "grad_norm": 3.0058188438415527,
      "learning_rate": 2.1191919191919192e-05,
      "loss": 0.0558,
      "step": 5804
    },
    {
      "epoch": 1.2089,
      "grad_norm": 3.548701286315918,
      "learning_rate": 2.1186868686868688e-05,
      "loss": 0.3204,
      "step": 5805
    },
    {
      "epoch": 1.209,
      "grad_norm": 2.5077269077301025,
      "learning_rate": 2.1181818181818184e-05,
      "loss": 0.1235,
      "step": 5806
    },
    {
      "epoch": 1.2091,
      "grad_norm": 2.730156421661377,
      "learning_rate": 2.117676767676768e-05,
      "loss": 0.0868,
      "step": 5807
    },
    {
      "epoch": 1.2092,
      "grad_norm": 3.28835391998291,
      "learning_rate": 2.1171717171717172e-05,
      "loss": 0.2736,
      "step": 5808
    },
    {
      "epoch": 1.2093,
      "grad_norm": 8.879287719726562,
      "learning_rate": 2.116666666666667e-05,
      "loss": 0.4526,
      "step": 5809
    },
    {
      "epoch": 1.2094,
      "grad_norm": 6.481984615325928,
      "learning_rate": 2.1161616161616164e-05,
      "loss": 0.3441,
      "step": 5810
    },
    {
      "epoch": 1.2095,
      "grad_norm": 1.8123502731323242,
      "learning_rate": 2.1156565656565657e-05,
      "loss": 0.1921,
      "step": 5811
    },
    {
      "epoch": 1.2096,
      "grad_norm": 2.0720722675323486,
      "learning_rate": 2.1151515151515153e-05,
      "loss": 0.0928,
      "step": 5812
    },
    {
      "epoch": 1.2097,
      "grad_norm": 10.186979293823242,
      "learning_rate": 2.114646464646465e-05,
      "loss": 0.2926,
      "step": 5813
    },
    {
      "epoch": 1.2098,
      "grad_norm": 2.232257604598999,
      "learning_rate": 2.114141414141414e-05,
      "loss": 0.076,
      "step": 5814
    },
    {
      "epoch": 1.2099,
      "grad_norm": 5.01591682434082,
      "learning_rate": 2.1136363636363638e-05,
      "loss": 0.247,
      "step": 5815
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.464776039123535,
      "learning_rate": 2.1131313131313134e-05,
      "loss": 0.4825,
      "step": 5816
    },
    {
      "epoch": 1.2101,
      "grad_norm": 3.6082284450531006,
      "learning_rate": 2.1126262626262626e-05,
      "loss": 0.2987,
      "step": 5817
    },
    {
      "epoch": 1.2102,
      "grad_norm": 4.447847843170166,
      "learning_rate": 2.1121212121212122e-05,
      "loss": 0.2917,
      "step": 5818
    },
    {
      "epoch": 1.2103,
      "grad_norm": 1.262362003326416,
      "learning_rate": 2.1116161616161618e-05,
      "loss": 0.0158,
      "step": 5819
    },
    {
      "epoch": 1.2104,
      "grad_norm": 2.5513365268707275,
      "learning_rate": 2.111111111111111e-05,
      "loss": 0.1114,
      "step": 5820
    },
    {
      "epoch": 1.2105,
      "grad_norm": 3.2960288524627686,
      "learning_rate": 2.1106060606060607e-05,
      "loss": 0.0948,
      "step": 5821
    },
    {
      "epoch": 1.2106,
      "grad_norm": 2.4400415420532227,
      "learning_rate": 2.1101010101010103e-05,
      "loss": 0.0336,
      "step": 5822
    },
    {
      "epoch": 1.2107,
      "grad_norm": 3.122178077697754,
      "learning_rate": 2.1095959595959595e-05,
      "loss": 0.0928,
      "step": 5823
    },
    {
      "epoch": 1.2107999999999999,
      "grad_norm": 2.5114083290100098,
      "learning_rate": 2.109090909090909e-05,
      "loss": 0.1219,
      "step": 5824
    },
    {
      "epoch": 1.2109,
      "grad_norm": 3.0899882316589355,
      "learning_rate": 2.1085858585858587e-05,
      "loss": 0.0799,
      "step": 5825
    },
    {
      "epoch": 1.211,
      "grad_norm": 4.275166034698486,
      "learning_rate": 2.108080808080808e-05,
      "loss": 0.1577,
      "step": 5826
    },
    {
      "epoch": 1.2111,
      "grad_norm": 15.348180770874023,
      "learning_rate": 2.1075757575757576e-05,
      "loss": 0.307,
      "step": 5827
    },
    {
      "epoch": 1.2112,
      "grad_norm": 4.42926549911499,
      "learning_rate": 2.1070707070707072e-05,
      "loss": 0.1304,
      "step": 5828
    },
    {
      "epoch": 1.2113,
      "grad_norm": 5.688600540161133,
      "learning_rate": 2.1065656565656568e-05,
      "loss": 0.3227,
      "step": 5829
    },
    {
      "epoch": 1.2114,
      "grad_norm": 5.349828720092773,
      "learning_rate": 2.106060606060606e-05,
      "loss": 0.2161,
      "step": 5830
    },
    {
      "epoch": 1.2115,
      "grad_norm": 5.852518558502197,
      "learning_rate": 2.1055555555555556e-05,
      "loss": 0.0952,
      "step": 5831
    },
    {
      "epoch": 1.2116,
      "grad_norm": 7.535603046417236,
      "learning_rate": 2.1050505050505052e-05,
      "loss": 0.2131,
      "step": 5832
    },
    {
      "epoch": 1.2117,
      "grad_norm": 6.452699184417725,
      "learning_rate": 2.1045454545454545e-05,
      "loss": 0.319,
      "step": 5833
    },
    {
      "epoch": 1.2118,
      "grad_norm": 5.990747451782227,
      "learning_rate": 2.104040404040404e-05,
      "loss": 0.4189,
      "step": 5834
    },
    {
      "epoch": 1.2119,
      "grad_norm": 3.0741615295410156,
      "learning_rate": 2.1035353535353537e-05,
      "loss": 0.0903,
      "step": 5835
    },
    {
      "epoch": 1.212,
      "grad_norm": 6.970510959625244,
      "learning_rate": 2.103030303030303e-05,
      "loss": 0.1365,
      "step": 5836
    },
    {
      "epoch": 1.2121,
      "grad_norm": 1.7033504247665405,
      "learning_rate": 2.1025252525252526e-05,
      "loss": 0.0325,
      "step": 5837
    },
    {
      "epoch": 1.2122,
      "grad_norm": 2.0621438026428223,
      "learning_rate": 2.102020202020202e-05,
      "loss": 0.0493,
      "step": 5838
    },
    {
      "epoch": 1.2123,
      "grad_norm": 2.725093364715576,
      "learning_rate": 2.1015151515151514e-05,
      "loss": 0.2379,
      "step": 5839
    },
    {
      "epoch": 1.2124,
      "grad_norm": 5.955760955810547,
      "learning_rate": 2.101010101010101e-05,
      "loss": 0.4474,
      "step": 5840
    },
    {
      "epoch": 1.2125,
      "grad_norm": 4.785432815551758,
      "learning_rate": 2.1005050505050506e-05,
      "loss": 0.1503,
      "step": 5841
    },
    {
      "epoch": 1.2126000000000001,
      "grad_norm": 3.1583664417266846,
      "learning_rate": 2.1e-05,
      "loss": 0.1069,
      "step": 5842
    },
    {
      "epoch": 1.2127,
      "grad_norm": 5.606582164764404,
      "learning_rate": 2.0994949494949495e-05,
      "loss": 0.178,
      "step": 5843
    },
    {
      "epoch": 1.2128,
      "grad_norm": 6.8200273513793945,
      "learning_rate": 2.098989898989899e-05,
      "loss": 0.3105,
      "step": 5844
    },
    {
      "epoch": 1.2129,
      "grad_norm": 2.664614200592041,
      "learning_rate": 2.0984848484848483e-05,
      "loss": 0.1322,
      "step": 5845
    },
    {
      "epoch": 1.213,
      "grad_norm": 3.195796489715576,
      "learning_rate": 2.097979797979798e-05,
      "loss": 0.1283,
      "step": 5846
    },
    {
      "epoch": 1.2131,
      "grad_norm": 4.888011932373047,
      "learning_rate": 2.0974747474747475e-05,
      "loss": 0.4051,
      "step": 5847
    },
    {
      "epoch": 1.2132,
      "grad_norm": 3.012044668197632,
      "learning_rate": 2.096969696969697e-05,
      "loss": 0.056,
      "step": 5848
    },
    {
      "epoch": 1.2133,
      "grad_norm": 1.255158543586731,
      "learning_rate": 2.0964646464646464e-05,
      "loss": 0.0228,
      "step": 5849
    },
    {
      "epoch": 1.2134,
      "grad_norm": 6.871646881103516,
      "learning_rate": 2.095959595959596e-05,
      "loss": 0.1658,
      "step": 5850
    },
    {
      "epoch": 1.2135,
      "grad_norm": 4.2900919914245605,
      "learning_rate": 2.0954545454545456e-05,
      "loss": 0.1483,
      "step": 5851
    },
    {
      "epoch": 1.2136,
      "grad_norm": 3.0193068981170654,
      "learning_rate": 2.094949494949495e-05,
      "loss": 0.1286,
      "step": 5852
    },
    {
      "epoch": 1.2137,
      "grad_norm": 7.453155517578125,
      "learning_rate": 2.0944444444444445e-05,
      "loss": 0.1997,
      "step": 5853
    },
    {
      "epoch": 1.2138,
      "grad_norm": 7.240789413452148,
      "learning_rate": 2.093939393939394e-05,
      "loss": 0.4332,
      "step": 5854
    },
    {
      "epoch": 1.2139,
      "grad_norm": 2.649360418319702,
      "learning_rate": 2.0934343434343433e-05,
      "loss": 0.2225,
      "step": 5855
    },
    {
      "epoch": 1.214,
      "grad_norm": 2.101686477661133,
      "learning_rate": 2.092929292929293e-05,
      "loss": 0.0433,
      "step": 5856
    },
    {
      "epoch": 1.2141,
      "grad_norm": 16.639392852783203,
      "learning_rate": 2.0924242424242425e-05,
      "loss": 0.2686,
      "step": 5857
    },
    {
      "epoch": 1.2142,
      "grad_norm": 5.252928733825684,
      "learning_rate": 2.0919191919191918e-05,
      "loss": 0.1563,
      "step": 5858
    },
    {
      "epoch": 1.2143,
      "grad_norm": 5.314360618591309,
      "learning_rate": 2.0914141414141414e-05,
      "loss": 0.1772,
      "step": 5859
    },
    {
      "epoch": 1.2144,
      "grad_norm": 3.199838638305664,
      "learning_rate": 2.090909090909091e-05,
      "loss": 0.1658,
      "step": 5860
    },
    {
      "epoch": 1.2145,
      "grad_norm": 4.462727069854736,
      "learning_rate": 2.0904040404040406e-05,
      "loss": 0.1943,
      "step": 5861
    },
    {
      "epoch": 1.2146,
      "grad_norm": 5.093144416809082,
      "learning_rate": 2.0898989898989902e-05,
      "loss": 0.1042,
      "step": 5862
    },
    {
      "epoch": 1.2147000000000001,
      "grad_norm": 6.223134517669678,
      "learning_rate": 2.0893939393939398e-05,
      "loss": 0.1238,
      "step": 5863
    },
    {
      "epoch": 1.2147999999999999,
      "grad_norm": 5.2226104736328125,
      "learning_rate": 2.088888888888889e-05,
      "loss": 0.213,
      "step": 5864
    },
    {
      "epoch": 1.2149,
      "grad_norm": 3.952949047088623,
      "learning_rate": 2.0883838383838386e-05,
      "loss": 0.0731,
      "step": 5865
    },
    {
      "epoch": 1.215,
      "grad_norm": 9.218148231506348,
      "learning_rate": 2.0878787878787882e-05,
      "loss": 0.4859,
      "step": 5866
    },
    {
      "epoch": 1.2151,
      "grad_norm": 6.264392375946045,
      "learning_rate": 2.0873737373737375e-05,
      "loss": 0.1929,
      "step": 5867
    },
    {
      "epoch": 1.2152,
      "grad_norm": 3.596224784851074,
      "learning_rate": 2.086868686868687e-05,
      "loss": 0.1463,
      "step": 5868
    },
    {
      "epoch": 1.2153,
      "grad_norm": 5.597659111022949,
      "learning_rate": 2.0863636363636367e-05,
      "loss": 0.1992,
      "step": 5869
    },
    {
      "epoch": 1.2154,
      "grad_norm": 4.6832709312438965,
      "learning_rate": 2.085858585858586e-05,
      "loss": 0.2733,
      "step": 5870
    },
    {
      "epoch": 1.2155,
      "grad_norm": 5.037816047668457,
      "learning_rate": 2.0853535353535355e-05,
      "loss": 0.18,
      "step": 5871
    },
    {
      "epoch": 1.2156,
      "grad_norm": 9.545265197753906,
      "learning_rate": 2.084848484848485e-05,
      "loss": 0.0567,
      "step": 5872
    },
    {
      "epoch": 1.2157,
      "grad_norm": 1.4041467905044556,
      "learning_rate": 2.0843434343434344e-05,
      "loss": 0.0251,
      "step": 5873
    },
    {
      "epoch": 1.2158,
      "grad_norm": 3.9496750831604004,
      "learning_rate": 2.083838383838384e-05,
      "loss": 0.2081,
      "step": 5874
    },
    {
      "epoch": 1.2159,
      "grad_norm": 0.6769630312919617,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.005,
      "step": 5875
    },
    {
      "epoch": 1.216,
      "grad_norm": 3.9473376274108887,
      "learning_rate": 2.082828282828283e-05,
      "loss": 0.1131,
      "step": 5876
    },
    {
      "epoch": 1.2161,
      "grad_norm": 3.627176523208618,
      "learning_rate": 2.0823232323232325e-05,
      "loss": 0.1005,
      "step": 5877
    },
    {
      "epoch": 1.2162,
      "grad_norm": 3.2894773483276367,
      "learning_rate": 2.081818181818182e-05,
      "loss": 0.251,
      "step": 5878
    },
    {
      "epoch": 1.2163,
      "grad_norm": 2.41201114654541,
      "learning_rate": 2.0813131313131317e-05,
      "loss": 0.0492,
      "step": 5879
    },
    {
      "epoch": 1.2164,
      "grad_norm": 2.788662910461426,
      "learning_rate": 2.080808080808081e-05,
      "loss": 0.103,
      "step": 5880
    },
    {
      "epoch": 1.2165,
      "grad_norm": 1.9633915424346924,
      "learning_rate": 2.0803030303030305e-05,
      "loss": 0.0373,
      "step": 5881
    },
    {
      "epoch": 1.2166,
      "grad_norm": 4.46208381652832,
      "learning_rate": 2.07979797979798e-05,
      "loss": 0.0624,
      "step": 5882
    },
    {
      "epoch": 1.2167,
      "grad_norm": 4.065872669219971,
      "learning_rate": 2.0792929292929294e-05,
      "loss": 0.177,
      "step": 5883
    },
    {
      "epoch": 1.2168,
      "grad_norm": 3.7673563957214355,
      "learning_rate": 2.078787878787879e-05,
      "loss": 0.0306,
      "step": 5884
    },
    {
      "epoch": 1.2169,
      "grad_norm": 7.0911030769348145,
      "learning_rate": 2.0782828282828286e-05,
      "loss": 0.2333,
      "step": 5885
    },
    {
      "epoch": 1.217,
      "grad_norm": 4.278111457824707,
      "learning_rate": 2.077777777777778e-05,
      "loss": 0.1217,
      "step": 5886
    },
    {
      "epoch": 1.2171,
      "grad_norm": 3.063100814819336,
      "learning_rate": 2.0772727272727274e-05,
      "loss": 0.1224,
      "step": 5887
    },
    {
      "epoch": 1.2172,
      "grad_norm": 2.365896701812744,
      "learning_rate": 2.076767676767677e-05,
      "loss": 0.0568,
      "step": 5888
    },
    {
      "epoch": 1.2173,
      "grad_norm": 5.4366559982299805,
      "learning_rate": 2.0762626262626263e-05,
      "loss": 0.3386,
      "step": 5889
    },
    {
      "epoch": 1.2174,
      "grad_norm": 13.035219192504883,
      "learning_rate": 2.075757575757576e-05,
      "loss": 0.2503,
      "step": 5890
    },
    {
      "epoch": 1.2175,
      "grad_norm": 2.76212215423584,
      "learning_rate": 2.0752525252525255e-05,
      "loss": 0.1028,
      "step": 5891
    },
    {
      "epoch": 1.2176,
      "grad_norm": 9.283026695251465,
      "learning_rate": 2.0747474747474748e-05,
      "loss": 0.3708,
      "step": 5892
    },
    {
      "epoch": 1.2177,
      "grad_norm": 4.384361267089844,
      "learning_rate": 2.0742424242424244e-05,
      "loss": 0.3947,
      "step": 5893
    },
    {
      "epoch": 1.2178,
      "grad_norm": 0.791057288646698,
      "learning_rate": 2.073737373737374e-05,
      "loss": 0.0144,
      "step": 5894
    },
    {
      "epoch": 1.2179,
      "grad_norm": 1.7824095487594604,
      "learning_rate": 2.0732323232323232e-05,
      "loss": 0.0485,
      "step": 5895
    },
    {
      "epoch": 1.218,
      "grad_norm": 3.70459246635437,
      "learning_rate": 2.0727272727272728e-05,
      "loss": 0.2873,
      "step": 5896
    },
    {
      "epoch": 1.2181,
      "grad_norm": 4.416072845458984,
      "learning_rate": 2.0722222222222224e-05,
      "loss": 0.0446,
      "step": 5897
    },
    {
      "epoch": 1.2182,
      "grad_norm": 6.208083629608154,
      "learning_rate": 2.0717171717171717e-05,
      "loss": 0.7515,
      "step": 5898
    },
    {
      "epoch": 1.2183,
      "grad_norm": 6.59655237197876,
      "learning_rate": 2.0712121212121213e-05,
      "loss": 0.4854,
      "step": 5899
    },
    {
      "epoch": 1.2184,
      "grad_norm": 4.596811294555664,
      "learning_rate": 2.070707070707071e-05,
      "loss": 0.1379,
      "step": 5900
    },
    {
      "epoch": 1.2185,
      "grad_norm": 11.181085586547852,
      "learning_rate": 2.0702020202020205e-05,
      "loss": 0.3412,
      "step": 5901
    },
    {
      "epoch": 1.2186,
      "grad_norm": 1.561629056930542,
      "learning_rate": 2.0696969696969697e-05,
      "loss": 0.0242,
      "step": 5902
    },
    {
      "epoch": 1.2187000000000001,
      "grad_norm": 3.9470603466033936,
      "learning_rate": 2.0691919191919193e-05,
      "loss": 0.3465,
      "step": 5903
    },
    {
      "epoch": 1.2187999999999999,
      "grad_norm": 9.327719688415527,
      "learning_rate": 2.068686868686869e-05,
      "loss": 0.325,
      "step": 5904
    },
    {
      "epoch": 1.2189,
      "grad_norm": 1.24477219581604,
      "learning_rate": 2.0681818181818182e-05,
      "loss": 0.0193,
      "step": 5905
    },
    {
      "epoch": 1.219,
      "grad_norm": 5.37725305557251,
      "learning_rate": 2.0676767676767678e-05,
      "loss": 0.2029,
      "step": 5906
    },
    {
      "epoch": 1.2191,
      "grad_norm": 10.074589729309082,
      "learning_rate": 2.0671717171717174e-05,
      "loss": 0.1873,
      "step": 5907
    },
    {
      "epoch": 1.2192,
      "grad_norm": 3.567187547683716,
      "learning_rate": 2.0666666666666666e-05,
      "loss": 0.1061,
      "step": 5908
    },
    {
      "epoch": 1.2193,
      "grad_norm": 21.818374633789062,
      "learning_rate": 2.0661616161616162e-05,
      "loss": 0.0996,
      "step": 5909
    },
    {
      "epoch": 1.2194,
      "grad_norm": 4.893548011779785,
      "learning_rate": 2.065656565656566e-05,
      "loss": 0.4007,
      "step": 5910
    },
    {
      "epoch": 1.2195,
      "grad_norm": 3.7946150302886963,
      "learning_rate": 2.065151515151515e-05,
      "loss": 0.1161,
      "step": 5911
    },
    {
      "epoch": 1.2196,
      "grad_norm": 2.462076187133789,
      "learning_rate": 2.0646464646464647e-05,
      "loss": 0.0501,
      "step": 5912
    },
    {
      "epoch": 1.2197,
      "grad_norm": 5.5154242515563965,
      "learning_rate": 2.0641414141414143e-05,
      "loss": 0.3288,
      "step": 5913
    },
    {
      "epoch": 1.2198,
      "grad_norm": 32.264549255371094,
      "learning_rate": 2.0636363636363636e-05,
      "loss": 0.1183,
      "step": 5914
    },
    {
      "epoch": 1.2199,
      "grad_norm": 6.793822765350342,
      "learning_rate": 2.063131313131313e-05,
      "loss": 0.1664,
      "step": 5915
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.1956043243408203,
      "learning_rate": 2.0626262626262628e-05,
      "loss": 0.5938,
      "step": 5916
    },
    {
      "epoch": 1.2201,
      "grad_norm": 6.126714706420898,
      "learning_rate": 2.062121212121212e-05,
      "loss": 0.1196,
      "step": 5917
    },
    {
      "epoch": 1.2202,
      "grad_norm": 3.08986234664917,
      "learning_rate": 2.0616161616161616e-05,
      "loss": 0.0538,
      "step": 5918
    },
    {
      "epoch": 1.2203,
      "grad_norm": 13.02116584777832,
      "learning_rate": 2.0611111111111112e-05,
      "loss": 0.288,
      "step": 5919
    },
    {
      "epoch": 1.2204,
      "grad_norm": 4.477590084075928,
      "learning_rate": 2.0606060606060608e-05,
      "loss": 0.295,
      "step": 5920
    },
    {
      "epoch": 1.2205,
      "grad_norm": 1.4866985082626343,
      "learning_rate": 2.06010101010101e-05,
      "loss": 0.0407,
      "step": 5921
    },
    {
      "epoch": 1.2206,
      "grad_norm": 1.483109951019287,
      "learning_rate": 2.0595959595959597e-05,
      "loss": 0.03,
      "step": 5922
    },
    {
      "epoch": 1.2207,
      "grad_norm": 3.4703400135040283,
      "learning_rate": 2.0590909090909093e-05,
      "loss": 0.0627,
      "step": 5923
    },
    {
      "epoch": 1.2208,
      "grad_norm": 8.541455268859863,
      "learning_rate": 2.0585858585858585e-05,
      "loss": 0.5083,
      "step": 5924
    },
    {
      "epoch": 1.2209,
      "grad_norm": 1.658977746963501,
      "learning_rate": 2.058080808080808e-05,
      "loss": 0.0273,
      "step": 5925
    },
    {
      "epoch": 1.221,
      "grad_norm": 12.395578384399414,
      "learning_rate": 2.0575757575757577e-05,
      "loss": 0.2132,
      "step": 5926
    },
    {
      "epoch": 1.2211,
      "grad_norm": 2.6721343994140625,
      "learning_rate": 2.057070707070707e-05,
      "loss": 0.07,
      "step": 5927
    },
    {
      "epoch": 1.2212,
      "grad_norm": 1.5771363973617554,
      "learning_rate": 2.0565656565656566e-05,
      "loss": 0.022,
      "step": 5928
    },
    {
      "epoch": 1.2213,
      "grad_norm": 3.8478598594665527,
      "learning_rate": 2.0560606060606062e-05,
      "loss": 0.1531,
      "step": 5929
    },
    {
      "epoch": 1.2214,
      "grad_norm": 0.8878563642501831,
      "learning_rate": 2.0555555555555555e-05,
      "loss": 0.009,
      "step": 5930
    },
    {
      "epoch": 1.2215,
      "grad_norm": 4.188392162322998,
      "learning_rate": 2.055050505050505e-05,
      "loss": 0.1788,
      "step": 5931
    },
    {
      "epoch": 1.2216,
      "grad_norm": 2.31046724319458,
      "learning_rate": 2.0545454545454546e-05,
      "loss": 0.0291,
      "step": 5932
    },
    {
      "epoch": 1.2217,
      "grad_norm": 2.46016788482666,
      "learning_rate": 2.054040404040404e-05,
      "loss": 0.0504,
      "step": 5933
    },
    {
      "epoch": 1.2218,
      "grad_norm": 4.796803951263428,
      "learning_rate": 2.0535353535353535e-05,
      "loss": 0.2104,
      "step": 5934
    },
    {
      "epoch": 1.2219,
      "grad_norm": 2.348109006881714,
      "learning_rate": 2.053030303030303e-05,
      "loss": 0.0432,
      "step": 5935
    },
    {
      "epoch": 1.222,
      "grad_norm": 3.7959835529327393,
      "learning_rate": 2.0525252525252524e-05,
      "loss": 0.1552,
      "step": 5936
    },
    {
      "epoch": 1.2221,
      "grad_norm": 3.2858619689941406,
      "learning_rate": 2.052020202020202e-05,
      "loss": 0.0742,
      "step": 5937
    },
    {
      "epoch": 1.2222,
      "grad_norm": 4.0442728996276855,
      "learning_rate": 2.0515151515151516e-05,
      "loss": 0.1289,
      "step": 5938
    },
    {
      "epoch": 1.2223,
      "grad_norm": 5.086440086364746,
      "learning_rate": 2.0510101010101008e-05,
      "loss": 0.2922,
      "step": 5939
    },
    {
      "epoch": 1.2224,
      "grad_norm": 6.6067328453063965,
      "learning_rate": 2.0505050505050504e-05,
      "loss": 1.1778,
      "step": 5940
    },
    {
      "epoch": 1.2225,
      "grad_norm": 7.876432418823242,
      "learning_rate": 2.05e-05,
      "loss": 0.4204,
      "step": 5941
    },
    {
      "epoch": 1.2226,
      "grad_norm": 1.0780342817306519,
      "learning_rate": 2.0494949494949496e-05,
      "loss": 0.0233,
      "step": 5942
    },
    {
      "epoch": 1.2227000000000001,
      "grad_norm": 6.849465847015381,
      "learning_rate": 2.048989898989899e-05,
      "loss": 0.2047,
      "step": 5943
    },
    {
      "epoch": 1.2227999999999999,
      "grad_norm": 5.944521427154541,
      "learning_rate": 2.0484848484848485e-05,
      "loss": 0.3667,
      "step": 5944
    },
    {
      "epoch": 1.2229,
      "grad_norm": 1.7640084028244019,
      "learning_rate": 2.047979797979798e-05,
      "loss": 0.191,
      "step": 5945
    },
    {
      "epoch": 1.223,
      "grad_norm": 26.938274383544922,
      "learning_rate": 2.0474747474747473e-05,
      "loss": 0.1512,
      "step": 5946
    },
    {
      "epoch": 1.2231,
      "grad_norm": 2.250765323638916,
      "learning_rate": 2.0469696969696973e-05,
      "loss": 0.0834,
      "step": 5947
    },
    {
      "epoch": 1.2232,
      "grad_norm": 2.545100450515747,
      "learning_rate": 2.0464646464646465e-05,
      "loss": 0.0955,
      "step": 5948
    },
    {
      "epoch": 1.2233,
      "grad_norm": 2.533658266067505,
      "learning_rate": 2.045959595959596e-05,
      "loss": 0.0473,
      "step": 5949
    },
    {
      "epoch": 1.2234,
      "grad_norm": 0.7512052059173584,
      "learning_rate": 2.0454545454545457e-05,
      "loss": 0.0188,
      "step": 5950
    },
    {
      "epoch": 1.2235,
      "grad_norm": 10.77377986907959,
      "learning_rate": 2.0449494949494953e-05,
      "loss": 0.4129,
      "step": 5951
    },
    {
      "epoch": 1.2236,
      "grad_norm": 4.987110614776611,
      "learning_rate": 2.0444444444444446e-05,
      "loss": 0.1891,
      "step": 5952
    },
    {
      "epoch": 1.2237,
      "grad_norm": 5.7508111000061035,
      "learning_rate": 2.0439393939393942e-05,
      "loss": 0.2657,
      "step": 5953
    },
    {
      "epoch": 1.2238,
      "grad_norm": 2.3981478214263916,
      "learning_rate": 2.0434343434343438e-05,
      "loss": 0.1751,
      "step": 5954
    },
    {
      "epoch": 1.2239,
      "grad_norm": 5.939002990722656,
      "learning_rate": 2.042929292929293e-05,
      "loss": 0.2642,
      "step": 5955
    },
    {
      "epoch": 1.224,
      "grad_norm": 5.232685565948486,
      "learning_rate": 2.0424242424242427e-05,
      "loss": 0.307,
      "step": 5956
    },
    {
      "epoch": 1.2241,
      "grad_norm": 16.60964584350586,
      "learning_rate": 2.0419191919191923e-05,
      "loss": 0.2177,
      "step": 5957
    },
    {
      "epoch": 1.2242,
      "grad_norm": 9.86539363861084,
      "learning_rate": 2.0414141414141415e-05,
      "loss": 0.2993,
      "step": 5958
    },
    {
      "epoch": 1.2243,
      "grad_norm": 5.059289455413818,
      "learning_rate": 2.040909090909091e-05,
      "loss": 0.0753,
      "step": 5959
    },
    {
      "epoch": 1.2244,
      "grad_norm": 6.031787395477295,
      "learning_rate": 2.0404040404040407e-05,
      "loss": 0.2814,
      "step": 5960
    },
    {
      "epoch": 1.2245,
      "grad_norm": 3.5872702598571777,
      "learning_rate": 2.03989898989899e-05,
      "loss": 0.2769,
      "step": 5961
    },
    {
      "epoch": 1.2246,
      "grad_norm": 5.593194007873535,
      "learning_rate": 2.0393939393939396e-05,
      "loss": 0.2527,
      "step": 5962
    },
    {
      "epoch": 1.2247,
      "grad_norm": 8.47677993774414,
      "learning_rate": 2.0388888888888892e-05,
      "loss": 0.3927,
      "step": 5963
    },
    {
      "epoch": 1.2248,
      "grad_norm": 0.9032896757125854,
      "learning_rate": 2.0383838383838384e-05,
      "loss": 0.0136,
      "step": 5964
    },
    {
      "epoch": 1.2248999999999999,
      "grad_norm": 1.5464133024215698,
      "learning_rate": 2.037878787878788e-05,
      "loss": 0.0142,
      "step": 5965
    },
    {
      "epoch": 1.225,
      "grad_norm": 8.688304901123047,
      "learning_rate": 2.0373737373737376e-05,
      "loss": 0.1225,
      "step": 5966
    },
    {
      "epoch": 1.2251,
      "grad_norm": 21.864614486694336,
      "learning_rate": 2.036868686868687e-05,
      "loss": 0.2815,
      "step": 5967
    },
    {
      "epoch": 1.2252,
      "grad_norm": 2.5249791145324707,
      "learning_rate": 2.0363636363636365e-05,
      "loss": 0.086,
      "step": 5968
    },
    {
      "epoch": 1.2253,
      "grad_norm": 5.625858783721924,
      "learning_rate": 2.035858585858586e-05,
      "loss": 0.2907,
      "step": 5969
    },
    {
      "epoch": 1.2254,
      "grad_norm": 1.2156603336334229,
      "learning_rate": 2.0353535353535353e-05,
      "loss": 0.0179,
      "step": 5970
    },
    {
      "epoch": 1.2255,
      "grad_norm": 20.727737426757812,
      "learning_rate": 2.034848484848485e-05,
      "loss": 0.5556,
      "step": 5971
    },
    {
      "epoch": 1.2256,
      "grad_norm": 2.8942339420318604,
      "learning_rate": 2.0343434343434345e-05,
      "loss": 0.2366,
      "step": 5972
    },
    {
      "epoch": 1.2257,
      "grad_norm": 1.6645158529281616,
      "learning_rate": 2.033838383838384e-05,
      "loss": 0.0474,
      "step": 5973
    },
    {
      "epoch": 1.2258,
      "grad_norm": 3.9845621585845947,
      "learning_rate": 2.0333333333333334e-05,
      "loss": 0.3465,
      "step": 5974
    },
    {
      "epoch": 1.2259,
      "grad_norm": 1.8992160558700562,
      "learning_rate": 2.032828282828283e-05,
      "loss": 0.0314,
      "step": 5975
    },
    {
      "epoch": 1.226,
      "grad_norm": 13.998970985412598,
      "learning_rate": 2.0323232323232326e-05,
      "loss": 0.3755,
      "step": 5976
    },
    {
      "epoch": 1.2261,
      "grad_norm": 1.8952186107635498,
      "learning_rate": 2.031818181818182e-05,
      "loss": 0.0157,
      "step": 5977
    },
    {
      "epoch": 1.2262,
      "grad_norm": 8.717466354370117,
      "learning_rate": 2.0313131313131315e-05,
      "loss": 0.1238,
      "step": 5978
    },
    {
      "epoch": 1.2263,
      "grad_norm": 4.412684917449951,
      "learning_rate": 2.030808080808081e-05,
      "loss": 0.2691,
      "step": 5979
    },
    {
      "epoch": 1.2264,
      "grad_norm": 9.938445091247559,
      "learning_rate": 2.0303030303030303e-05,
      "loss": 0.2394,
      "step": 5980
    },
    {
      "epoch": 1.2265,
      "grad_norm": 3.2740888595581055,
      "learning_rate": 2.02979797979798e-05,
      "loss": 0.0865,
      "step": 5981
    },
    {
      "epoch": 1.2266,
      "grad_norm": 4.414059162139893,
      "learning_rate": 2.0292929292929295e-05,
      "loss": 0.3474,
      "step": 5982
    },
    {
      "epoch": 1.2267000000000001,
      "grad_norm": 9.714601516723633,
      "learning_rate": 2.0287878787878788e-05,
      "loss": 0.1881,
      "step": 5983
    },
    {
      "epoch": 1.2268,
      "grad_norm": 1.7984507083892822,
      "learning_rate": 2.0282828282828284e-05,
      "loss": 0.0232,
      "step": 5984
    },
    {
      "epoch": 1.2269,
      "grad_norm": 7.495674133300781,
      "learning_rate": 2.027777777777778e-05,
      "loss": 0.2683,
      "step": 5985
    },
    {
      "epoch": 1.227,
      "grad_norm": 5.192502021789551,
      "learning_rate": 2.0272727272727272e-05,
      "loss": 0.1195,
      "step": 5986
    },
    {
      "epoch": 1.2271,
      "grad_norm": 0.6200819611549377,
      "learning_rate": 2.026767676767677e-05,
      "loss": 0.0085,
      "step": 5987
    },
    {
      "epoch": 1.2272,
      "grad_norm": 1.6154625415802002,
      "learning_rate": 2.0262626262626264e-05,
      "loss": 0.0256,
      "step": 5988
    },
    {
      "epoch": 1.2273,
      "grad_norm": 3.5850133895874023,
      "learning_rate": 2.0257575757575757e-05,
      "loss": 0.0978,
      "step": 5989
    },
    {
      "epoch": 1.2274,
      "grad_norm": 14.655338287353516,
      "learning_rate": 2.0252525252525253e-05,
      "loss": 0.2488,
      "step": 5990
    },
    {
      "epoch": 1.2275,
      "grad_norm": 3.9058289527893066,
      "learning_rate": 2.024747474747475e-05,
      "loss": 0.1322,
      "step": 5991
    },
    {
      "epoch": 1.2276,
      "grad_norm": 3.2846195697784424,
      "learning_rate": 2.0242424242424245e-05,
      "loss": 0.1396,
      "step": 5992
    },
    {
      "epoch": 1.2277,
      "grad_norm": 6.303680419921875,
      "learning_rate": 2.0237373737373738e-05,
      "loss": 0.2998,
      "step": 5993
    },
    {
      "epoch": 1.2278,
      "grad_norm": 11.706016540527344,
      "learning_rate": 2.0232323232323234e-05,
      "loss": 0.8586,
      "step": 5994
    },
    {
      "epoch": 1.2279,
      "grad_norm": 4.248395919799805,
      "learning_rate": 2.022727272727273e-05,
      "loss": 0.2828,
      "step": 5995
    },
    {
      "epoch": 1.228,
      "grad_norm": 6.115786075592041,
      "learning_rate": 2.0222222222222222e-05,
      "loss": 0.2616,
      "step": 5996
    },
    {
      "epoch": 1.2281,
      "grad_norm": 6.524404525756836,
      "learning_rate": 2.0217171717171718e-05,
      "loss": 0.085,
      "step": 5997
    },
    {
      "epoch": 1.2282,
      "grad_norm": 2.256260633468628,
      "learning_rate": 2.0212121212121214e-05,
      "loss": 0.2728,
      "step": 5998
    },
    {
      "epoch": 1.2283,
      "grad_norm": 0.8582990765571594,
      "learning_rate": 2.0207070707070707e-05,
      "loss": 0.0093,
      "step": 5999
    },
    {
      "epoch": 1.2284,
      "grad_norm": 2.1961309909820557,
      "learning_rate": 2.0202020202020203e-05,
      "loss": 0.0394,
      "step": 6000
    },
    {
      "epoch": 1.2285,
      "grad_norm": 15.419182777404785,
      "learning_rate": 2.01969696969697e-05,
      "loss": 0.8118,
      "step": 6001
    },
    {
      "epoch": 1.2286,
      "grad_norm": 3.0020992755889893,
      "learning_rate": 2.019191919191919e-05,
      "loss": 0.1048,
      "step": 6002
    },
    {
      "epoch": 1.2287,
      "grad_norm": 8.241303443908691,
      "learning_rate": 2.0186868686868687e-05,
      "loss": 0.2209,
      "step": 6003
    },
    {
      "epoch": 1.2288000000000001,
      "grad_norm": 1.317479133605957,
      "learning_rate": 2.0181818181818183e-05,
      "loss": 0.0202,
      "step": 6004
    },
    {
      "epoch": 1.2288999999999999,
      "grad_norm": 4.734952449798584,
      "learning_rate": 2.0176767676767676e-05,
      "loss": 0.2139,
      "step": 6005
    },
    {
      "epoch": 1.229,
      "grad_norm": 8.29056167602539,
      "learning_rate": 2.0171717171717172e-05,
      "loss": 0.1239,
      "step": 6006
    },
    {
      "epoch": 1.2291,
      "grad_norm": 8.162137031555176,
      "learning_rate": 2.0166666666666668e-05,
      "loss": 0.2411,
      "step": 6007
    },
    {
      "epoch": 1.2292,
      "grad_norm": 20.167505264282227,
      "learning_rate": 2.016161616161616e-05,
      "loss": 0.2343,
      "step": 6008
    },
    {
      "epoch": 1.2293,
      "grad_norm": 3.585515260696411,
      "learning_rate": 2.0156565656565656e-05,
      "loss": 0.2701,
      "step": 6009
    },
    {
      "epoch": 1.2294,
      "grad_norm": 2.4269895553588867,
      "learning_rate": 2.0151515151515152e-05,
      "loss": 0.2729,
      "step": 6010
    },
    {
      "epoch": 1.2295,
      "grad_norm": 2.16790509223938,
      "learning_rate": 2.0146464646464645e-05,
      "loss": 0.0472,
      "step": 6011
    },
    {
      "epoch": 1.2296,
      "grad_norm": 3.886824607849121,
      "learning_rate": 2.014141414141414e-05,
      "loss": 0.2274,
      "step": 6012
    },
    {
      "epoch": 1.2297,
      "grad_norm": 4.436333656311035,
      "learning_rate": 2.0136363636363637e-05,
      "loss": 0.2583,
      "step": 6013
    },
    {
      "epoch": 1.2298,
      "grad_norm": 25.07602882385254,
      "learning_rate": 2.0131313131313133e-05,
      "loss": 0.2803,
      "step": 6014
    },
    {
      "epoch": 1.2299,
      "grad_norm": 11.961418151855469,
      "learning_rate": 2.0126262626262626e-05,
      "loss": 0.4624,
      "step": 6015
    },
    {
      "epoch": 1.23,
      "grad_norm": 4.187346458435059,
      "learning_rate": 2.012121212121212e-05,
      "loss": 0.2666,
      "step": 6016
    },
    {
      "epoch": 1.2301,
      "grad_norm": 2.867523670196533,
      "learning_rate": 2.0116161616161618e-05,
      "loss": 0.0608,
      "step": 6017
    },
    {
      "epoch": 1.2302,
      "grad_norm": 1.5766957998275757,
      "learning_rate": 2.011111111111111e-05,
      "loss": 0.0753,
      "step": 6018
    },
    {
      "epoch": 1.2303,
      "grad_norm": 3.011096239089966,
      "learning_rate": 2.0106060606060606e-05,
      "loss": 0.162,
      "step": 6019
    },
    {
      "epoch": 1.2304,
      "grad_norm": 1.659752368927002,
      "learning_rate": 2.0101010101010102e-05,
      "loss": 0.0354,
      "step": 6020
    },
    {
      "epoch": 1.2305,
      "grad_norm": 2.458332061767578,
      "learning_rate": 2.0095959595959595e-05,
      "loss": 0.2437,
      "step": 6021
    },
    {
      "epoch": 1.2306,
      "grad_norm": 5.209084510803223,
      "learning_rate": 2.009090909090909e-05,
      "loss": 0.4107,
      "step": 6022
    },
    {
      "epoch": 1.2307,
      "grad_norm": 4.496373653411865,
      "learning_rate": 2.0085858585858587e-05,
      "loss": 0.3492,
      "step": 6023
    },
    {
      "epoch": 1.2308,
      "grad_norm": 1.46846604347229,
      "learning_rate": 2.008080808080808e-05,
      "loss": 0.048,
      "step": 6024
    },
    {
      "epoch": 1.2309,
      "grad_norm": 4.541207313537598,
      "learning_rate": 2.0075757575757575e-05,
      "loss": 0.3925,
      "step": 6025
    },
    {
      "epoch": 1.231,
      "grad_norm": 5.334872722625732,
      "learning_rate": 2.007070707070707e-05,
      "loss": 0.3835,
      "step": 6026
    },
    {
      "epoch": 1.2311,
      "grad_norm": 21.223695755004883,
      "learning_rate": 2.0065656565656564e-05,
      "loss": 0.1187,
      "step": 6027
    },
    {
      "epoch": 1.2312,
      "grad_norm": 0.8433763384819031,
      "learning_rate": 2.006060606060606e-05,
      "loss": 0.0084,
      "step": 6028
    },
    {
      "epoch": 1.2313,
      "grad_norm": 7.437657356262207,
      "learning_rate": 2.0055555555555556e-05,
      "loss": 0.2123,
      "step": 6029
    },
    {
      "epoch": 1.2314,
      "grad_norm": 5.6929426193237305,
      "learning_rate": 2.005050505050505e-05,
      "loss": 0.1859,
      "step": 6030
    },
    {
      "epoch": 1.2315,
      "grad_norm": 3.0661277770996094,
      "learning_rate": 2.0045454545454544e-05,
      "loss": 0.3105,
      "step": 6031
    },
    {
      "epoch": 1.2316,
      "grad_norm": 4.064093112945557,
      "learning_rate": 2.0040404040404044e-05,
      "loss": 0.1149,
      "step": 6032
    },
    {
      "epoch": 1.2317,
      "grad_norm": 3.724564552307129,
      "learning_rate": 2.0035353535353536e-05,
      "loss": 0.2412,
      "step": 6033
    },
    {
      "epoch": 1.2318,
      "grad_norm": 4.4315948486328125,
      "learning_rate": 2.0030303030303032e-05,
      "loss": 0.1899,
      "step": 6034
    },
    {
      "epoch": 1.2319,
      "grad_norm": 6.046931743621826,
      "learning_rate": 2.002525252525253e-05,
      "loss": 0.4082,
      "step": 6035
    },
    {
      "epoch": 1.232,
      "grad_norm": 1.3493540287017822,
      "learning_rate": 2.002020202020202e-05,
      "loss": 0.0402,
      "step": 6036
    },
    {
      "epoch": 1.2321,
      "grad_norm": 6.614625453948975,
      "learning_rate": 2.0015151515151517e-05,
      "loss": 0.5164,
      "step": 6037
    },
    {
      "epoch": 1.2322,
      "grad_norm": 0.8310726881027222,
      "learning_rate": 2.0010101010101013e-05,
      "loss": 0.019,
      "step": 6038
    },
    {
      "epoch": 1.2323,
      "grad_norm": 4.202789306640625,
      "learning_rate": 2.0005050505050506e-05,
      "loss": 0.2163,
      "step": 6039
    },
    {
      "epoch": 1.2324,
      "grad_norm": 3.9338881969451904,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 6040
    },
    {
      "epoch": 1.2325,
      "grad_norm": 3.5584590435028076,
      "learning_rate": 1.9994949494949498e-05,
      "loss": 0.0322,
      "step": 6041
    },
    {
      "epoch": 1.2326,
      "grad_norm": 3.822939157485962,
      "learning_rate": 1.998989898989899e-05,
      "loss": 0.6155,
      "step": 6042
    },
    {
      "epoch": 1.2327,
      "grad_norm": 2.186025619506836,
      "learning_rate": 1.9984848484848486e-05,
      "loss": 0.226,
      "step": 6043
    },
    {
      "epoch": 1.2328000000000001,
      "grad_norm": 4.483946323394775,
      "learning_rate": 1.9979797979797982e-05,
      "loss": 0.3834,
      "step": 6044
    },
    {
      "epoch": 1.2328999999999999,
      "grad_norm": 1.8768768310546875,
      "learning_rate": 1.9974747474747478e-05,
      "loss": 0.0481,
      "step": 6045
    },
    {
      "epoch": 1.233,
      "grad_norm": 2.6090457439422607,
      "learning_rate": 1.996969696969697e-05,
      "loss": 0.2106,
      "step": 6046
    },
    {
      "epoch": 1.2331,
      "grad_norm": 6.6834259033203125,
      "learning_rate": 1.9964646464646467e-05,
      "loss": 0.2326,
      "step": 6047
    },
    {
      "epoch": 1.2332,
      "grad_norm": 2.780083656311035,
      "learning_rate": 1.9959595959595963e-05,
      "loss": 0.2412,
      "step": 6048
    },
    {
      "epoch": 1.2333,
      "grad_norm": 2.3335320949554443,
      "learning_rate": 1.9954545454545455e-05,
      "loss": 0.0467,
      "step": 6049
    },
    {
      "epoch": 1.2334,
      "grad_norm": 11.25678539276123,
      "learning_rate": 1.994949494949495e-05,
      "loss": 0.2417,
      "step": 6050
    },
    {
      "epoch": 1.2335,
      "grad_norm": 4.208549499511719,
      "learning_rate": 1.9944444444444447e-05,
      "loss": 0.3434,
      "step": 6051
    },
    {
      "epoch": 1.2336,
      "grad_norm": 3.4154305458068848,
      "learning_rate": 1.993939393939394e-05,
      "loss": 0.0855,
      "step": 6052
    },
    {
      "epoch": 1.2337,
      "grad_norm": 3.7776248455047607,
      "learning_rate": 1.9934343434343436e-05,
      "loss": 0.2846,
      "step": 6053
    },
    {
      "epoch": 1.2338,
      "grad_norm": 3.9872066974639893,
      "learning_rate": 1.9929292929292932e-05,
      "loss": 0.1039,
      "step": 6054
    },
    {
      "epoch": 1.2339,
      "grad_norm": 20.243940353393555,
      "learning_rate": 1.9924242424242425e-05,
      "loss": 0.0719,
      "step": 6055
    },
    {
      "epoch": 1.234,
      "grad_norm": 7.080476760864258,
      "learning_rate": 1.991919191919192e-05,
      "loss": 0.4322,
      "step": 6056
    },
    {
      "epoch": 1.2341,
      "grad_norm": 7.936539173126221,
      "learning_rate": 1.9914141414141417e-05,
      "loss": 0.0775,
      "step": 6057
    },
    {
      "epoch": 1.2342,
      "grad_norm": 3.11917781829834,
      "learning_rate": 1.990909090909091e-05,
      "loss": 0.1319,
      "step": 6058
    },
    {
      "epoch": 1.2343,
      "grad_norm": 22.455459594726562,
      "learning_rate": 1.9904040404040405e-05,
      "loss": 0.3547,
      "step": 6059
    },
    {
      "epoch": 1.2344,
      "grad_norm": 3.5371527671813965,
      "learning_rate": 1.98989898989899e-05,
      "loss": 0.0851,
      "step": 6060
    },
    {
      "epoch": 1.2345,
      "grad_norm": 4.795179843902588,
      "learning_rate": 1.9893939393939394e-05,
      "loss": 0.1571,
      "step": 6061
    },
    {
      "epoch": 1.2346,
      "grad_norm": 4.4850006103515625,
      "learning_rate": 1.988888888888889e-05,
      "loss": 0.2356,
      "step": 6062
    },
    {
      "epoch": 1.2347,
      "grad_norm": 3.100567579269409,
      "learning_rate": 1.9883838383838386e-05,
      "loss": 0.0877,
      "step": 6063
    },
    {
      "epoch": 1.2348,
      "grad_norm": 3.4261984825134277,
      "learning_rate": 1.987878787878788e-05,
      "loss": 0.1368,
      "step": 6064
    },
    {
      "epoch": 1.2349,
      "grad_norm": 3.3618624210357666,
      "learning_rate": 1.9873737373737374e-05,
      "loss": 0.1098,
      "step": 6065
    },
    {
      "epoch": 1.2349999999999999,
      "grad_norm": 0.7348311543464661,
      "learning_rate": 1.986868686868687e-05,
      "loss": 0.0123,
      "step": 6066
    },
    {
      "epoch": 1.2351,
      "grad_norm": 4.5125732421875,
      "learning_rate": 1.9863636363636366e-05,
      "loss": 0.1914,
      "step": 6067
    },
    {
      "epoch": 1.2352,
      "grad_norm": 4.641914367675781,
      "learning_rate": 1.985858585858586e-05,
      "loss": 0.114,
      "step": 6068
    },
    {
      "epoch": 1.2353,
      "grad_norm": 5.1436567306518555,
      "learning_rate": 1.9853535353535355e-05,
      "loss": 0.0994,
      "step": 6069
    },
    {
      "epoch": 1.2354,
      "grad_norm": 3.7501044273376465,
      "learning_rate": 1.984848484848485e-05,
      "loss": 0.1523,
      "step": 6070
    },
    {
      "epoch": 1.2355,
      "grad_norm": 14.788339614868164,
      "learning_rate": 1.9843434343434343e-05,
      "loss": 0.1836,
      "step": 6071
    },
    {
      "epoch": 1.2356,
      "grad_norm": 3.430124282836914,
      "learning_rate": 1.983838383838384e-05,
      "loss": 0.2444,
      "step": 6072
    },
    {
      "epoch": 1.2357,
      "grad_norm": 5.253172874450684,
      "learning_rate": 1.9833333333333335e-05,
      "loss": 0.1524,
      "step": 6073
    },
    {
      "epoch": 1.2358,
      "grad_norm": 16.339378356933594,
      "learning_rate": 1.9828282828282828e-05,
      "loss": 0.1275,
      "step": 6074
    },
    {
      "epoch": 1.2359,
      "grad_norm": 1.8609111309051514,
      "learning_rate": 1.9823232323232324e-05,
      "loss": 0.0332,
      "step": 6075
    },
    {
      "epoch": 1.236,
      "grad_norm": 17.37805938720703,
      "learning_rate": 1.981818181818182e-05,
      "loss": 0.2442,
      "step": 6076
    },
    {
      "epoch": 1.2361,
      "grad_norm": 5.287741184234619,
      "learning_rate": 1.9813131313131313e-05,
      "loss": 0.3769,
      "step": 6077
    },
    {
      "epoch": 1.2362,
      "grad_norm": 3.1966235637664795,
      "learning_rate": 1.980808080808081e-05,
      "loss": 0.0813,
      "step": 6078
    },
    {
      "epoch": 1.2363,
      "grad_norm": 5.830831527709961,
      "learning_rate": 1.9803030303030305e-05,
      "loss": 0.1767,
      "step": 6079
    },
    {
      "epoch": 1.2364,
      "grad_norm": 4.221264362335205,
      "learning_rate": 1.9797979797979797e-05,
      "loss": 0.142,
      "step": 6080
    },
    {
      "epoch": 1.2365,
      "grad_norm": 3.2159252166748047,
      "learning_rate": 1.9792929292929293e-05,
      "loss": 0.2521,
      "step": 6081
    },
    {
      "epoch": 1.2366,
      "grad_norm": 3.3987579345703125,
      "learning_rate": 1.978787878787879e-05,
      "loss": 0.3106,
      "step": 6082
    },
    {
      "epoch": 1.2367,
      "grad_norm": 5.497638702392578,
      "learning_rate": 1.9782828282828285e-05,
      "loss": 0.7654,
      "step": 6083
    },
    {
      "epoch": 1.2368000000000001,
      "grad_norm": 1.8941113948822021,
      "learning_rate": 1.9777777777777778e-05,
      "loss": 0.2062,
      "step": 6084
    },
    {
      "epoch": 1.2368999999999999,
      "grad_norm": 2.6390860080718994,
      "learning_rate": 1.9772727272727274e-05,
      "loss": 0.0323,
      "step": 6085
    },
    {
      "epoch": 1.237,
      "grad_norm": 8.076960563659668,
      "learning_rate": 1.976767676767677e-05,
      "loss": 0.6256,
      "step": 6086
    },
    {
      "epoch": 1.2371,
      "grad_norm": 3.9151535034179688,
      "learning_rate": 1.9762626262626262e-05,
      "loss": 0.1244,
      "step": 6087
    },
    {
      "epoch": 1.2372,
      "grad_norm": 1.4270097017288208,
      "learning_rate": 1.975757575757576e-05,
      "loss": 0.0805,
      "step": 6088
    },
    {
      "epoch": 1.2373,
      "grad_norm": 3.806138753890991,
      "learning_rate": 1.9752525252525254e-05,
      "loss": 0.1415,
      "step": 6089
    },
    {
      "epoch": 1.2374,
      "grad_norm": 2.021663188934326,
      "learning_rate": 1.9747474747474747e-05,
      "loss": 0.0349,
      "step": 6090
    },
    {
      "epoch": 1.2375,
      "grad_norm": 4.451749324798584,
      "learning_rate": 1.9742424242424243e-05,
      "loss": 0.2736,
      "step": 6091
    },
    {
      "epoch": 1.2376,
      "grad_norm": 3.664038896560669,
      "learning_rate": 1.973737373737374e-05,
      "loss": 0.2854,
      "step": 6092
    },
    {
      "epoch": 1.2377,
      "grad_norm": 4.929820537567139,
      "learning_rate": 1.973232323232323e-05,
      "loss": 0.048,
      "step": 6093
    },
    {
      "epoch": 1.2378,
      "grad_norm": 3.995549201965332,
      "learning_rate": 1.9727272727272728e-05,
      "loss": 0.2325,
      "step": 6094
    },
    {
      "epoch": 1.2379,
      "grad_norm": 6.056914329528809,
      "learning_rate": 1.9722222222222224e-05,
      "loss": 0.1002,
      "step": 6095
    },
    {
      "epoch": 1.238,
      "grad_norm": 5.537181377410889,
      "learning_rate": 1.9717171717171716e-05,
      "loss": 0.1801,
      "step": 6096
    },
    {
      "epoch": 1.2381,
      "grad_norm": 2.3198835849761963,
      "learning_rate": 1.9712121212121212e-05,
      "loss": 0.2495,
      "step": 6097
    },
    {
      "epoch": 1.2382,
      "grad_norm": 3.85622501373291,
      "learning_rate": 1.9707070707070708e-05,
      "loss": 0.3168,
      "step": 6098
    },
    {
      "epoch": 1.2383,
      "grad_norm": 7.559925556182861,
      "learning_rate": 1.97020202020202e-05,
      "loss": 0.4494,
      "step": 6099
    },
    {
      "epoch": 1.2384,
      "grad_norm": 4.429924964904785,
      "learning_rate": 1.9696969696969697e-05,
      "loss": 0.6699,
      "step": 6100
    },
    {
      "epoch": 1.2385,
      "grad_norm": 2.4530246257781982,
      "learning_rate": 1.9691919191919193e-05,
      "loss": 0.1055,
      "step": 6101
    },
    {
      "epoch": 1.2386,
      "grad_norm": 5.461737632751465,
      "learning_rate": 1.9686868686868685e-05,
      "loss": 0.3863,
      "step": 6102
    },
    {
      "epoch": 1.2387,
      "grad_norm": 3.777616024017334,
      "learning_rate": 1.968181818181818e-05,
      "loss": 0.1453,
      "step": 6103
    },
    {
      "epoch": 1.2388,
      "grad_norm": 9.67583179473877,
      "learning_rate": 1.9676767676767677e-05,
      "loss": 0.4238,
      "step": 6104
    },
    {
      "epoch": 1.2389000000000001,
      "grad_norm": 2.3296477794647217,
      "learning_rate": 1.9671717171717173e-05,
      "loss": 0.1049,
      "step": 6105
    },
    {
      "epoch": 1.2389999999999999,
      "grad_norm": 13.83845329284668,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 0.3385,
      "step": 6106
    },
    {
      "epoch": 1.2391,
      "grad_norm": 5.374402046203613,
      "learning_rate": 1.9661616161616162e-05,
      "loss": 0.367,
      "step": 6107
    },
    {
      "epoch": 1.2392,
      "grad_norm": 6.32444429397583,
      "learning_rate": 1.9656565656565658e-05,
      "loss": 0.6487,
      "step": 6108
    },
    {
      "epoch": 1.2393,
      "grad_norm": 5.7791948318481445,
      "learning_rate": 1.965151515151515e-05,
      "loss": 0.2717,
      "step": 6109
    },
    {
      "epoch": 1.2394,
      "grad_norm": 2.1171071529388428,
      "learning_rate": 1.9646464646464646e-05,
      "loss": 0.2209,
      "step": 6110
    },
    {
      "epoch": 1.2395,
      "grad_norm": 6.0373992919921875,
      "learning_rate": 1.9641414141414142e-05,
      "loss": 0.0566,
      "step": 6111
    },
    {
      "epoch": 1.2396,
      "grad_norm": 5.235950469970703,
      "learning_rate": 1.9636363636363635e-05,
      "loss": 0.1091,
      "step": 6112
    },
    {
      "epoch": 1.2397,
      "grad_norm": 1.3393303155899048,
      "learning_rate": 1.963131313131313e-05,
      "loss": 0.0171,
      "step": 6113
    },
    {
      "epoch": 1.2398,
      "grad_norm": 2.2584497928619385,
      "learning_rate": 1.9626262626262627e-05,
      "loss": 0.0472,
      "step": 6114
    },
    {
      "epoch": 1.2399,
      "grad_norm": 2.252572536468506,
      "learning_rate": 1.962121212121212e-05,
      "loss": 0.0618,
      "step": 6115
    },
    {
      "epoch": 1.24,
      "grad_norm": 3.428971767425537,
      "learning_rate": 1.9616161616161616e-05,
      "loss": 0.2788,
      "step": 6116
    },
    {
      "epoch": 1.2401,
      "grad_norm": 0.969836413860321,
      "learning_rate": 1.9611111111111115e-05,
      "loss": 0.0217,
      "step": 6117
    },
    {
      "epoch": 1.2402,
      "grad_norm": 0.9311358332633972,
      "learning_rate": 1.9606060606060608e-05,
      "loss": 0.0229,
      "step": 6118
    },
    {
      "epoch": 1.2403,
      "grad_norm": 25.790307998657227,
      "learning_rate": 1.9601010101010104e-05,
      "loss": 0.6498,
      "step": 6119
    },
    {
      "epoch": 1.2404,
      "grad_norm": 4.884109020233154,
      "learning_rate": 1.95959595959596e-05,
      "loss": 0.0574,
      "step": 6120
    },
    {
      "epoch": 1.2405,
      "grad_norm": 2.8395094871520996,
      "learning_rate": 1.9590909090909092e-05,
      "loss": 0.052,
      "step": 6121
    },
    {
      "epoch": 1.2406,
      "grad_norm": 5.018804550170898,
      "learning_rate": 1.9585858585858588e-05,
      "loss": 0.6705,
      "step": 6122
    },
    {
      "epoch": 1.2407,
      "grad_norm": 3.3996169567108154,
      "learning_rate": 1.9580808080808084e-05,
      "loss": 0.1172,
      "step": 6123
    },
    {
      "epoch": 1.2408,
      "grad_norm": 5.292490005493164,
      "learning_rate": 1.9575757575757577e-05,
      "loss": 0.0769,
      "step": 6124
    },
    {
      "epoch": 1.2409,
      "grad_norm": 3.881319046020508,
      "learning_rate": 1.9570707070707073e-05,
      "loss": 0.1738,
      "step": 6125
    },
    {
      "epoch": 1.241,
      "grad_norm": 2.0070641040802,
      "learning_rate": 1.956565656565657e-05,
      "loss": 0.0326,
      "step": 6126
    },
    {
      "epoch": 1.2411,
      "grad_norm": 1.6455402374267578,
      "learning_rate": 1.956060606060606e-05,
      "loss": 0.0589,
      "step": 6127
    },
    {
      "epoch": 1.2412,
      "grad_norm": 15.566871643066406,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 0.254,
      "step": 6128
    },
    {
      "epoch": 1.2413,
      "grad_norm": 2.64431095123291,
      "learning_rate": 1.9550505050505053e-05,
      "loss": 0.171,
      "step": 6129
    },
    {
      "epoch": 1.2414,
      "grad_norm": 9.08087158203125,
      "learning_rate": 1.9545454545454546e-05,
      "loss": 0.8207,
      "step": 6130
    },
    {
      "epoch": 1.2415,
      "grad_norm": 9.387184143066406,
      "learning_rate": 1.9540404040404042e-05,
      "loss": 0.3193,
      "step": 6131
    },
    {
      "epoch": 1.2416,
      "grad_norm": 3.317033290863037,
      "learning_rate": 1.9535353535353538e-05,
      "loss": 0.0794,
      "step": 6132
    },
    {
      "epoch": 1.2417,
      "grad_norm": 2.82035493850708,
      "learning_rate": 1.953030303030303e-05,
      "loss": 0.0938,
      "step": 6133
    },
    {
      "epoch": 1.2418,
      "grad_norm": 1.1608692407608032,
      "learning_rate": 1.9525252525252526e-05,
      "loss": 0.0147,
      "step": 6134
    },
    {
      "epoch": 1.2419,
      "grad_norm": 5.129396915435791,
      "learning_rate": 1.9520202020202022e-05,
      "loss": 0.4146,
      "step": 6135
    },
    {
      "epoch": 1.242,
      "grad_norm": 2.384561061859131,
      "learning_rate": 1.951515151515152e-05,
      "loss": 0.0593,
      "step": 6136
    },
    {
      "epoch": 1.2421,
      "grad_norm": 1.4172139167785645,
      "learning_rate": 1.951010101010101e-05,
      "loss": 0.0463,
      "step": 6137
    },
    {
      "epoch": 1.2422,
      "grad_norm": 2.815081834793091,
      "learning_rate": 1.9505050505050507e-05,
      "loss": 0.2129,
      "step": 6138
    },
    {
      "epoch": 1.2423,
      "grad_norm": 6.611221790313721,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.0744,
      "step": 6139
    },
    {
      "epoch": 1.2424,
      "grad_norm": 3.594994306564331,
      "learning_rate": 1.9494949494949496e-05,
      "loss": 0.1439,
      "step": 6140
    },
    {
      "epoch": 1.2425,
      "grad_norm": 2.076230525970459,
      "learning_rate": 1.948989898989899e-05,
      "loss": 0.038,
      "step": 6141
    },
    {
      "epoch": 1.2426,
      "grad_norm": 3.1296913623809814,
      "learning_rate": 1.9484848484848488e-05,
      "loss": 0.3115,
      "step": 6142
    },
    {
      "epoch": 1.2427,
      "grad_norm": 1.421118140220642,
      "learning_rate": 1.947979797979798e-05,
      "loss": 0.0153,
      "step": 6143
    },
    {
      "epoch": 1.2428,
      "grad_norm": 4.794938564300537,
      "learning_rate": 1.9474747474747476e-05,
      "loss": 0.1222,
      "step": 6144
    },
    {
      "epoch": 1.2429000000000001,
      "grad_norm": 2.244291305541992,
      "learning_rate": 1.9469696969696972e-05,
      "loss": 0.0582,
      "step": 6145
    },
    {
      "epoch": 1.2429999999999999,
      "grad_norm": 1.86780846118927,
      "learning_rate": 1.9464646464646465e-05,
      "loss": 0.0387,
      "step": 6146
    },
    {
      "epoch": 1.2431,
      "grad_norm": 6.3164167404174805,
      "learning_rate": 1.945959595959596e-05,
      "loss": 0.1886,
      "step": 6147
    },
    {
      "epoch": 1.2432,
      "grad_norm": 4.242239475250244,
      "learning_rate": 1.9454545454545457e-05,
      "loss": 0.1725,
      "step": 6148
    },
    {
      "epoch": 1.2433,
      "grad_norm": 8.920600891113281,
      "learning_rate": 1.944949494949495e-05,
      "loss": 0.1307,
      "step": 6149
    },
    {
      "epoch": 1.2434,
      "grad_norm": 5.33392333984375,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.3985,
      "step": 6150
    },
    {
      "epoch": 1.2435,
      "grad_norm": 1.6551294326782227,
      "learning_rate": 1.943939393939394e-05,
      "loss": 0.0272,
      "step": 6151
    },
    {
      "epoch": 1.2436,
      "grad_norm": 2.8844547271728516,
      "learning_rate": 1.9434343434343434e-05,
      "loss": 0.2421,
      "step": 6152
    },
    {
      "epoch": 1.2437,
      "grad_norm": 4.108041286468506,
      "learning_rate": 1.942929292929293e-05,
      "loss": 0.0833,
      "step": 6153
    },
    {
      "epoch": 1.2438,
      "grad_norm": 6.173274517059326,
      "learning_rate": 1.9424242424242426e-05,
      "loss": 0.0971,
      "step": 6154
    },
    {
      "epoch": 1.2439,
      "grad_norm": 2.595707654953003,
      "learning_rate": 1.9419191919191922e-05,
      "loss": 0.2471,
      "step": 6155
    },
    {
      "epoch": 1.244,
      "grad_norm": 9.076233863830566,
      "learning_rate": 1.9414141414141415e-05,
      "loss": 0.1179,
      "step": 6156
    },
    {
      "epoch": 1.2441,
      "grad_norm": 3.3523075580596924,
      "learning_rate": 1.940909090909091e-05,
      "loss": 0.2444,
      "step": 6157
    },
    {
      "epoch": 1.2442,
      "grad_norm": 11.717608451843262,
      "learning_rate": 1.9404040404040407e-05,
      "loss": 0.2602,
      "step": 6158
    },
    {
      "epoch": 1.2443,
      "grad_norm": 1.268429160118103,
      "learning_rate": 1.93989898989899e-05,
      "loss": 0.018,
      "step": 6159
    },
    {
      "epoch": 1.2444,
      "grad_norm": 1.5723882913589478,
      "learning_rate": 1.9393939393939395e-05,
      "loss": 0.0248,
      "step": 6160
    },
    {
      "epoch": 1.2445,
      "grad_norm": 3.4583702087402344,
      "learning_rate": 1.938888888888889e-05,
      "loss": 0.2613,
      "step": 6161
    },
    {
      "epoch": 1.2446,
      "grad_norm": 8.42194938659668,
      "learning_rate": 1.9383838383838384e-05,
      "loss": 0.1141,
      "step": 6162
    },
    {
      "epoch": 1.2447,
      "grad_norm": 3.669978380203247,
      "learning_rate": 1.937878787878788e-05,
      "loss": 0.1637,
      "step": 6163
    },
    {
      "epoch": 1.2448,
      "grad_norm": 2.993828773498535,
      "learning_rate": 1.9373737373737376e-05,
      "loss": 0.0967,
      "step": 6164
    },
    {
      "epoch": 1.2449,
      "grad_norm": 2.9348838329315186,
      "learning_rate": 1.9368686868686868e-05,
      "loss": 0.0236,
      "step": 6165
    },
    {
      "epoch": 1.245,
      "grad_norm": 2.502685070037842,
      "learning_rate": 1.9363636363636364e-05,
      "loss": 0.0641,
      "step": 6166
    },
    {
      "epoch": 1.2451,
      "grad_norm": 1.9998047351837158,
      "learning_rate": 1.935858585858586e-05,
      "loss": 0.0201,
      "step": 6167
    },
    {
      "epoch": 1.2452,
      "grad_norm": 1.9756022691726685,
      "learning_rate": 1.9353535353535353e-05,
      "loss": 0.0331,
      "step": 6168
    },
    {
      "epoch": 1.2453,
      "grad_norm": 4.9745564460754395,
      "learning_rate": 1.934848484848485e-05,
      "loss": 0.0677,
      "step": 6169
    },
    {
      "epoch": 1.2454,
      "grad_norm": 3.317308187484741,
      "learning_rate": 1.9343434343434345e-05,
      "loss": 0.1056,
      "step": 6170
    },
    {
      "epoch": 1.2455,
      "grad_norm": 4.496276378631592,
      "learning_rate": 1.9338383838383837e-05,
      "loss": 0.0863,
      "step": 6171
    },
    {
      "epoch": 1.2456,
      "grad_norm": 10.490632057189941,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 0.2054,
      "step": 6172
    },
    {
      "epoch": 1.2457,
      "grad_norm": 3.279313325881958,
      "learning_rate": 1.932828282828283e-05,
      "loss": 0.2238,
      "step": 6173
    },
    {
      "epoch": 1.2458,
      "grad_norm": 8.254262924194336,
      "learning_rate": 1.9323232323232322e-05,
      "loss": 0.0879,
      "step": 6174
    },
    {
      "epoch": 1.2459,
      "grad_norm": 1.8048357963562012,
      "learning_rate": 1.9318181818181818e-05,
      "loss": 0.0109,
      "step": 6175
    },
    {
      "epoch": 1.246,
      "grad_norm": 4.80662202835083,
      "learning_rate": 1.9313131313131314e-05,
      "loss": 0.1383,
      "step": 6176
    },
    {
      "epoch": 1.2461,
      "grad_norm": 1.6084132194519043,
      "learning_rate": 1.930808080808081e-05,
      "loss": 0.0271,
      "step": 6177
    },
    {
      "epoch": 1.2462,
      "grad_norm": 3.2161550521850586,
      "learning_rate": 1.9303030303030303e-05,
      "loss": 0.3034,
      "step": 6178
    },
    {
      "epoch": 1.2463,
      "grad_norm": 2.1178953647613525,
      "learning_rate": 1.92979797979798e-05,
      "loss": 0.2331,
      "step": 6179
    },
    {
      "epoch": 1.2464,
      "grad_norm": 6.601255893707275,
      "learning_rate": 1.9292929292929295e-05,
      "loss": 0.1656,
      "step": 6180
    },
    {
      "epoch": 1.2465,
      "grad_norm": 5.75533390045166,
      "learning_rate": 1.9287878787878787e-05,
      "loss": 0.1382,
      "step": 6181
    },
    {
      "epoch": 1.2466,
      "grad_norm": 2.583468437194824,
      "learning_rate": 1.9282828282828283e-05,
      "loss": 0.0671,
      "step": 6182
    },
    {
      "epoch": 1.2467,
      "grad_norm": 1.134522795677185,
      "learning_rate": 1.927777777777778e-05,
      "loss": 0.0099,
      "step": 6183
    },
    {
      "epoch": 1.2468,
      "grad_norm": 6.838954448699951,
      "learning_rate": 1.9272727272727272e-05,
      "loss": 0.6823,
      "step": 6184
    },
    {
      "epoch": 1.2469000000000001,
      "grad_norm": 3.739121198654175,
      "learning_rate": 1.9267676767676768e-05,
      "loss": 0.1558,
      "step": 6185
    },
    {
      "epoch": 1.2469999999999999,
      "grad_norm": 4.951013565063477,
      "learning_rate": 1.9262626262626264e-05,
      "loss": 0.2149,
      "step": 6186
    },
    {
      "epoch": 1.2471,
      "grad_norm": 9.41068172454834,
      "learning_rate": 1.9257575757575756e-05,
      "loss": 0.3619,
      "step": 6187
    },
    {
      "epoch": 1.2472,
      "grad_norm": 6.729616165161133,
      "learning_rate": 1.9252525252525252e-05,
      "loss": 0.4576,
      "step": 6188
    },
    {
      "epoch": 1.2473,
      "grad_norm": 3.814450263977051,
      "learning_rate": 1.924747474747475e-05,
      "loss": 0.2867,
      "step": 6189
    },
    {
      "epoch": 1.2474,
      "grad_norm": 3.0289018154144287,
      "learning_rate": 1.924242424242424e-05,
      "loss": 0.1012,
      "step": 6190
    },
    {
      "epoch": 1.2475,
      "grad_norm": 20.628812789916992,
      "learning_rate": 1.9237373737373737e-05,
      "loss": 0.3131,
      "step": 6191
    },
    {
      "epoch": 1.2476,
      "grad_norm": 5.570855140686035,
      "learning_rate": 1.9232323232323233e-05,
      "loss": 0.1904,
      "step": 6192
    },
    {
      "epoch": 1.2477,
      "grad_norm": 1.1989052295684814,
      "learning_rate": 1.9227272727272726e-05,
      "loss": 0.0102,
      "step": 6193
    },
    {
      "epoch": 1.2478,
      "grad_norm": 2.956742286682129,
      "learning_rate": 1.922222222222222e-05,
      "loss": 0.5527,
      "step": 6194
    },
    {
      "epoch": 1.2479,
      "grad_norm": 3.45316743850708,
      "learning_rate": 1.9217171717171718e-05,
      "loss": 0.0535,
      "step": 6195
    },
    {
      "epoch": 1.248,
      "grad_norm": 5.672026634216309,
      "learning_rate": 1.9212121212121213e-05,
      "loss": 0.4026,
      "step": 6196
    },
    {
      "epoch": 1.2481,
      "grad_norm": 5.1493024826049805,
      "learning_rate": 1.9207070707070706e-05,
      "loss": 0.1431,
      "step": 6197
    },
    {
      "epoch": 1.2482,
      "grad_norm": 3.9200613498687744,
      "learning_rate": 1.9202020202020202e-05,
      "loss": 0.1233,
      "step": 6198
    },
    {
      "epoch": 1.2483,
      "grad_norm": 1.416495442390442,
      "learning_rate": 1.9196969696969698e-05,
      "loss": 0.0189,
      "step": 6199
    },
    {
      "epoch": 1.2484,
      "grad_norm": 1.9180177450180054,
      "learning_rate": 1.919191919191919e-05,
      "loss": 0.0775,
      "step": 6200
    },
    {
      "epoch": 1.2485,
      "grad_norm": 5.539704322814941,
      "learning_rate": 1.9186868686868687e-05,
      "loss": 0.0336,
      "step": 6201
    },
    {
      "epoch": 1.2486,
      "grad_norm": 7.401930809020996,
      "learning_rate": 1.9181818181818183e-05,
      "loss": 0.1145,
      "step": 6202
    },
    {
      "epoch": 1.2487,
      "grad_norm": 0.23345524072647095,
      "learning_rate": 1.917676767676768e-05,
      "loss": 0.0022,
      "step": 6203
    },
    {
      "epoch": 1.2488,
      "grad_norm": 7.720966815948486,
      "learning_rate": 1.9171717171717175e-05,
      "loss": 0.0978,
      "step": 6204
    },
    {
      "epoch": 1.2489,
      "grad_norm": 7.041057109832764,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 0.3407,
      "step": 6205
    },
    {
      "epoch": 1.249,
      "grad_norm": 12.814632415771484,
      "learning_rate": 1.9161616161616163e-05,
      "loss": 0.2102,
      "step": 6206
    },
    {
      "epoch": 1.2490999999999999,
      "grad_norm": 12.698078155517578,
      "learning_rate": 1.915656565656566e-05,
      "loss": 0.3769,
      "step": 6207
    },
    {
      "epoch": 1.2492,
      "grad_norm": 2.8154568672180176,
      "learning_rate": 1.9151515151515155e-05,
      "loss": 0.1171,
      "step": 6208
    },
    {
      "epoch": 1.2493,
      "grad_norm": 0.9030357003211975,
      "learning_rate": 1.9146464646464648e-05,
      "loss": 0.0228,
      "step": 6209
    },
    {
      "epoch": 1.2494,
      "grad_norm": 8.947032928466797,
      "learning_rate": 1.9141414141414144e-05,
      "loss": 0.3663,
      "step": 6210
    },
    {
      "epoch": 1.2495,
      "grad_norm": 8.401777267456055,
      "learning_rate": 1.913636363636364e-05,
      "loss": 0.2006,
      "step": 6211
    },
    {
      "epoch": 1.2496,
      "grad_norm": 5.621592998504639,
      "learning_rate": 1.9131313131313132e-05,
      "loss": 0.1917,
      "step": 6212
    },
    {
      "epoch": 1.2497,
      "grad_norm": 3.1765356063842773,
      "learning_rate": 1.912626262626263e-05,
      "loss": 0.3229,
      "step": 6213
    },
    {
      "epoch": 1.2498,
      "grad_norm": 2.4445059299468994,
      "learning_rate": 1.9121212121212124e-05,
      "loss": 0.0418,
      "step": 6214
    },
    {
      "epoch": 1.2499,
      "grad_norm": 2.8028533458709717,
      "learning_rate": 1.9116161616161617e-05,
      "loss": 0.1266,
      "step": 6215
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.278815746307373,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 0.3148,
      "step": 6216
    },
    {
      "epoch": 1.2501,
      "grad_norm": 2.2112507820129395,
      "learning_rate": 1.910606060606061e-05,
      "loss": 0.2135,
      "step": 6217
    },
    {
      "epoch": 1.2502,
      "grad_norm": 2.5200185775756836,
      "learning_rate": 1.91010101010101e-05,
      "loss": 0.0704,
      "step": 6218
    },
    {
      "epoch": 1.2503,
      "grad_norm": 4.977707862854004,
      "learning_rate": 1.9095959595959598e-05,
      "loss": 0.2588,
      "step": 6219
    },
    {
      "epoch": 1.2504,
      "grad_norm": 1.3628106117248535,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 0.0379,
      "step": 6220
    },
    {
      "epoch": 1.2505,
      "grad_norm": 4.095950603485107,
      "learning_rate": 1.9085858585858586e-05,
      "loss": 0.2988,
      "step": 6221
    },
    {
      "epoch": 1.2506,
      "grad_norm": 3.0624496936798096,
      "learning_rate": 1.9080808080808082e-05,
      "loss": 0.0694,
      "step": 6222
    },
    {
      "epoch": 1.2507,
      "grad_norm": 7.357241153717041,
      "learning_rate": 1.9075757575757578e-05,
      "loss": 0.3964,
      "step": 6223
    },
    {
      "epoch": 1.2508,
      "grad_norm": 15.985004425048828,
      "learning_rate": 1.907070707070707e-05,
      "loss": 0.3357,
      "step": 6224
    },
    {
      "epoch": 1.2509000000000001,
      "grad_norm": 9.992728233337402,
      "learning_rate": 1.9065656565656567e-05,
      "loss": 0.2463,
      "step": 6225
    },
    {
      "epoch": 1.251,
      "grad_norm": 3.3286073207855225,
      "learning_rate": 1.9060606060606063e-05,
      "loss": 0.2801,
      "step": 6226
    },
    {
      "epoch": 1.2511,
      "grad_norm": 8.65282154083252,
      "learning_rate": 1.905555555555556e-05,
      "loss": 0.4019,
      "step": 6227
    },
    {
      "epoch": 1.2511999999999999,
      "grad_norm": 4.285361289978027,
      "learning_rate": 1.905050505050505e-05,
      "loss": 0.098,
      "step": 6228
    },
    {
      "epoch": 1.2513,
      "grad_norm": 4.49790096282959,
      "learning_rate": 1.9045454545454547e-05,
      "loss": 0.2087,
      "step": 6229
    },
    {
      "epoch": 1.2514,
      "grad_norm": 13.81605339050293,
      "learning_rate": 1.9040404040404043e-05,
      "loss": 0.4455,
      "step": 6230
    },
    {
      "epoch": 1.2515,
      "grad_norm": 5.722465991973877,
      "learning_rate": 1.9035353535353536e-05,
      "loss": 0.093,
      "step": 6231
    },
    {
      "epoch": 1.2516,
      "grad_norm": 16.355661392211914,
      "learning_rate": 1.9030303030303032e-05,
      "loss": 0.2843,
      "step": 6232
    },
    {
      "epoch": 1.2517,
      "grad_norm": 7.224652290344238,
      "learning_rate": 1.9025252525252528e-05,
      "loss": 0.3305,
      "step": 6233
    },
    {
      "epoch": 1.2518,
      "grad_norm": 1.6957660913467407,
      "learning_rate": 1.902020202020202e-05,
      "loss": 0.036,
      "step": 6234
    },
    {
      "epoch": 1.2519,
      "grad_norm": 5.195170879364014,
      "learning_rate": 1.9015151515151516e-05,
      "loss": 0.2146,
      "step": 6235
    },
    {
      "epoch": 1.252,
      "grad_norm": 4.727704048156738,
      "learning_rate": 1.9010101010101012e-05,
      "loss": 0.152,
      "step": 6236
    },
    {
      "epoch": 1.2521,
      "grad_norm": 4.690859794616699,
      "learning_rate": 1.9005050505050505e-05,
      "loss": 0.2987,
      "step": 6237
    },
    {
      "epoch": 1.2522,
      "grad_norm": 2.9610512256622314,
      "learning_rate": 1.9e-05,
      "loss": 0.0529,
      "step": 6238
    },
    {
      "epoch": 1.2523,
      "grad_norm": 0.9151058197021484,
      "learning_rate": 1.8994949494949497e-05,
      "loss": 0.0177,
      "step": 6239
    },
    {
      "epoch": 1.2524,
      "grad_norm": 10.659239768981934,
      "learning_rate": 1.898989898989899e-05,
      "loss": 0.1038,
      "step": 6240
    },
    {
      "epoch": 1.2525,
      "grad_norm": 9.24109935760498,
      "learning_rate": 1.8984848484848486e-05,
      "loss": 0.4,
      "step": 6241
    },
    {
      "epoch": 1.2526,
      "grad_norm": 4.330331802368164,
      "learning_rate": 1.897979797979798e-05,
      "loss": 0.2496,
      "step": 6242
    },
    {
      "epoch": 1.2527,
      "grad_norm": 13.717588424682617,
      "learning_rate": 1.8974747474747474e-05,
      "loss": 0.2395,
      "step": 6243
    },
    {
      "epoch": 1.2528000000000001,
      "grad_norm": 1.990870714187622,
      "learning_rate": 1.896969696969697e-05,
      "loss": 0.0688,
      "step": 6244
    },
    {
      "epoch": 1.2529,
      "grad_norm": 2.2706682682037354,
      "learning_rate": 1.8964646464646466e-05,
      "loss": 0.1036,
      "step": 6245
    },
    {
      "epoch": 1.2530000000000001,
      "grad_norm": 9.851455688476562,
      "learning_rate": 1.895959595959596e-05,
      "loss": 0.5377,
      "step": 6246
    },
    {
      "epoch": 1.2530999999999999,
      "grad_norm": 8.529763221740723,
      "learning_rate": 1.8954545454545455e-05,
      "loss": 0.114,
      "step": 6247
    },
    {
      "epoch": 1.2532,
      "grad_norm": 33.41804504394531,
      "learning_rate": 1.894949494949495e-05,
      "loss": 0.0831,
      "step": 6248
    },
    {
      "epoch": 1.2533,
      "grad_norm": 2.6491522789001465,
      "learning_rate": 1.8944444444444447e-05,
      "loss": 0.0845,
      "step": 6249
    },
    {
      "epoch": 1.2534,
      "grad_norm": 3.319546699523926,
      "learning_rate": 1.893939393939394e-05,
      "loss": 0.0426,
      "step": 6250
    },
    {
      "epoch": 1.2535,
      "grad_norm": 1.6935075521469116,
      "learning_rate": 1.8934343434343435e-05,
      "loss": 0.0483,
      "step": 6251
    },
    {
      "epoch": 1.2536,
      "grad_norm": 2.457643747329712,
      "learning_rate": 1.892929292929293e-05,
      "loss": 0.049,
      "step": 6252
    },
    {
      "epoch": 1.2537,
      "grad_norm": 1.7013722658157349,
      "learning_rate": 1.8924242424242424e-05,
      "loss": 0.044,
      "step": 6253
    },
    {
      "epoch": 1.2538,
      "grad_norm": 1.9566733837127686,
      "learning_rate": 1.891919191919192e-05,
      "loss": 0.0258,
      "step": 6254
    },
    {
      "epoch": 1.2539,
      "grad_norm": 0.8872674703598022,
      "learning_rate": 1.8914141414141416e-05,
      "loss": 0.0206,
      "step": 6255
    },
    {
      "epoch": 1.254,
      "grad_norm": 9.094738960266113,
      "learning_rate": 1.890909090909091e-05,
      "loss": 0.8127,
      "step": 6256
    },
    {
      "epoch": 1.2541,
      "grad_norm": 8.278011322021484,
      "learning_rate": 1.8904040404040405e-05,
      "loss": 0.3878,
      "step": 6257
    },
    {
      "epoch": 1.2542,
      "grad_norm": 1.017379879951477,
      "learning_rate": 1.88989898989899e-05,
      "loss": 0.051,
      "step": 6258
    },
    {
      "epoch": 1.2543,
      "grad_norm": 3.1895344257354736,
      "learning_rate": 1.8893939393939393e-05,
      "loss": 0.0202,
      "step": 6259
    },
    {
      "epoch": 1.2544,
      "grad_norm": 4.799853324890137,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.1684,
      "step": 6260
    },
    {
      "epoch": 1.2545,
      "grad_norm": 9.209220886230469,
      "learning_rate": 1.8883838383838385e-05,
      "loss": 0.37,
      "step": 6261
    },
    {
      "epoch": 1.2546,
      "grad_norm": 9.548107147216797,
      "learning_rate": 1.8878787878787878e-05,
      "loss": 0.1875,
      "step": 6262
    },
    {
      "epoch": 1.2547,
      "grad_norm": 3.625953435897827,
      "learning_rate": 1.8873737373737374e-05,
      "loss": 0.0957,
      "step": 6263
    },
    {
      "epoch": 1.2548,
      "grad_norm": 10.908575057983398,
      "learning_rate": 1.886868686868687e-05,
      "loss": 0.6872,
      "step": 6264
    },
    {
      "epoch": 1.2549000000000001,
      "grad_norm": 6.562564849853516,
      "learning_rate": 1.8863636363636362e-05,
      "loss": 0.1977,
      "step": 6265
    },
    {
      "epoch": 1.255,
      "grad_norm": 1.585446834564209,
      "learning_rate": 1.8858585858585858e-05,
      "loss": 0.0308,
      "step": 6266
    },
    {
      "epoch": 1.2551,
      "grad_norm": 7.149526596069336,
      "learning_rate": 1.8853535353535354e-05,
      "loss": 0.5298,
      "step": 6267
    },
    {
      "epoch": 1.2551999999999999,
      "grad_norm": 4.744585037231445,
      "learning_rate": 1.884848484848485e-05,
      "loss": 0.1326,
      "step": 6268
    },
    {
      "epoch": 1.2553,
      "grad_norm": 55.41714096069336,
      "learning_rate": 1.8843434343434343e-05,
      "loss": 0.3762,
      "step": 6269
    },
    {
      "epoch": 1.2554,
      "grad_norm": 6.471060752868652,
      "learning_rate": 1.883838383838384e-05,
      "loss": 0.2688,
      "step": 6270
    },
    {
      "epoch": 1.2555,
      "grad_norm": 6.524525165557861,
      "learning_rate": 1.8833333333333335e-05,
      "loss": 0.3153,
      "step": 6271
    },
    {
      "epoch": 1.2556,
      "grad_norm": 3.6855297088623047,
      "learning_rate": 1.8828282828282827e-05,
      "loss": 0.2448,
      "step": 6272
    },
    {
      "epoch": 1.2557,
      "grad_norm": 6.207939624786377,
      "learning_rate": 1.8823232323232323e-05,
      "loss": 0.114,
      "step": 6273
    },
    {
      "epoch": 1.2558,
      "grad_norm": 4.992859363555908,
      "learning_rate": 1.881818181818182e-05,
      "loss": 0.2802,
      "step": 6274
    },
    {
      "epoch": 1.2559,
      "grad_norm": 6.975537300109863,
      "learning_rate": 1.8813131313131312e-05,
      "loss": 0.1183,
      "step": 6275
    },
    {
      "epoch": 1.256,
      "grad_norm": 2.9032535552978516,
      "learning_rate": 1.8808080808080808e-05,
      "loss": 0.266,
      "step": 6276
    },
    {
      "epoch": 1.2561,
      "grad_norm": 4.501613140106201,
      "learning_rate": 1.8803030303030304e-05,
      "loss": 0.1123,
      "step": 6277
    },
    {
      "epoch": 1.2562,
      "grad_norm": 3.781432628631592,
      "learning_rate": 1.8797979797979797e-05,
      "loss": 0.0878,
      "step": 6278
    },
    {
      "epoch": 1.2563,
      "grad_norm": 7.130786895751953,
      "learning_rate": 1.8792929292929293e-05,
      "loss": 0.2437,
      "step": 6279
    },
    {
      "epoch": 1.2564,
      "grad_norm": 5.3871870040893555,
      "learning_rate": 1.878787878787879e-05,
      "loss": 0.2313,
      "step": 6280
    },
    {
      "epoch": 1.2565,
      "grad_norm": 4.079959392547607,
      "learning_rate": 1.878282828282828e-05,
      "loss": 0.1295,
      "step": 6281
    },
    {
      "epoch": 1.2566,
      "grad_norm": 3.9905786514282227,
      "learning_rate": 1.8777777777777777e-05,
      "loss": 0.0903,
      "step": 6282
    },
    {
      "epoch": 1.2567,
      "grad_norm": 3.914095401763916,
      "learning_rate": 1.8772727272727273e-05,
      "loss": 0.1116,
      "step": 6283
    },
    {
      "epoch": 1.2568,
      "grad_norm": 3.252932548522949,
      "learning_rate": 1.8767676767676766e-05,
      "loss": 0.0688,
      "step": 6284
    },
    {
      "epoch": 1.2569,
      "grad_norm": 2.7784905433654785,
      "learning_rate": 1.8762626262626262e-05,
      "loss": 0.1193,
      "step": 6285
    },
    {
      "epoch": 1.2570000000000001,
      "grad_norm": 35.22977066040039,
      "learning_rate": 1.8757575757575758e-05,
      "loss": 0.1574,
      "step": 6286
    },
    {
      "epoch": 1.2570999999999999,
      "grad_norm": 0.1940050572156906,
      "learning_rate": 1.875252525252525e-05,
      "loss": 0.0023,
      "step": 6287
    },
    {
      "epoch": 1.2572,
      "grad_norm": 4.541574954986572,
      "learning_rate": 1.874747474747475e-05,
      "loss": 0.2014,
      "step": 6288
    },
    {
      "epoch": 1.2572999999999999,
      "grad_norm": 3.8086133003234863,
      "learning_rate": 1.8742424242424246e-05,
      "loss": 0.0652,
      "step": 6289
    },
    {
      "epoch": 1.2574,
      "grad_norm": 9.319779396057129,
      "learning_rate": 1.873737373737374e-05,
      "loss": 0.1398,
      "step": 6290
    },
    {
      "epoch": 1.2575,
      "grad_norm": 14.171106338500977,
      "learning_rate": 1.8732323232323234e-05,
      "loss": 0.1527,
      "step": 6291
    },
    {
      "epoch": 1.2576,
      "grad_norm": 1.8149325847625732,
      "learning_rate": 1.872727272727273e-05,
      "loss": 0.0323,
      "step": 6292
    },
    {
      "epoch": 1.2577,
      "grad_norm": 11.160685539245605,
      "learning_rate": 1.8722222222222223e-05,
      "loss": 0.1419,
      "step": 6293
    },
    {
      "epoch": 1.2578,
      "grad_norm": 5.139343738555908,
      "learning_rate": 1.871717171717172e-05,
      "loss": 0.1421,
      "step": 6294
    },
    {
      "epoch": 1.2579,
      "grad_norm": 4.467588901519775,
      "learning_rate": 1.8712121212121215e-05,
      "loss": 0.1186,
      "step": 6295
    },
    {
      "epoch": 1.258,
      "grad_norm": 1.9287080764770508,
      "learning_rate": 1.8707070707070707e-05,
      "loss": 0.0788,
      "step": 6296
    },
    {
      "epoch": 1.2581,
      "grad_norm": 2.236910581588745,
      "learning_rate": 1.8702020202020203e-05,
      "loss": 0.0555,
      "step": 6297
    },
    {
      "epoch": 1.2582,
      "grad_norm": 2.025270938873291,
      "learning_rate": 1.86969696969697e-05,
      "loss": 0.0417,
      "step": 6298
    },
    {
      "epoch": 1.2583,
      "grad_norm": 4.720226764678955,
      "learning_rate": 1.8691919191919195e-05,
      "loss": 0.1683,
      "step": 6299
    },
    {
      "epoch": 1.2584,
      "grad_norm": 6.7825093269348145,
      "learning_rate": 1.8686868686868688e-05,
      "loss": 0.1812,
      "step": 6300
    },
    {
      "epoch": 1.2585,
      "grad_norm": 5.756746292114258,
      "learning_rate": 1.8681818181818184e-05,
      "loss": 0.1735,
      "step": 6301
    },
    {
      "epoch": 1.2586,
      "grad_norm": 10.722757339477539,
      "learning_rate": 1.867676767676768e-05,
      "loss": 0.1821,
      "step": 6302
    },
    {
      "epoch": 1.2587,
      "grad_norm": 8.380599975585938,
      "learning_rate": 1.8671717171717173e-05,
      "loss": 0.2679,
      "step": 6303
    },
    {
      "epoch": 1.2588,
      "grad_norm": 6.6173529624938965,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.4025,
      "step": 6304
    },
    {
      "epoch": 1.2589000000000001,
      "grad_norm": 3.554232358932495,
      "learning_rate": 1.8661616161616165e-05,
      "loss": 0.1328,
      "step": 6305
    },
    {
      "epoch": 1.259,
      "grad_norm": 3.9217259883880615,
      "learning_rate": 1.8656565656565657e-05,
      "loss": 0.0712,
      "step": 6306
    },
    {
      "epoch": 1.2591,
      "grad_norm": 24.390127182006836,
      "learning_rate": 1.8651515151515153e-05,
      "loss": 0.2193,
      "step": 6307
    },
    {
      "epoch": 1.2591999999999999,
      "grad_norm": 4.1412248611450195,
      "learning_rate": 1.864646464646465e-05,
      "loss": 0.1761,
      "step": 6308
    },
    {
      "epoch": 1.2593,
      "grad_norm": 6.738348007202148,
      "learning_rate": 1.8641414141414142e-05,
      "loss": 0.0945,
      "step": 6309
    },
    {
      "epoch": 1.2594,
      "grad_norm": 7.1068572998046875,
      "learning_rate": 1.8636363636363638e-05,
      "loss": 0.1636,
      "step": 6310
    },
    {
      "epoch": 1.2595,
      "grad_norm": 13.094812393188477,
      "learning_rate": 1.8631313131313134e-05,
      "loss": 0.1387,
      "step": 6311
    },
    {
      "epoch": 1.2596,
      "grad_norm": 5.0602240562438965,
      "learning_rate": 1.8626262626262626e-05,
      "loss": 0.1586,
      "step": 6312
    },
    {
      "epoch": 1.2597,
      "grad_norm": 2.206672430038452,
      "learning_rate": 1.8621212121212122e-05,
      "loss": 0.0394,
      "step": 6313
    },
    {
      "epoch": 1.2598,
      "grad_norm": 7.849245071411133,
      "learning_rate": 1.861616161616162e-05,
      "loss": 0.2306,
      "step": 6314
    },
    {
      "epoch": 1.2599,
      "grad_norm": 1.782950758934021,
      "learning_rate": 1.861111111111111e-05,
      "loss": 0.0596,
      "step": 6315
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.349958419799805,
      "learning_rate": 1.8606060606060607e-05,
      "loss": 0.0761,
      "step": 6316
    },
    {
      "epoch": 1.2601,
      "grad_norm": 3.876981735229492,
      "learning_rate": 1.8601010101010103e-05,
      "loss": 0.0919,
      "step": 6317
    },
    {
      "epoch": 1.2602,
      "grad_norm": 6.572679042816162,
      "learning_rate": 1.8595959595959596e-05,
      "loss": 0.082,
      "step": 6318
    },
    {
      "epoch": 1.2603,
      "grad_norm": 8.483963012695312,
      "learning_rate": 1.859090909090909e-05,
      "loss": 0.1403,
      "step": 6319
    },
    {
      "epoch": 1.2604,
      "grad_norm": 1.8397881984710693,
      "learning_rate": 1.8585858585858588e-05,
      "loss": 0.0867,
      "step": 6320
    },
    {
      "epoch": 1.2605,
      "grad_norm": 2.2622294425964355,
      "learning_rate": 1.8580808080808084e-05,
      "loss": 0.0459,
      "step": 6321
    },
    {
      "epoch": 1.2606,
      "grad_norm": 5.374152183532715,
      "learning_rate": 1.8575757575757576e-05,
      "loss": 0.0689,
      "step": 6322
    },
    {
      "epoch": 1.2607,
      "grad_norm": 1.1306732892990112,
      "learning_rate": 1.8570707070707072e-05,
      "loss": 0.022,
      "step": 6323
    },
    {
      "epoch": 1.2608,
      "grad_norm": 4.414181232452393,
      "learning_rate": 1.8565656565656568e-05,
      "loss": 0.2377,
      "step": 6324
    },
    {
      "epoch": 1.2609,
      "grad_norm": 2.8654868602752686,
      "learning_rate": 1.856060606060606e-05,
      "loss": 0.0754,
      "step": 6325
    },
    {
      "epoch": 1.2610000000000001,
      "grad_norm": 1.7293665409088135,
      "learning_rate": 1.8555555555555557e-05,
      "loss": 0.0476,
      "step": 6326
    },
    {
      "epoch": 1.2610999999999999,
      "grad_norm": 2.7324860095977783,
      "learning_rate": 1.8550505050505053e-05,
      "loss": 0.0915,
      "step": 6327
    },
    {
      "epoch": 1.2612,
      "grad_norm": 2.4365217685699463,
      "learning_rate": 1.8545454545454545e-05,
      "loss": 0.0127,
      "step": 6328
    },
    {
      "epoch": 1.2612999999999999,
      "grad_norm": 7.356318950653076,
      "learning_rate": 1.854040404040404e-05,
      "loss": 0.1683,
      "step": 6329
    },
    {
      "epoch": 1.2614,
      "grad_norm": 4.0801897048950195,
      "learning_rate": 1.8535353535353537e-05,
      "loss": 0.3045,
      "step": 6330
    },
    {
      "epoch": 1.2615,
      "grad_norm": 1.2136008739471436,
      "learning_rate": 1.853030303030303e-05,
      "loss": 0.0172,
      "step": 6331
    },
    {
      "epoch": 1.2616,
      "grad_norm": 6.575067520141602,
      "learning_rate": 1.8525252525252526e-05,
      "loss": 0.2488,
      "step": 6332
    },
    {
      "epoch": 1.2617,
      "grad_norm": 5.131925582885742,
      "learning_rate": 1.8520202020202022e-05,
      "loss": 0.1438,
      "step": 6333
    },
    {
      "epoch": 1.2618,
      "grad_norm": 4.161630630493164,
      "learning_rate": 1.8515151515151514e-05,
      "loss": 0.0983,
      "step": 6334
    },
    {
      "epoch": 1.2619,
      "grad_norm": 1.7737208604812622,
      "learning_rate": 1.851010101010101e-05,
      "loss": 0.0187,
      "step": 6335
    },
    {
      "epoch": 1.262,
      "grad_norm": 4.848827838897705,
      "learning_rate": 1.8505050505050506e-05,
      "loss": 0.1571,
      "step": 6336
    },
    {
      "epoch": 1.2621,
      "grad_norm": 1.6318329572677612,
      "learning_rate": 1.85e-05,
      "loss": 0.0685,
      "step": 6337
    },
    {
      "epoch": 1.2622,
      "grad_norm": 4.483133792877197,
      "learning_rate": 1.8494949494949495e-05,
      "loss": 0.0781,
      "step": 6338
    },
    {
      "epoch": 1.2623,
      "grad_norm": 1.9471728801727295,
      "learning_rate": 1.848989898989899e-05,
      "loss": 0.047,
      "step": 6339
    },
    {
      "epoch": 1.2624,
      "grad_norm": 4.567713260650635,
      "learning_rate": 1.8484848484848487e-05,
      "loss": 0.1993,
      "step": 6340
    },
    {
      "epoch": 1.2625,
      "grad_norm": 4.713589668273926,
      "learning_rate": 1.847979797979798e-05,
      "loss": 0.137,
      "step": 6341
    },
    {
      "epoch": 1.2626,
      "grad_norm": 2.6855804920196533,
      "learning_rate": 1.8474747474747476e-05,
      "loss": 0.0143,
      "step": 6342
    },
    {
      "epoch": 1.2627,
      "grad_norm": 2.8090341091156006,
      "learning_rate": 1.846969696969697e-05,
      "loss": 0.0519,
      "step": 6343
    },
    {
      "epoch": 1.2628,
      "grad_norm": 5.014294624328613,
      "learning_rate": 1.8464646464646464e-05,
      "loss": 0.2162,
      "step": 6344
    },
    {
      "epoch": 1.2629000000000001,
      "grad_norm": 2.389406681060791,
      "learning_rate": 1.845959595959596e-05,
      "loss": 0.0179,
      "step": 6345
    },
    {
      "epoch": 1.263,
      "grad_norm": 2.9241394996643066,
      "learning_rate": 1.8454545454545456e-05,
      "loss": 0.0696,
      "step": 6346
    },
    {
      "epoch": 1.2631000000000001,
      "grad_norm": 5.2900495529174805,
      "learning_rate": 1.844949494949495e-05,
      "loss": 0.1463,
      "step": 6347
    },
    {
      "epoch": 1.2631999999999999,
      "grad_norm": 32.04691696166992,
      "learning_rate": 1.8444444444444445e-05,
      "loss": 0.2666,
      "step": 6348
    },
    {
      "epoch": 1.2633,
      "grad_norm": 0.8339242935180664,
      "learning_rate": 1.843939393939394e-05,
      "loss": 0.0168,
      "step": 6349
    },
    {
      "epoch": 1.2634,
      "grad_norm": 21.20917510986328,
      "learning_rate": 1.8434343434343433e-05,
      "loss": 0.4634,
      "step": 6350
    },
    {
      "epoch": 1.2635,
      "grad_norm": 4.137879848480225,
      "learning_rate": 1.842929292929293e-05,
      "loss": 0.1578,
      "step": 6351
    },
    {
      "epoch": 1.2636,
      "grad_norm": 0.8036760091781616,
      "learning_rate": 1.8424242424242425e-05,
      "loss": 0.0063,
      "step": 6352
    },
    {
      "epoch": 1.2637,
      "grad_norm": 1.9028263092041016,
      "learning_rate": 1.8419191919191918e-05,
      "loss": 0.0412,
      "step": 6353
    },
    {
      "epoch": 1.2638,
      "grad_norm": 15.248907089233398,
      "learning_rate": 1.8414141414141414e-05,
      "loss": 0.1995,
      "step": 6354
    },
    {
      "epoch": 1.2639,
      "grad_norm": 9.265148162841797,
      "learning_rate": 1.840909090909091e-05,
      "loss": 0.4659,
      "step": 6355
    },
    {
      "epoch": 1.264,
      "grad_norm": 1.3965951204299927,
      "learning_rate": 1.8404040404040403e-05,
      "loss": 0.0405,
      "step": 6356
    },
    {
      "epoch": 1.2641,
      "grad_norm": 2.5238988399505615,
      "learning_rate": 1.83989898989899e-05,
      "loss": 0.0549,
      "step": 6357
    },
    {
      "epoch": 1.2642,
      "grad_norm": 4.028842449188232,
      "learning_rate": 1.8393939393939395e-05,
      "loss": 0.062,
      "step": 6358
    },
    {
      "epoch": 1.2643,
      "grad_norm": 4.342535495758057,
      "learning_rate": 1.838888888888889e-05,
      "loss": 0.0614,
      "step": 6359
    },
    {
      "epoch": 1.2644,
      "grad_norm": 11.37582015991211,
      "learning_rate": 1.8383838383838383e-05,
      "loss": 0.3052,
      "step": 6360
    },
    {
      "epoch": 1.2645,
      "grad_norm": 7.021024227142334,
      "learning_rate": 1.837878787878788e-05,
      "loss": 0.2325,
      "step": 6361
    },
    {
      "epoch": 1.2646,
      "grad_norm": 1.6001437902450562,
      "learning_rate": 1.8373737373737375e-05,
      "loss": 0.0565,
      "step": 6362
    },
    {
      "epoch": 1.2647,
      "grad_norm": 16.752002716064453,
      "learning_rate": 1.8368686868686868e-05,
      "loss": 0.586,
      "step": 6363
    },
    {
      "epoch": 1.2648,
      "grad_norm": 3.044355869293213,
      "learning_rate": 1.8363636363636364e-05,
      "loss": 0.0256,
      "step": 6364
    },
    {
      "epoch": 1.2649,
      "grad_norm": 1.956779956817627,
      "learning_rate": 1.835858585858586e-05,
      "loss": 0.1048,
      "step": 6365
    },
    {
      "epoch": 1.2650000000000001,
      "grad_norm": 8.065112113952637,
      "learning_rate": 1.8353535353535352e-05,
      "loss": 0.3403,
      "step": 6366
    },
    {
      "epoch": 1.2651,
      "grad_norm": 7.293356895446777,
      "learning_rate": 1.8348484848484848e-05,
      "loss": 0.3142,
      "step": 6367
    },
    {
      "epoch": 1.2652,
      "grad_norm": 2.6521072387695312,
      "learning_rate": 1.8343434343434344e-05,
      "loss": 0.0918,
      "step": 6368
    },
    {
      "epoch": 1.2652999999999999,
      "grad_norm": 2.431319236755371,
      "learning_rate": 1.8338383838383837e-05,
      "loss": 0.0822,
      "step": 6369
    },
    {
      "epoch": 1.2654,
      "grad_norm": 2.05281400680542,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.0348,
      "step": 6370
    },
    {
      "epoch": 1.2655,
      "grad_norm": 10.45687484741211,
      "learning_rate": 1.832828282828283e-05,
      "loss": 0.0431,
      "step": 6371
    },
    {
      "epoch": 1.2656,
      "grad_norm": 3.5949437618255615,
      "learning_rate": 1.832323232323232e-05,
      "loss": 0.1103,
      "step": 6372
    },
    {
      "epoch": 1.2657,
      "grad_norm": 6.111138343811035,
      "learning_rate": 1.831818181818182e-05,
      "loss": 0.1893,
      "step": 6373
    },
    {
      "epoch": 1.2658,
      "grad_norm": 7.132574081420898,
      "learning_rate": 1.8313131313131317e-05,
      "loss": 0.2524,
      "step": 6374
    },
    {
      "epoch": 1.2659,
      "grad_norm": 2.4085593223571777,
      "learning_rate": 1.830808080808081e-05,
      "loss": 0.0424,
      "step": 6375
    },
    {
      "epoch": 1.266,
      "grad_norm": 18.49786949157715,
      "learning_rate": 1.8303030303030305e-05,
      "loss": 0.5554,
      "step": 6376
    },
    {
      "epoch": 1.2661,
      "grad_norm": 3.0487442016601562,
      "learning_rate": 1.82979797979798e-05,
      "loss": 0.1702,
      "step": 6377
    },
    {
      "epoch": 1.2662,
      "grad_norm": 8.864136695861816,
      "learning_rate": 1.8292929292929294e-05,
      "loss": 0.2757,
      "step": 6378
    },
    {
      "epoch": 1.2663,
      "grad_norm": 2.726062297821045,
      "learning_rate": 1.828787878787879e-05,
      "loss": 0.0528,
      "step": 6379
    },
    {
      "epoch": 1.2664,
      "grad_norm": 36.16083526611328,
      "learning_rate": 1.8282828282828286e-05,
      "loss": 0.1437,
      "step": 6380
    },
    {
      "epoch": 1.2665,
      "grad_norm": 5.203984260559082,
      "learning_rate": 1.827777777777778e-05,
      "loss": 0.1688,
      "step": 6381
    },
    {
      "epoch": 1.2666,
      "grad_norm": 1.6919376850128174,
      "learning_rate": 1.8272727272727275e-05,
      "loss": 0.0498,
      "step": 6382
    },
    {
      "epoch": 1.2667,
      "grad_norm": 3.330361843109131,
      "learning_rate": 1.826767676767677e-05,
      "loss": 0.1432,
      "step": 6383
    },
    {
      "epoch": 1.2668,
      "grad_norm": 6.653658866882324,
      "learning_rate": 1.8262626262626263e-05,
      "loss": 0.161,
      "step": 6384
    },
    {
      "epoch": 1.2669000000000001,
      "grad_norm": 13.582550048828125,
      "learning_rate": 1.825757575757576e-05,
      "loss": 0.2566,
      "step": 6385
    },
    {
      "epoch": 1.267,
      "grad_norm": 30.368968963623047,
      "learning_rate": 1.8252525252525255e-05,
      "loss": 0.1722,
      "step": 6386
    },
    {
      "epoch": 1.2671000000000001,
      "grad_norm": 9.30049991607666,
      "learning_rate": 1.8247474747474748e-05,
      "loss": 0.3628,
      "step": 6387
    },
    {
      "epoch": 1.2671999999999999,
      "grad_norm": 6.964273929595947,
      "learning_rate": 1.8242424242424244e-05,
      "loss": 0.4121,
      "step": 6388
    },
    {
      "epoch": 1.2673,
      "grad_norm": 3.3769030570983887,
      "learning_rate": 1.823737373737374e-05,
      "loss": 0.1409,
      "step": 6389
    },
    {
      "epoch": 1.2674,
      "grad_norm": 6.007580757141113,
      "learning_rate": 1.8232323232323232e-05,
      "loss": 0.0576,
      "step": 6390
    },
    {
      "epoch": 1.2675,
      "grad_norm": 4.612805366516113,
      "learning_rate": 1.822727272727273e-05,
      "loss": 0.2198,
      "step": 6391
    },
    {
      "epoch": 1.2676,
      "grad_norm": 39.06838607788086,
      "learning_rate": 1.8222222222222224e-05,
      "loss": 1.2492,
      "step": 6392
    },
    {
      "epoch": 1.2677,
      "grad_norm": 39.71403503417969,
      "learning_rate": 1.821717171717172e-05,
      "loss": 1.0255,
      "step": 6393
    },
    {
      "epoch": 1.2678,
      "grad_norm": 34.61613082885742,
      "learning_rate": 1.8212121212121213e-05,
      "loss": 0.7307,
      "step": 6394
    },
    {
      "epoch": 1.2679,
      "grad_norm": 6.033215522766113,
      "learning_rate": 1.820707070707071e-05,
      "loss": 0.3227,
      "step": 6395
    },
    {
      "epoch": 1.268,
      "grad_norm": 6.574586868286133,
      "learning_rate": 1.8202020202020205e-05,
      "loss": 0.0988,
      "step": 6396
    },
    {
      "epoch": 1.2681,
      "grad_norm": 3.2710561752319336,
      "learning_rate": 1.8196969696969697e-05,
      "loss": 0.0988,
      "step": 6397
    },
    {
      "epoch": 1.2682,
      "grad_norm": 7.37160587310791,
      "learning_rate": 1.8191919191919193e-05,
      "loss": 0.2507,
      "step": 6398
    },
    {
      "epoch": 1.2683,
      "grad_norm": 5.725250720977783,
      "learning_rate": 1.818686868686869e-05,
      "loss": 0.1556,
      "step": 6399
    },
    {
      "epoch": 1.2684,
      "grad_norm": 0.7680919170379639,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.0071,
      "step": 6400
    },
    {
      "epoch": 1.2685,
      "grad_norm": 2.8578009605407715,
      "learning_rate": 1.8176767676767678e-05,
      "loss": 0.0799,
      "step": 6401
    },
    {
      "epoch": 1.2686,
      "grad_norm": 5.234334945678711,
      "learning_rate": 1.8171717171717174e-05,
      "loss": 0.2861,
      "step": 6402
    },
    {
      "epoch": 1.2687,
      "grad_norm": 2.385767698287964,
      "learning_rate": 1.8166666666666667e-05,
      "loss": 0.0199,
      "step": 6403
    },
    {
      "epoch": 1.2688,
      "grad_norm": 2.22210693359375,
      "learning_rate": 1.8161616161616163e-05,
      "loss": 0.0503,
      "step": 6404
    },
    {
      "epoch": 1.2689,
      "grad_norm": 2.656566858291626,
      "learning_rate": 1.815656565656566e-05,
      "loss": 0.049,
      "step": 6405
    },
    {
      "epoch": 1.2690000000000001,
      "grad_norm": 7.506990432739258,
      "learning_rate": 1.815151515151515e-05,
      "loss": 0.2711,
      "step": 6406
    },
    {
      "epoch": 1.2691,
      "grad_norm": 24.40779685974121,
      "learning_rate": 1.8146464646464647e-05,
      "loss": 0.2979,
      "step": 6407
    },
    {
      "epoch": 1.2692,
      "grad_norm": 3.139869451522827,
      "learning_rate": 1.8141414141414143e-05,
      "loss": 0.0273,
      "step": 6408
    },
    {
      "epoch": 1.2692999999999999,
      "grad_norm": 7.55941915512085,
      "learning_rate": 1.8136363636363636e-05,
      "loss": 0.4801,
      "step": 6409
    },
    {
      "epoch": 1.2694,
      "grad_norm": 2.203887701034546,
      "learning_rate": 1.8131313131313132e-05,
      "loss": 0.0564,
      "step": 6410
    },
    {
      "epoch": 1.2695,
      "grad_norm": 6.417689323425293,
      "learning_rate": 1.8126262626262628e-05,
      "loss": 0.2412,
      "step": 6411
    },
    {
      "epoch": 1.2696,
      "grad_norm": 4.016369819641113,
      "learning_rate": 1.8121212121212124e-05,
      "loss": 0.1981,
      "step": 6412
    },
    {
      "epoch": 1.2697,
      "grad_norm": 2.8822529315948486,
      "learning_rate": 1.8116161616161616e-05,
      "loss": 0.0887,
      "step": 6413
    },
    {
      "epoch": 1.2698,
      "grad_norm": 5.952798843383789,
      "learning_rate": 1.8111111111111112e-05,
      "loss": 0.2455,
      "step": 6414
    },
    {
      "epoch": 1.2699,
      "grad_norm": 12.848515510559082,
      "learning_rate": 1.810606060606061e-05,
      "loss": 0.1013,
      "step": 6415
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.615299701690674,
      "learning_rate": 1.81010101010101e-05,
      "loss": 0.199,
      "step": 6416
    },
    {
      "epoch": 1.2701,
      "grad_norm": 5.398725509643555,
      "learning_rate": 1.8095959595959597e-05,
      "loss": 0.1557,
      "step": 6417
    },
    {
      "epoch": 1.2702,
      "grad_norm": 5.099421501159668,
      "learning_rate": 1.8090909090909093e-05,
      "loss": 0.2389,
      "step": 6418
    },
    {
      "epoch": 1.2703,
      "grad_norm": 4.352939128875732,
      "learning_rate": 1.8085858585858586e-05,
      "loss": 0.1427,
      "step": 6419
    },
    {
      "epoch": 1.2704,
      "grad_norm": 8.86424732208252,
      "learning_rate": 1.808080808080808e-05,
      "loss": 0.3067,
      "step": 6420
    },
    {
      "epoch": 1.2705,
      "grad_norm": 4.03536319732666,
      "learning_rate": 1.8075757575757578e-05,
      "loss": 0.067,
      "step": 6421
    },
    {
      "epoch": 1.2706,
      "grad_norm": 5.507758140563965,
      "learning_rate": 1.807070707070707e-05,
      "loss": 0.1767,
      "step": 6422
    },
    {
      "epoch": 1.2707,
      "grad_norm": 5.204480171203613,
      "learning_rate": 1.8065656565656566e-05,
      "loss": 0.1794,
      "step": 6423
    },
    {
      "epoch": 1.2708,
      "grad_norm": 4.476762771606445,
      "learning_rate": 1.8060606060606062e-05,
      "loss": 0.3032,
      "step": 6424
    },
    {
      "epoch": 1.2709,
      "grad_norm": 4.969857692718506,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 0.271,
      "step": 6425
    },
    {
      "epoch": 1.271,
      "grad_norm": 4.969899654388428,
      "learning_rate": 1.805050505050505e-05,
      "loss": 0.1778,
      "step": 6426
    },
    {
      "epoch": 1.2711000000000001,
      "grad_norm": 7.159968852996826,
      "learning_rate": 1.8045454545454547e-05,
      "loss": 0.3535,
      "step": 6427
    },
    {
      "epoch": 1.2711999999999999,
      "grad_norm": 4.381277561187744,
      "learning_rate": 1.804040404040404e-05,
      "loss": 0.1005,
      "step": 6428
    },
    {
      "epoch": 1.2713,
      "grad_norm": 2.4746265411376953,
      "learning_rate": 1.8035353535353535e-05,
      "loss": 0.0661,
      "step": 6429
    },
    {
      "epoch": 1.2713999999999999,
      "grad_norm": 4.832862854003906,
      "learning_rate": 1.803030303030303e-05,
      "loss": 0.1796,
      "step": 6430
    },
    {
      "epoch": 1.2715,
      "grad_norm": 3.2027320861816406,
      "learning_rate": 1.8025252525252527e-05,
      "loss": 0.0777,
      "step": 6431
    },
    {
      "epoch": 1.2716,
      "grad_norm": 3.2742016315460205,
      "learning_rate": 1.802020202020202e-05,
      "loss": 0.104,
      "step": 6432
    },
    {
      "epoch": 1.2717,
      "grad_norm": 8.31515121459961,
      "learning_rate": 1.8015151515151516e-05,
      "loss": 0.1396,
      "step": 6433
    },
    {
      "epoch": 1.2718,
      "grad_norm": 3.446011543273926,
      "learning_rate": 1.8010101010101012e-05,
      "loss": 0.1231,
      "step": 6434
    },
    {
      "epoch": 1.2719,
      "grad_norm": 11.876331329345703,
      "learning_rate": 1.8005050505050504e-05,
      "loss": 0.2159,
      "step": 6435
    },
    {
      "epoch": 1.272,
      "grad_norm": 2.914818286895752,
      "learning_rate": 1.8e-05,
      "loss": 0.0405,
      "step": 6436
    },
    {
      "epoch": 1.2721,
      "grad_norm": 6.44588041305542,
      "learning_rate": 1.7994949494949496e-05,
      "loss": 0.3235,
      "step": 6437
    },
    {
      "epoch": 1.2722,
      "grad_norm": 2.092862844467163,
      "learning_rate": 1.798989898989899e-05,
      "loss": 0.0762,
      "step": 6438
    },
    {
      "epoch": 1.2723,
      "grad_norm": 4.738187789916992,
      "learning_rate": 1.7984848484848485e-05,
      "loss": 0.0743,
      "step": 6439
    },
    {
      "epoch": 1.2724,
      "grad_norm": 4.550585746765137,
      "learning_rate": 1.797979797979798e-05,
      "loss": 0.1821,
      "step": 6440
    },
    {
      "epoch": 1.2725,
      "grad_norm": 4.238790035247803,
      "learning_rate": 1.7974747474747474e-05,
      "loss": 0.0734,
      "step": 6441
    },
    {
      "epoch": 1.2726,
      "grad_norm": 5.314282417297363,
      "learning_rate": 1.796969696969697e-05,
      "loss": 0.2169,
      "step": 6442
    },
    {
      "epoch": 1.2727,
      "grad_norm": 7.211373329162598,
      "learning_rate": 1.7964646464646466e-05,
      "loss": 0.1461,
      "step": 6443
    },
    {
      "epoch": 1.2728,
      "grad_norm": 4.156303405761719,
      "learning_rate": 1.7959595959595958e-05,
      "loss": 0.1405,
      "step": 6444
    },
    {
      "epoch": 1.2729,
      "grad_norm": 5.708178997039795,
      "learning_rate": 1.7954545454545454e-05,
      "loss": 0.2686,
      "step": 6445
    },
    {
      "epoch": 1.2730000000000001,
      "grad_norm": 1.7301126718521118,
      "learning_rate": 1.794949494949495e-05,
      "loss": 0.0354,
      "step": 6446
    },
    {
      "epoch": 1.2731,
      "grad_norm": 2.5199031829833984,
      "learning_rate": 1.7944444444444443e-05,
      "loss": 0.0538,
      "step": 6447
    },
    {
      "epoch": 1.2732,
      "grad_norm": 3.784543514251709,
      "learning_rate": 1.793939393939394e-05,
      "loss": 0.1484,
      "step": 6448
    },
    {
      "epoch": 1.2732999999999999,
      "grad_norm": 14.153253555297852,
      "learning_rate": 1.7934343434343435e-05,
      "loss": 0.2157,
      "step": 6449
    },
    {
      "epoch": 1.2734,
      "grad_norm": 4.932328224182129,
      "learning_rate": 1.7929292929292927e-05,
      "loss": 0.136,
      "step": 6450
    },
    {
      "epoch": 1.2735,
      "grad_norm": 3.4402809143066406,
      "learning_rate": 1.7924242424242423e-05,
      "loss": 0.1374,
      "step": 6451
    },
    {
      "epoch": 1.2736,
      "grad_norm": 4.480948448181152,
      "learning_rate": 1.791919191919192e-05,
      "loss": 0.2403,
      "step": 6452
    },
    {
      "epoch": 1.2737,
      "grad_norm": 4.946902751922607,
      "learning_rate": 1.7914141414141415e-05,
      "loss": 0.1106,
      "step": 6453
    },
    {
      "epoch": 1.2738,
      "grad_norm": 8.643705368041992,
      "learning_rate": 1.7909090909090908e-05,
      "loss": 0.298,
      "step": 6454
    },
    {
      "epoch": 1.2739,
      "grad_norm": 5.667552947998047,
      "learning_rate": 1.7904040404040404e-05,
      "loss": 0.1978,
      "step": 6455
    },
    {
      "epoch": 1.274,
      "grad_norm": 2.4836642742156982,
      "learning_rate": 1.78989898989899e-05,
      "loss": 0.0764,
      "step": 6456
    },
    {
      "epoch": 1.2741,
      "grad_norm": 10.118660926818848,
      "learning_rate": 1.7893939393939393e-05,
      "loss": 0.2836,
      "step": 6457
    },
    {
      "epoch": 1.2742,
      "grad_norm": 2.5441417694091797,
      "learning_rate": 1.788888888888889e-05,
      "loss": 0.104,
      "step": 6458
    },
    {
      "epoch": 1.2743,
      "grad_norm": 3.2051103115081787,
      "learning_rate": 1.7883838383838385e-05,
      "loss": 0.0559,
      "step": 6459
    },
    {
      "epoch": 1.2744,
      "grad_norm": 1.7942622900009155,
      "learning_rate": 1.787878787878788e-05,
      "loss": 0.0576,
      "step": 6460
    },
    {
      "epoch": 1.2745,
      "grad_norm": 4.381035327911377,
      "learning_rate": 1.7873737373737376e-05,
      "loss": 0.2337,
      "step": 6461
    },
    {
      "epoch": 1.2746,
      "grad_norm": 8.941059112548828,
      "learning_rate": 1.7868686868686872e-05,
      "loss": 0.4989,
      "step": 6462
    },
    {
      "epoch": 1.2747,
      "grad_norm": 6.133901119232178,
      "learning_rate": 1.7863636363636365e-05,
      "loss": 0.1673,
      "step": 6463
    },
    {
      "epoch": 1.2748,
      "grad_norm": 2.9751899242401123,
      "learning_rate": 1.785858585858586e-05,
      "loss": 0.0715,
      "step": 6464
    },
    {
      "epoch": 1.2749,
      "grad_norm": 2.854611873626709,
      "learning_rate": 1.7853535353535357e-05,
      "loss": 0.0684,
      "step": 6465
    },
    {
      "epoch": 1.275,
      "grad_norm": 2.7964019775390625,
      "learning_rate": 1.784848484848485e-05,
      "loss": 0.1189,
      "step": 6466
    },
    {
      "epoch": 1.2751000000000001,
      "grad_norm": 2.390651226043701,
      "learning_rate": 1.7843434343434346e-05,
      "loss": 0.0797,
      "step": 6467
    },
    {
      "epoch": 1.2752,
      "grad_norm": 4.166101455688477,
      "learning_rate": 1.783838383838384e-05,
      "loss": 0.1116,
      "step": 6468
    },
    {
      "epoch": 1.2753,
      "grad_norm": 3.0572900772094727,
      "learning_rate": 1.7833333333333334e-05,
      "loss": 0.0375,
      "step": 6469
    },
    {
      "epoch": 1.2753999999999999,
      "grad_norm": 59.00664138793945,
      "learning_rate": 1.782828282828283e-05,
      "loss": 0.2639,
      "step": 6470
    },
    {
      "epoch": 1.2755,
      "grad_norm": 5.632651329040527,
      "learning_rate": 1.7823232323232326e-05,
      "loss": 0.205,
      "step": 6471
    },
    {
      "epoch": 1.2756,
      "grad_norm": 5.643845081329346,
      "learning_rate": 1.781818181818182e-05,
      "loss": 0.1375,
      "step": 6472
    },
    {
      "epoch": 1.2757,
      "grad_norm": 1.525572419166565,
      "learning_rate": 1.7813131313131315e-05,
      "loss": 0.0538,
      "step": 6473
    },
    {
      "epoch": 1.2758,
      "grad_norm": 3.455275297164917,
      "learning_rate": 1.780808080808081e-05,
      "loss": 0.1156,
      "step": 6474
    },
    {
      "epoch": 1.2759,
      "grad_norm": 1.022060751914978,
      "learning_rate": 1.7803030303030303e-05,
      "loss": 0.0182,
      "step": 6475
    },
    {
      "epoch": 1.276,
      "grad_norm": 4.546195983886719,
      "learning_rate": 1.77979797979798e-05,
      "loss": 0.1202,
      "step": 6476
    },
    {
      "epoch": 1.2761,
      "grad_norm": 4.834536552429199,
      "learning_rate": 1.7792929292929295e-05,
      "loss": 0.0763,
      "step": 6477
    },
    {
      "epoch": 1.2762,
      "grad_norm": 6.882696628570557,
      "learning_rate": 1.7787878787878788e-05,
      "loss": 0.1825,
      "step": 6478
    },
    {
      "epoch": 1.2763,
      "grad_norm": 6.97481107711792,
      "learning_rate": 1.7782828282828284e-05,
      "loss": 0.1861,
      "step": 6479
    },
    {
      "epoch": 1.2764,
      "grad_norm": 4.93814754486084,
      "learning_rate": 1.777777777777778e-05,
      "loss": 0.1942,
      "step": 6480
    },
    {
      "epoch": 1.2765,
      "grad_norm": 2.183678150177002,
      "learning_rate": 1.7772727272727273e-05,
      "loss": 0.0177,
      "step": 6481
    },
    {
      "epoch": 1.2766,
      "grad_norm": 6.989164352416992,
      "learning_rate": 1.776767676767677e-05,
      "loss": 0.1698,
      "step": 6482
    },
    {
      "epoch": 1.2767,
      "grad_norm": 7.618488788604736,
      "learning_rate": 1.7762626262626265e-05,
      "loss": 0.1542,
      "step": 6483
    },
    {
      "epoch": 1.2768,
      "grad_norm": 1.9103233814239502,
      "learning_rate": 1.775757575757576e-05,
      "loss": 0.058,
      "step": 6484
    },
    {
      "epoch": 1.2769,
      "grad_norm": 4.877736568450928,
      "learning_rate": 1.7752525252525253e-05,
      "loss": 0.1036,
      "step": 6485
    },
    {
      "epoch": 1.2770000000000001,
      "grad_norm": 7.634633541107178,
      "learning_rate": 1.774747474747475e-05,
      "loss": 0.0646,
      "step": 6486
    },
    {
      "epoch": 1.2771,
      "grad_norm": 2.775357484817505,
      "learning_rate": 1.7742424242424245e-05,
      "loss": 0.0907,
      "step": 6487
    },
    {
      "epoch": 1.2772000000000001,
      "grad_norm": 1.973780870437622,
      "learning_rate": 1.7737373737373738e-05,
      "loss": 0.1034,
      "step": 6488
    },
    {
      "epoch": 1.2772999999999999,
      "grad_norm": 3.806384563446045,
      "learning_rate": 1.7732323232323234e-05,
      "loss": 0.1247,
      "step": 6489
    },
    {
      "epoch": 1.2774,
      "grad_norm": 7.117270469665527,
      "learning_rate": 1.772727272727273e-05,
      "loss": 0.2323,
      "step": 6490
    },
    {
      "epoch": 1.2775,
      "grad_norm": 3.7107036113739014,
      "learning_rate": 1.7722222222222222e-05,
      "loss": 0.1344,
      "step": 6491
    },
    {
      "epoch": 1.2776,
      "grad_norm": 0.7446317076683044,
      "learning_rate": 1.771717171717172e-05,
      "loss": 0.011,
      "step": 6492
    },
    {
      "epoch": 1.2777,
      "grad_norm": 2.723132848739624,
      "learning_rate": 1.7712121212121214e-05,
      "loss": 0.0795,
      "step": 6493
    },
    {
      "epoch": 1.2778,
      "grad_norm": 13.418065071105957,
      "learning_rate": 1.7707070707070707e-05,
      "loss": 0.1454,
      "step": 6494
    },
    {
      "epoch": 1.2779,
      "grad_norm": 4.143869400024414,
      "learning_rate": 1.7702020202020203e-05,
      "loss": 0.2009,
      "step": 6495
    },
    {
      "epoch": 1.278,
      "grad_norm": 2.1088883876800537,
      "learning_rate": 1.76969696969697e-05,
      "loss": 0.0465,
      "step": 6496
    },
    {
      "epoch": 1.2781,
      "grad_norm": 8.030044555664062,
      "learning_rate": 1.769191919191919e-05,
      "loss": 0.2181,
      "step": 6497
    },
    {
      "epoch": 1.2782,
      "grad_norm": 2.15513014793396,
      "learning_rate": 1.7686868686868687e-05,
      "loss": 0.0393,
      "step": 6498
    },
    {
      "epoch": 1.2783,
      "grad_norm": 5.545630931854248,
      "learning_rate": 1.7681818181818183e-05,
      "loss": 0.1868,
      "step": 6499
    },
    {
      "epoch": 1.2784,
      "grad_norm": 4.856614112854004,
      "learning_rate": 1.7676767676767676e-05,
      "loss": 0.1916,
      "step": 6500
    },
    {
      "epoch": 1.2785,
      "grad_norm": 2.161466360092163,
      "learning_rate": 1.7671717171717172e-05,
      "loss": 0.0435,
      "step": 6501
    },
    {
      "epoch": 1.2786,
      "grad_norm": 5.812962055206299,
      "learning_rate": 1.7666666666666668e-05,
      "loss": 0.3048,
      "step": 6502
    },
    {
      "epoch": 1.2787,
      "grad_norm": 7.997537612915039,
      "learning_rate": 1.7661616161616164e-05,
      "loss": 0.0863,
      "step": 6503
    },
    {
      "epoch": 1.2788,
      "grad_norm": 4.165771007537842,
      "learning_rate": 1.7656565656565657e-05,
      "loss": 0.1754,
      "step": 6504
    },
    {
      "epoch": 1.2789,
      "grad_norm": 0.8089017271995544,
      "learning_rate": 1.7651515151515153e-05,
      "loss": 0.0099,
      "step": 6505
    },
    {
      "epoch": 1.279,
      "grad_norm": 10.241802215576172,
      "learning_rate": 1.764646464646465e-05,
      "loss": 0.2624,
      "step": 6506
    },
    {
      "epoch": 1.2791000000000001,
      "grad_norm": 3.1839311122894287,
      "learning_rate": 1.764141414141414e-05,
      "loss": 0.1668,
      "step": 6507
    },
    {
      "epoch": 1.2792,
      "grad_norm": 6.076528549194336,
      "learning_rate": 1.7636363636363637e-05,
      "loss": 0.2552,
      "step": 6508
    },
    {
      "epoch": 1.2793,
      "grad_norm": 4.146114826202393,
      "learning_rate": 1.7631313131313133e-05,
      "loss": 0.0833,
      "step": 6509
    },
    {
      "epoch": 1.2793999999999999,
      "grad_norm": 9.021429061889648,
      "learning_rate": 1.7626262626262626e-05,
      "loss": 0.179,
      "step": 6510
    },
    {
      "epoch": 1.2795,
      "grad_norm": 5.228076457977295,
      "learning_rate": 1.7621212121212122e-05,
      "loss": 0.1869,
      "step": 6511
    },
    {
      "epoch": 1.2796,
      "grad_norm": 1.6037161350250244,
      "learning_rate": 1.7616161616161618e-05,
      "loss": 0.0593,
      "step": 6512
    },
    {
      "epoch": 1.2797,
      "grad_norm": 60.39580535888672,
      "learning_rate": 1.761111111111111e-05,
      "loss": 0.0966,
      "step": 6513
    },
    {
      "epoch": 1.2798,
      "grad_norm": 4.62702751159668,
      "learning_rate": 1.7606060606060606e-05,
      "loss": 0.1153,
      "step": 6514
    },
    {
      "epoch": 1.2799,
      "grad_norm": 6.143311023712158,
      "learning_rate": 1.7601010101010102e-05,
      "loss": 0.2839,
      "step": 6515
    },
    {
      "epoch": 1.28,
      "grad_norm": 6.776322841644287,
      "learning_rate": 1.7595959595959595e-05,
      "loss": 0.1887,
      "step": 6516
    },
    {
      "epoch": 1.2801,
      "grad_norm": 1.6979749202728271,
      "learning_rate": 1.759090909090909e-05,
      "loss": 0.0593,
      "step": 6517
    },
    {
      "epoch": 1.2802,
      "grad_norm": 4.546590805053711,
      "learning_rate": 1.7585858585858587e-05,
      "loss": 0.1283,
      "step": 6518
    },
    {
      "epoch": 1.2803,
      "grad_norm": 2.1984753608703613,
      "learning_rate": 1.758080808080808e-05,
      "loss": 0.0605,
      "step": 6519
    },
    {
      "epoch": 1.2804,
      "grad_norm": 5.556955337524414,
      "learning_rate": 1.7575757575757576e-05,
      "loss": 0.2672,
      "step": 6520
    },
    {
      "epoch": 1.2805,
      "grad_norm": 7.506209373474121,
      "learning_rate": 1.757070707070707e-05,
      "loss": 0.133,
      "step": 6521
    },
    {
      "epoch": 1.2806,
      "grad_norm": 1.5517181158065796,
      "learning_rate": 1.7565656565656564e-05,
      "loss": 0.0363,
      "step": 6522
    },
    {
      "epoch": 1.2807,
      "grad_norm": 5.00507116317749,
      "learning_rate": 1.756060606060606e-05,
      "loss": 0.1973,
      "step": 6523
    },
    {
      "epoch": 1.2808,
      "grad_norm": 9.735578536987305,
      "learning_rate": 1.7555555555555556e-05,
      "loss": 0.4132,
      "step": 6524
    },
    {
      "epoch": 1.2809,
      "grad_norm": 1.164891004562378,
      "learning_rate": 1.7550505050505052e-05,
      "loss": 0.0173,
      "step": 6525
    },
    {
      "epoch": 1.2810000000000001,
      "grad_norm": 7.285132884979248,
      "learning_rate": 1.7545454545454545e-05,
      "loss": 0.0882,
      "step": 6526
    },
    {
      "epoch": 1.2811,
      "grad_norm": 9.903358459472656,
      "learning_rate": 1.754040404040404e-05,
      "loss": 0.0733,
      "step": 6527
    },
    {
      "epoch": 1.2812000000000001,
      "grad_norm": 4.064689636230469,
      "learning_rate": 1.7535353535353537e-05,
      "loss": 0.1184,
      "step": 6528
    },
    {
      "epoch": 1.2812999999999999,
      "grad_norm": 3.356759786605835,
      "learning_rate": 1.753030303030303e-05,
      "loss": 0.1222,
      "step": 6529
    },
    {
      "epoch": 1.2814,
      "grad_norm": 4.368165493011475,
      "learning_rate": 1.7525252525252525e-05,
      "loss": 0.2875,
      "step": 6530
    },
    {
      "epoch": 1.2814999999999999,
      "grad_norm": 6.883482456207275,
      "learning_rate": 1.752020202020202e-05,
      "loss": 0.3403,
      "step": 6531
    },
    {
      "epoch": 1.2816,
      "grad_norm": 5.729964733123779,
      "learning_rate": 1.7515151515151514e-05,
      "loss": 0.0771,
      "step": 6532
    },
    {
      "epoch": 1.2817,
      "grad_norm": 5.243563175201416,
      "learning_rate": 1.751010101010101e-05,
      "loss": 0.1664,
      "step": 6533
    },
    {
      "epoch": 1.2818,
      "grad_norm": 3.31870174407959,
      "learning_rate": 1.7505050505050506e-05,
      "loss": 0.0653,
      "step": 6534
    },
    {
      "epoch": 1.2819,
      "grad_norm": 2.808016538619995,
      "learning_rate": 1.75e-05,
      "loss": 0.0471,
      "step": 6535
    },
    {
      "epoch": 1.282,
      "grad_norm": 0.8584072589874268,
      "learning_rate": 1.7494949494949494e-05,
      "loss": 0.0216,
      "step": 6536
    },
    {
      "epoch": 1.2821,
      "grad_norm": 2.5239932537078857,
      "learning_rate": 1.748989898989899e-05,
      "loss": 0.1286,
      "step": 6537
    },
    {
      "epoch": 1.2822,
      "grad_norm": 4.629832744598389,
      "learning_rate": 1.7484848484848483e-05,
      "loss": 0.1818,
      "step": 6538
    },
    {
      "epoch": 1.2823,
      "grad_norm": 6.255419731140137,
      "learning_rate": 1.747979797979798e-05,
      "loss": 0.189,
      "step": 6539
    },
    {
      "epoch": 1.2824,
      "grad_norm": 1.282387375831604,
      "learning_rate": 1.7474747474747475e-05,
      "loss": 0.0224,
      "step": 6540
    },
    {
      "epoch": 1.2825,
      "grad_norm": 6.8037028312683105,
      "learning_rate": 1.7469696969696968e-05,
      "loss": 0.146,
      "step": 6541
    },
    {
      "epoch": 1.2826,
      "grad_norm": 5.004291534423828,
      "learning_rate": 1.7464646464646464e-05,
      "loss": 0.251,
      "step": 6542
    },
    {
      "epoch": 1.2827,
      "grad_norm": 1.1298021078109741,
      "learning_rate": 1.745959595959596e-05,
      "loss": 0.0224,
      "step": 6543
    },
    {
      "epoch": 1.2828,
      "grad_norm": 2.3989923000335693,
      "learning_rate": 1.7454545454545456e-05,
      "loss": 0.0649,
      "step": 6544
    },
    {
      "epoch": 1.2829,
      "grad_norm": 7.771114349365234,
      "learning_rate": 1.744949494949495e-05,
      "loss": 0.256,
      "step": 6545
    },
    {
      "epoch": 1.283,
      "grad_norm": 9.272039413452148,
      "learning_rate": 1.7444444444444448e-05,
      "loss": 0.2028,
      "step": 6546
    },
    {
      "epoch": 1.2831000000000001,
      "grad_norm": 2.7932302951812744,
      "learning_rate": 1.743939393939394e-05,
      "loss": 0.0528,
      "step": 6547
    },
    {
      "epoch": 1.2832,
      "grad_norm": 7.250854015350342,
      "learning_rate": 1.7434343434343436e-05,
      "loss": 0.1604,
      "step": 6548
    },
    {
      "epoch": 1.2833,
      "grad_norm": 4.755906105041504,
      "learning_rate": 1.7429292929292932e-05,
      "loss": 0.1373,
      "step": 6549
    },
    {
      "epoch": 1.2833999999999999,
      "grad_norm": 6.085086822509766,
      "learning_rate": 1.7424242424242425e-05,
      "loss": 0.2844,
      "step": 6550
    },
    {
      "epoch": 1.2835,
      "grad_norm": 3.0721263885498047,
      "learning_rate": 1.741919191919192e-05,
      "loss": 0.1002,
      "step": 6551
    },
    {
      "epoch": 1.2836,
      "grad_norm": 2.6689565181732178,
      "learning_rate": 1.7414141414141417e-05,
      "loss": 0.102,
      "step": 6552
    },
    {
      "epoch": 1.2837,
      "grad_norm": 9.372248649597168,
      "learning_rate": 1.740909090909091e-05,
      "loss": 0.1409,
      "step": 6553
    },
    {
      "epoch": 1.2838,
      "grad_norm": 3.671823740005493,
      "learning_rate": 1.7404040404040405e-05,
      "loss": 0.2032,
      "step": 6554
    },
    {
      "epoch": 1.2839,
      "grad_norm": 13.202364921569824,
      "learning_rate": 1.73989898989899e-05,
      "loss": 0.382,
      "step": 6555
    },
    {
      "epoch": 1.284,
      "grad_norm": 7.915292263031006,
      "learning_rate": 1.7393939393939397e-05,
      "loss": 0.1927,
      "step": 6556
    },
    {
      "epoch": 1.2841,
      "grad_norm": 3.0660932064056396,
      "learning_rate": 1.738888888888889e-05,
      "loss": 0.2395,
      "step": 6557
    },
    {
      "epoch": 1.2842,
      "grad_norm": 13.77719783782959,
      "learning_rate": 1.7383838383838386e-05,
      "loss": 0.3267,
      "step": 6558
    },
    {
      "epoch": 1.2843,
      "grad_norm": 6.904725551605225,
      "learning_rate": 1.7378787878787882e-05,
      "loss": 0.2594,
      "step": 6559
    },
    {
      "epoch": 1.2844,
      "grad_norm": 5.63423490524292,
      "learning_rate": 1.7373737373737375e-05,
      "loss": 0.2175,
      "step": 6560
    },
    {
      "epoch": 1.2845,
      "grad_norm": 1.27860426902771,
      "learning_rate": 1.736868686868687e-05,
      "loss": 0.0731,
      "step": 6561
    },
    {
      "epoch": 1.2846,
      "grad_norm": 9.315132141113281,
      "learning_rate": 1.7363636363636366e-05,
      "loss": 0.2469,
      "step": 6562
    },
    {
      "epoch": 1.2847,
      "grad_norm": 4.288934707641602,
      "learning_rate": 1.735858585858586e-05,
      "loss": 0.151,
      "step": 6563
    },
    {
      "epoch": 1.2848,
      "grad_norm": 6.695157051086426,
      "learning_rate": 1.7353535353535355e-05,
      "loss": 0.371,
      "step": 6564
    },
    {
      "epoch": 1.2849,
      "grad_norm": 2.6856555938720703,
      "learning_rate": 1.734848484848485e-05,
      "loss": 0.0761,
      "step": 6565
    },
    {
      "epoch": 1.285,
      "grad_norm": 2.187901496887207,
      "learning_rate": 1.7343434343434344e-05,
      "loss": 0.0507,
      "step": 6566
    },
    {
      "epoch": 1.2851,
      "grad_norm": 4.814215183258057,
      "learning_rate": 1.733838383838384e-05,
      "loss": 0.168,
      "step": 6567
    },
    {
      "epoch": 1.2852000000000001,
      "grad_norm": 4.422396659851074,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.1975,
      "step": 6568
    },
    {
      "epoch": 1.2852999999999999,
      "grad_norm": 5.755982875823975,
      "learning_rate": 1.7328282828282828e-05,
      "loss": 0.1613,
      "step": 6569
    },
    {
      "epoch": 1.2854,
      "grad_norm": 3.9870946407318115,
      "learning_rate": 1.7323232323232324e-05,
      "loss": 0.1723,
      "step": 6570
    },
    {
      "epoch": 1.2854999999999999,
      "grad_norm": 6.538243293762207,
      "learning_rate": 1.731818181818182e-05,
      "loss": 0.1678,
      "step": 6571
    },
    {
      "epoch": 1.2856,
      "grad_norm": 3.624570846557617,
      "learning_rate": 1.7313131313131313e-05,
      "loss": 0.1263,
      "step": 6572
    },
    {
      "epoch": 1.2857,
      "grad_norm": 5.848352432250977,
      "learning_rate": 1.730808080808081e-05,
      "loss": 0.2172,
      "step": 6573
    },
    {
      "epoch": 1.2858,
      "grad_norm": 7.564940452575684,
      "learning_rate": 1.7303030303030305e-05,
      "loss": 0.1677,
      "step": 6574
    },
    {
      "epoch": 1.2859,
      "grad_norm": 3.360706329345703,
      "learning_rate": 1.72979797979798e-05,
      "loss": 0.1304,
      "step": 6575
    },
    {
      "epoch": 1.286,
      "grad_norm": 7.598694324493408,
      "learning_rate": 1.7292929292929293e-05,
      "loss": 0.3426,
      "step": 6576
    },
    {
      "epoch": 1.2861,
      "grad_norm": 1.43954598903656,
      "learning_rate": 1.728787878787879e-05,
      "loss": 0.0767,
      "step": 6577
    },
    {
      "epoch": 1.2862,
      "grad_norm": 2.524876117706299,
      "learning_rate": 1.7282828282828285e-05,
      "loss": 0.0568,
      "step": 6578
    },
    {
      "epoch": 1.2863,
      "grad_norm": 3.14316463470459,
      "learning_rate": 1.7277777777777778e-05,
      "loss": 0.0565,
      "step": 6579
    },
    {
      "epoch": 1.2864,
      "grad_norm": 3.8004722595214844,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 0.1951,
      "step": 6580
    },
    {
      "epoch": 1.2865,
      "grad_norm": 4.585323333740234,
      "learning_rate": 1.726767676767677e-05,
      "loss": 0.1319,
      "step": 6581
    },
    {
      "epoch": 1.2866,
      "grad_norm": 5.191866874694824,
      "learning_rate": 1.7262626262626263e-05,
      "loss": 0.2099,
      "step": 6582
    },
    {
      "epoch": 1.2867,
      "grad_norm": 2.2244467735290527,
      "learning_rate": 1.725757575757576e-05,
      "loss": 0.0479,
      "step": 6583
    },
    {
      "epoch": 1.2868,
      "grad_norm": 5.570510387420654,
      "learning_rate": 1.7252525252525255e-05,
      "loss": 0.1688,
      "step": 6584
    },
    {
      "epoch": 1.2869,
      "grad_norm": 1.4203871488571167,
      "learning_rate": 1.7247474747474747e-05,
      "loss": 0.0254,
      "step": 6585
    },
    {
      "epoch": 1.287,
      "grad_norm": 3.1566317081451416,
      "learning_rate": 1.7242424242424243e-05,
      "loss": 0.1324,
      "step": 6586
    },
    {
      "epoch": 1.2871000000000001,
      "grad_norm": 1.258914828300476,
      "learning_rate": 1.723737373737374e-05,
      "loss": 0.0762,
      "step": 6587
    },
    {
      "epoch": 1.2872,
      "grad_norm": 4.238008499145508,
      "learning_rate": 1.7232323232323232e-05,
      "loss": 0.1051,
      "step": 6588
    },
    {
      "epoch": 1.2873,
      "grad_norm": 6.283780097961426,
      "learning_rate": 1.7227272727272728e-05,
      "loss": 0.5457,
      "step": 6589
    },
    {
      "epoch": 1.2873999999999999,
      "grad_norm": 8.065881729125977,
      "learning_rate": 1.7222222222222224e-05,
      "loss": 0.2001,
      "step": 6590
    },
    {
      "epoch": 1.2875,
      "grad_norm": 4.464353084564209,
      "learning_rate": 1.7217171717171716e-05,
      "loss": 0.1433,
      "step": 6591
    },
    {
      "epoch": 1.2876,
      "grad_norm": 4.512223720550537,
      "learning_rate": 1.7212121212121212e-05,
      "loss": 0.0721,
      "step": 6592
    },
    {
      "epoch": 1.2877,
      "grad_norm": 1.7157918214797974,
      "learning_rate": 1.720707070707071e-05,
      "loss": 0.0261,
      "step": 6593
    },
    {
      "epoch": 1.2878,
      "grad_norm": 6.950909614562988,
      "learning_rate": 1.72020202020202e-05,
      "loss": 0.2547,
      "step": 6594
    },
    {
      "epoch": 1.2879,
      "grad_norm": 4.51042366027832,
      "learning_rate": 1.7196969696969697e-05,
      "loss": 0.1752,
      "step": 6595
    },
    {
      "epoch": 1.288,
      "grad_norm": 4.235248565673828,
      "learning_rate": 1.7191919191919193e-05,
      "loss": 0.1105,
      "step": 6596
    },
    {
      "epoch": 1.2881,
      "grad_norm": 4.014166831970215,
      "learning_rate": 1.718686868686869e-05,
      "loss": 0.1378,
      "step": 6597
    },
    {
      "epoch": 1.2882,
      "grad_norm": 5.544291019439697,
      "learning_rate": 1.718181818181818e-05,
      "loss": 0.2655,
      "step": 6598
    },
    {
      "epoch": 1.2883,
      "grad_norm": 4.595393180847168,
      "learning_rate": 1.7176767676767677e-05,
      "loss": 0.0858,
      "step": 6599
    },
    {
      "epoch": 1.2884,
      "grad_norm": 16.631269454956055,
      "learning_rate": 1.7171717171717173e-05,
      "loss": 0.3442,
      "step": 6600
    },
    {
      "epoch": 1.2885,
      "grad_norm": 2.939002752304077,
      "learning_rate": 1.7166666666666666e-05,
      "loss": 0.1387,
      "step": 6601
    },
    {
      "epoch": 1.2886,
      "grad_norm": 6.185410022735596,
      "learning_rate": 1.7161616161616162e-05,
      "loss": 0.0452,
      "step": 6602
    },
    {
      "epoch": 1.2887,
      "grad_norm": 33.671783447265625,
      "learning_rate": 1.7156565656565658e-05,
      "loss": 0.3866,
      "step": 6603
    },
    {
      "epoch": 1.2888,
      "grad_norm": 6.622661113739014,
      "learning_rate": 1.715151515151515e-05,
      "loss": 0.2041,
      "step": 6604
    },
    {
      "epoch": 1.2889,
      "grad_norm": 1.9900033473968506,
      "learning_rate": 1.7146464646464647e-05,
      "loss": 0.0271,
      "step": 6605
    },
    {
      "epoch": 1.289,
      "grad_norm": 5.347684383392334,
      "learning_rate": 1.7141414141414143e-05,
      "loss": 0.0857,
      "step": 6606
    },
    {
      "epoch": 1.2891,
      "grad_norm": 1.615678071975708,
      "learning_rate": 1.7136363636363635e-05,
      "loss": 0.0482,
      "step": 6607
    },
    {
      "epoch": 1.2892000000000001,
      "grad_norm": 2.7353146076202393,
      "learning_rate": 1.713131313131313e-05,
      "loss": 0.0445,
      "step": 6608
    },
    {
      "epoch": 1.2893,
      "grad_norm": 3.7659530639648438,
      "learning_rate": 1.7126262626262627e-05,
      "loss": 0.1704,
      "step": 6609
    },
    {
      "epoch": 1.2894,
      "grad_norm": 1.8091179132461548,
      "learning_rate": 1.712121212121212e-05,
      "loss": 0.0401,
      "step": 6610
    },
    {
      "epoch": 1.2894999999999999,
      "grad_norm": 2.232652425765991,
      "learning_rate": 1.7116161616161616e-05,
      "loss": 0.0282,
      "step": 6611
    },
    {
      "epoch": 1.2896,
      "grad_norm": 16.239675521850586,
      "learning_rate": 1.7111111111111112e-05,
      "loss": 0.6242,
      "step": 6612
    },
    {
      "epoch": 1.2897,
      "grad_norm": 4.625651836395264,
      "learning_rate": 1.7106060606060604e-05,
      "loss": 0.2007,
      "step": 6613
    },
    {
      "epoch": 1.2898,
      "grad_norm": 9.564905166625977,
      "learning_rate": 1.71010101010101e-05,
      "loss": 0.4712,
      "step": 6614
    },
    {
      "epoch": 1.2899,
      "grad_norm": 12.70189380645752,
      "learning_rate": 1.7095959595959596e-05,
      "loss": 0.3859,
      "step": 6615
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.626531600952148,
      "learning_rate": 1.7090909090909092e-05,
      "loss": 0.1338,
      "step": 6616
    },
    {
      "epoch": 1.2901,
      "grad_norm": 2.4386157989501953,
      "learning_rate": 1.7085858585858585e-05,
      "loss": 0.0342,
      "step": 6617
    },
    {
      "epoch": 1.2902,
      "grad_norm": 3.211615562438965,
      "learning_rate": 1.708080808080808e-05,
      "loss": 0.0869,
      "step": 6618
    },
    {
      "epoch": 1.2903,
      "grad_norm": 6.2923970222473145,
      "learning_rate": 1.7075757575757577e-05,
      "loss": 0.2401,
      "step": 6619
    },
    {
      "epoch": 1.2904,
      "grad_norm": 3.994316816329956,
      "learning_rate": 1.707070707070707e-05,
      "loss": 0.0553,
      "step": 6620
    },
    {
      "epoch": 1.2905,
      "grad_norm": 3.917008638381958,
      "learning_rate": 1.7065656565656566e-05,
      "loss": 0.1227,
      "step": 6621
    },
    {
      "epoch": 1.2906,
      "grad_norm": 6.289501190185547,
      "learning_rate": 1.706060606060606e-05,
      "loss": 0.1413,
      "step": 6622
    },
    {
      "epoch": 1.2907,
      "grad_norm": 3.9912219047546387,
      "learning_rate": 1.7055555555555554e-05,
      "loss": 0.1163,
      "step": 6623
    },
    {
      "epoch": 1.2908,
      "grad_norm": 2.7376134395599365,
      "learning_rate": 1.705050505050505e-05,
      "loss": 0.0833,
      "step": 6624
    },
    {
      "epoch": 1.2909,
      "grad_norm": 5.965973854064941,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.2185,
      "step": 6625
    },
    {
      "epoch": 1.291,
      "grad_norm": 9.81987190246582,
      "learning_rate": 1.704040404040404e-05,
      "loss": 0.2852,
      "step": 6626
    },
    {
      "epoch": 1.2911000000000001,
      "grad_norm": 4.111846923828125,
      "learning_rate": 1.7035353535353535e-05,
      "loss": 0.1465,
      "step": 6627
    },
    {
      "epoch": 1.2912,
      "grad_norm": 8.863398551940918,
      "learning_rate": 1.703030303030303e-05,
      "loss": 0.2511,
      "step": 6628
    },
    {
      "epoch": 1.2913000000000001,
      "grad_norm": 24.628440856933594,
      "learning_rate": 1.7025252525252527e-05,
      "loss": 0.1778,
      "step": 6629
    },
    {
      "epoch": 1.2913999999999999,
      "grad_norm": 4.185924530029297,
      "learning_rate": 1.7020202020202023e-05,
      "loss": 0.2254,
      "step": 6630
    },
    {
      "epoch": 1.2915,
      "grad_norm": 3.620344400405884,
      "learning_rate": 1.701515151515152e-05,
      "loss": 0.1757,
      "step": 6631
    },
    {
      "epoch": 1.2916,
      "grad_norm": 14.018586158752441,
      "learning_rate": 1.701010101010101e-05,
      "loss": 0.3155,
      "step": 6632
    },
    {
      "epoch": 1.2917,
      "grad_norm": 4.656787395477295,
      "learning_rate": 1.7005050505050507e-05,
      "loss": 0.2046,
      "step": 6633
    },
    {
      "epoch": 1.2918,
      "grad_norm": 7.087305545806885,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.2031,
      "step": 6634
    },
    {
      "epoch": 1.2919,
      "grad_norm": 1.1842877864837646,
      "learning_rate": 1.6994949494949496e-05,
      "loss": 0.0125,
      "step": 6635
    },
    {
      "epoch": 1.292,
      "grad_norm": 3.785111665725708,
      "learning_rate": 1.6989898989898992e-05,
      "loss": 0.04,
      "step": 6636
    },
    {
      "epoch": 1.2921,
      "grad_norm": 6.892505168914795,
      "learning_rate": 1.6984848484848488e-05,
      "loss": 0.4524,
      "step": 6637
    },
    {
      "epoch": 1.2922,
      "grad_norm": 3.9125235080718994,
      "learning_rate": 1.697979797979798e-05,
      "loss": 0.1607,
      "step": 6638
    },
    {
      "epoch": 1.2923,
      "grad_norm": 5.985437393188477,
      "learning_rate": 1.6974747474747476e-05,
      "loss": 0.1786,
      "step": 6639
    },
    {
      "epoch": 1.2924,
      "grad_norm": 22.370227813720703,
      "learning_rate": 1.6969696969696972e-05,
      "loss": 0.2342,
      "step": 6640
    },
    {
      "epoch": 1.2925,
      "grad_norm": 2.8891143798828125,
      "learning_rate": 1.6964646464646465e-05,
      "loss": 0.1117,
      "step": 6641
    },
    {
      "epoch": 1.2926,
      "grad_norm": 5.000120162963867,
      "learning_rate": 1.695959595959596e-05,
      "loss": 0.1454,
      "step": 6642
    },
    {
      "epoch": 1.2927,
      "grad_norm": 6.19069766998291,
      "learning_rate": 1.6954545454545457e-05,
      "loss": 0.2669,
      "step": 6643
    },
    {
      "epoch": 1.2928,
      "grad_norm": 3.757333517074585,
      "learning_rate": 1.694949494949495e-05,
      "loss": 0.1384,
      "step": 6644
    },
    {
      "epoch": 1.2929,
      "grad_norm": 2.6117238998413086,
      "learning_rate": 1.6944444444444446e-05,
      "loss": 0.0856,
      "step": 6645
    },
    {
      "epoch": 1.293,
      "grad_norm": 2.594026565551758,
      "learning_rate": 1.693939393939394e-05,
      "loss": 0.0393,
      "step": 6646
    },
    {
      "epoch": 1.2931,
      "grad_norm": 5.009931564331055,
      "learning_rate": 1.6934343434343438e-05,
      "loss": 0.2344,
      "step": 6647
    },
    {
      "epoch": 1.2932000000000001,
      "grad_norm": 3.05248761177063,
      "learning_rate": 1.692929292929293e-05,
      "loss": 0.0463,
      "step": 6648
    },
    {
      "epoch": 1.2933,
      "grad_norm": 1.1854758262634277,
      "learning_rate": 1.6924242424242426e-05,
      "loss": 0.0262,
      "step": 6649
    },
    {
      "epoch": 1.2934,
      "grad_norm": 2.6934149265289307,
      "learning_rate": 1.6919191919191922e-05,
      "loss": 0.058,
      "step": 6650
    },
    {
      "epoch": 1.2934999999999999,
      "grad_norm": 11.759737968444824,
      "learning_rate": 1.6914141414141415e-05,
      "loss": 0.1876,
      "step": 6651
    },
    {
      "epoch": 1.2936,
      "grad_norm": 1.9152570962905884,
      "learning_rate": 1.690909090909091e-05,
      "loss": 0.0715,
      "step": 6652
    },
    {
      "epoch": 1.2937,
      "grad_norm": 3.988737106323242,
      "learning_rate": 1.6904040404040407e-05,
      "loss": 0.1268,
      "step": 6653
    },
    {
      "epoch": 1.2938,
      "grad_norm": 5.513370513916016,
      "learning_rate": 1.68989898989899e-05,
      "loss": 0.1187,
      "step": 6654
    },
    {
      "epoch": 1.2939,
      "grad_norm": 9.650033950805664,
      "learning_rate": 1.6893939393939395e-05,
      "loss": 0.5028,
      "step": 6655
    },
    {
      "epoch": 1.294,
      "grad_norm": 3.1856443881988525,
      "learning_rate": 1.688888888888889e-05,
      "loss": 0.0407,
      "step": 6656
    },
    {
      "epoch": 1.2941,
      "grad_norm": 3.692202568054199,
      "learning_rate": 1.6883838383838384e-05,
      "loss": 0.1438,
      "step": 6657
    },
    {
      "epoch": 1.2942,
      "grad_norm": 3.0582275390625,
      "learning_rate": 1.687878787878788e-05,
      "loss": 0.0829,
      "step": 6658
    },
    {
      "epoch": 1.2943,
      "grad_norm": 8.451763153076172,
      "learning_rate": 1.6873737373737376e-05,
      "loss": 0.3646,
      "step": 6659
    },
    {
      "epoch": 1.2944,
      "grad_norm": 2.789213180541992,
      "learning_rate": 1.686868686868687e-05,
      "loss": 0.0505,
      "step": 6660
    },
    {
      "epoch": 1.2945,
      "grad_norm": 6.578148365020752,
      "learning_rate": 1.6863636363636364e-05,
      "loss": 0.0296,
      "step": 6661
    },
    {
      "epoch": 1.2946,
      "grad_norm": 3.3888306617736816,
      "learning_rate": 1.685858585858586e-05,
      "loss": 0.2261,
      "step": 6662
    },
    {
      "epoch": 1.2947,
      "grad_norm": 7.725007057189941,
      "learning_rate": 1.6853535353535353e-05,
      "loss": 0.5151,
      "step": 6663
    },
    {
      "epoch": 1.2948,
      "grad_norm": 7.14291524887085,
      "learning_rate": 1.684848484848485e-05,
      "loss": 0.1616,
      "step": 6664
    },
    {
      "epoch": 1.2949,
      "grad_norm": 4.743415355682373,
      "learning_rate": 1.6843434343434345e-05,
      "loss": 0.2535,
      "step": 6665
    },
    {
      "epoch": 1.295,
      "grad_norm": 12.774060249328613,
      "learning_rate": 1.683838383838384e-05,
      "loss": 0.3842,
      "step": 6666
    },
    {
      "epoch": 1.2951,
      "grad_norm": 14.582109451293945,
      "learning_rate": 1.6833333333333334e-05,
      "loss": 0.509,
      "step": 6667
    },
    {
      "epoch": 1.2952,
      "grad_norm": 7.596268653869629,
      "learning_rate": 1.682828282828283e-05,
      "loss": 0.1139,
      "step": 6668
    },
    {
      "epoch": 1.2953000000000001,
      "grad_norm": 62.3545036315918,
      "learning_rate": 1.6823232323232326e-05,
      "loss": 0.6395,
      "step": 6669
    },
    {
      "epoch": 1.2953999999999999,
      "grad_norm": 3.4531168937683105,
      "learning_rate": 1.6818181818181818e-05,
      "loss": 0.1185,
      "step": 6670
    },
    {
      "epoch": 1.2955,
      "grad_norm": 9.13180160522461,
      "learning_rate": 1.6813131313131314e-05,
      "loss": 0.2153,
      "step": 6671
    },
    {
      "epoch": 1.2955999999999999,
      "grad_norm": 7.542859077453613,
      "learning_rate": 1.680808080808081e-05,
      "loss": 0.1557,
      "step": 6672
    },
    {
      "epoch": 1.2957,
      "grad_norm": 3.1040282249450684,
      "learning_rate": 1.6803030303030303e-05,
      "loss": 0.1013,
      "step": 6673
    },
    {
      "epoch": 1.2958,
      "grad_norm": 5.223938941955566,
      "learning_rate": 1.67979797979798e-05,
      "loss": 0.1832,
      "step": 6674
    },
    {
      "epoch": 1.2959,
      "grad_norm": 2.622129201889038,
      "learning_rate": 1.6792929292929295e-05,
      "loss": 0.1088,
      "step": 6675
    },
    {
      "epoch": 1.296,
      "grad_norm": 2.723219871520996,
      "learning_rate": 1.6787878787878787e-05,
      "loss": 0.0772,
      "step": 6676
    },
    {
      "epoch": 1.2961,
      "grad_norm": 12.303427696228027,
      "learning_rate": 1.6782828282828283e-05,
      "loss": 0.3303,
      "step": 6677
    },
    {
      "epoch": 1.2962,
      "grad_norm": 4.724796772003174,
      "learning_rate": 1.677777777777778e-05,
      "loss": 0.1756,
      "step": 6678
    },
    {
      "epoch": 1.2963,
      "grad_norm": 3.4868781566619873,
      "learning_rate": 1.6772727272727272e-05,
      "loss": 0.0528,
      "step": 6679
    },
    {
      "epoch": 1.2964,
      "grad_norm": 6.453860759735107,
      "learning_rate": 1.6767676767676768e-05,
      "loss": 0.2351,
      "step": 6680
    },
    {
      "epoch": 1.2965,
      "grad_norm": 130.853271484375,
      "learning_rate": 1.6762626262626264e-05,
      "loss": 0.2848,
      "step": 6681
    },
    {
      "epoch": 1.2966,
      "grad_norm": 6.641577243804932,
      "learning_rate": 1.6757575757575757e-05,
      "loss": 0.3347,
      "step": 6682
    },
    {
      "epoch": 1.2967,
      "grad_norm": 4.5535478591918945,
      "learning_rate": 1.6752525252525253e-05,
      "loss": 0.0899,
      "step": 6683
    },
    {
      "epoch": 1.2968,
      "grad_norm": 2.99111008644104,
      "learning_rate": 1.674747474747475e-05,
      "loss": 0.088,
      "step": 6684
    },
    {
      "epoch": 1.2969,
      "grad_norm": 4.835123062133789,
      "learning_rate": 1.674242424242424e-05,
      "loss": 0.2206,
      "step": 6685
    },
    {
      "epoch": 1.297,
      "grad_norm": 5.036015033721924,
      "learning_rate": 1.6737373737373737e-05,
      "loss": 0.0922,
      "step": 6686
    },
    {
      "epoch": 1.2971,
      "grad_norm": 16.216922760009766,
      "learning_rate": 1.6732323232323233e-05,
      "loss": 0.4415,
      "step": 6687
    },
    {
      "epoch": 1.2972000000000001,
      "grad_norm": 6.831161975860596,
      "learning_rate": 1.672727272727273e-05,
      "loss": 0.0929,
      "step": 6688
    },
    {
      "epoch": 1.2973,
      "grad_norm": 9.522482872009277,
      "learning_rate": 1.6722222222222222e-05,
      "loss": 0.1777,
      "step": 6689
    },
    {
      "epoch": 1.2974,
      "grad_norm": 10.829620361328125,
      "learning_rate": 1.6717171717171718e-05,
      "loss": 0.5097,
      "step": 6690
    },
    {
      "epoch": 1.2974999999999999,
      "grad_norm": 5.165419578552246,
      "learning_rate": 1.6712121212121214e-05,
      "loss": 0.1011,
      "step": 6691
    },
    {
      "epoch": 1.2976,
      "grad_norm": 4.977964878082275,
      "learning_rate": 1.6707070707070706e-05,
      "loss": 0.2319,
      "step": 6692
    },
    {
      "epoch": 1.2977,
      "grad_norm": 1.3810763359069824,
      "learning_rate": 1.6702020202020202e-05,
      "loss": 0.0194,
      "step": 6693
    },
    {
      "epoch": 1.2978,
      "grad_norm": 6.427859783172607,
      "learning_rate": 1.6696969696969698e-05,
      "loss": 0.3535,
      "step": 6694
    },
    {
      "epoch": 1.2979,
      "grad_norm": 6.8155741691589355,
      "learning_rate": 1.669191919191919e-05,
      "loss": 0.1408,
      "step": 6695
    },
    {
      "epoch": 1.298,
      "grad_norm": 4.286326885223389,
      "learning_rate": 1.6686868686868687e-05,
      "loss": 0.1361,
      "step": 6696
    },
    {
      "epoch": 1.2981,
      "grad_norm": 8.819520950317383,
      "learning_rate": 1.6681818181818183e-05,
      "loss": 0.3444,
      "step": 6697
    },
    {
      "epoch": 1.2982,
      "grad_norm": 1.405287265777588,
      "learning_rate": 1.6676767676767675e-05,
      "loss": 0.0292,
      "step": 6698
    },
    {
      "epoch": 1.2983,
      "grad_norm": 3.6132559776306152,
      "learning_rate": 1.667171717171717e-05,
      "loss": 0.1542,
      "step": 6699
    },
    {
      "epoch": 1.2984,
      "grad_norm": 6.009148120880127,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2273,
      "step": 6700
    },
    {
      "epoch": 1.2985,
      "grad_norm": 0.9532508254051208,
      "learning_rate": 1.666161616161616e-05,
      "loss": 0.0241,
      "step": 6701
    },
    {
      "epoch": 1.2986,
      "grad_norm": 12.008414268493652,
      "learning_rate": 1.6656565656565656e-05,
      "loss": 0.2166,
      "step": 6702
    },
    {
      "epoch": 1.2987,
      "grad_norm": 4.674522876739502,
      "learning_rate": 1.6651515151515152e-05,
      "loss": 0.072,
      "step": 6703
    },
    {
      "epoch": 1.2988,
      "grad_norm": 3.582273006439209,
      "learning_rate": 1.6646464646464645e-05,
      "loss": 0.1183,
      "step": 6704
    },
    {
      "epoch": 1.2989,
      "grad_norm": 3.7157142162323,
      "learning_rate": 1.664141414141414e-05,
      "loss": 0.1187,
      "step": 6705
    },
    {
      "epoch": 1.299,
      "grad_norm": 6.273159503936768,
      "learning_rate": 1.6636363636363637e-05,
      "loss": 0.2548,
      "step": 6706
    },
    {
      "epoch": 1.2991,
      "grad_norm": 13.104290008544922,
      "learning_rate": 1.6631313131313133e-05,
      "loss": 0.5989,
      "step": 6707
    },
    {
      "epoch": 1.2992,
      "grad_norm": 4.389108180999756,
      "learning_rate": 1.6626262626262625e-05,
      "loss": 0.1675,
      "step": 6708
    },
    {
      "epoch": 1.2993000000000001,
      "grad_norm": 4.603024005889893,
      "learning_rate": 1.662121212121212e-05,
      "loss": 0.2098,
      "step": 6709
    },
    {
      "epoch": 1.2993999999999999,
      "grad_norm": 4.2344207763671875,
      "learning_rate": 1.6616161616161617e-05,
      "loss": 0.1388,
      "step": 6710
    },
    {
      "epoch": 1.2995,
      "grad_norm": 4.868001937866211,
      "learning_rate": 1.661111111111111e-05,
      "loss": 0.3552,
      "step": 6711
    },
    {
      "epoch": 1.2995999999999999,
      "grad_norm": 2.3106632232666016,
      "learning_rate": 1.6606060606060606e-05,
      "loss": 0.0106,
      "step": 6712
    },
    {
      "epoch": 1.2997,
      "grad_norm": 13.660812377929688,
      "learning_rate": 1.6601010101010102e-05,
      "loss": 0.1492,
      "step": 6713
    },
    {
      "epoch": 1.2998,
      "grad_norm": 4.678943157196045,
      "learning_rate": 1.6595959595959594e-05,
      "loss": 0.1262,
      "step": 6714
    },
    {
      "epoch": 1.2999,
      "grad_norm": 4.609287738800049,
      "learning_rate": 1.6590909090909094e-05,
      "loss": 0.0706,
      "step": 6715
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.1607770919799805,
      "learning_rate": 1.6585858585858586e-05,
      "loss": 0.0125,
      "step": 6716
    },
    {
      "epoch": 1.3001,
      "grad_norm": 1.5212444067001343,
      "learning_rate": 1.6580808080808082e-05,
      "loss": 0.0549,
      "step": 6717
    },
    {
      "epoch": 1.3002,
      "grad_norm": 3.139565944671631,
      "learning_rate": 1.657575757575758e-05,
      "loss": 0.0823,
      "step": 6718
    },
    {
      "epoch": 1.3003,
      "grad_norm": 6.947129249572754,
      "learning_rate": 1.6570707070707074e-05,
      "loss": 0.1189,
      "step": 6719
    },
    {
      "epoch": 1.3004,
      "grad_norm": 0.8543640375137329,
      "learning_rate": 1.6565656565656567e-05,
      "loss": 0.0202,
      "step": 6720
    },
    {
      "epoch": 1.3005,
      "grad_norm": 3.5757620334625244,
      "learning_rate": 1.6560606060606063e-05,
      "loss": 0.0711,
      "step": 6721
    },
    {
      "epoch": 1.3006,
      "grad_norm": 6.222956657409668,
      "learning_rate": 1.655555555555556e-05,
      "loss": 0.1484,
      "step": 6722
    },
    {
      "epoch": 1.3007,
      "grad_norm": 6.056316375732422,
      "learning_rate": 1.655050505050505e-05,
      "loss": 0.2194,
      "step": 6723
    },
    {
      "epoch": 1.3008,
      "grad_norm": 9.440180778503418,
      "learning_rate": 1.6545454545454548e-05,
      "loss": 0.049,
      "step": 6724
    },
    {
      "epoch": 1.3009,
      "grad_norm": 3.6393752098083496,
      "learning_rate": 1.6540404040404044e-05,
      "loss": 0.108,
      "step": 6725
    },
    {
      "epoch": 1.301,
      "grad_norm": 15.035807609558105,
      "learning_rate": 1.6535353535353536e-05,
      "loss": 0.4311,
      "step": 6726
    },
    {
      "epoch": 1.3011,
      "grad_norm": 2.9239187240600586,
      "learning_rate": 1.6530303030303032e-05,
      "loss": 0.0625,
      "step": 6727
    },
    {
      "epoch": 1.3012000000000001,
      "grad_norm": 7.896549224853516,
      "learning_rate": 1.6525252525252528e-05,
      "loss": 0.1588,
      "step": 6728
    },
    {
      "epoch": 1.3013,
      "grad_norm": 11.797721862792969,
      "learning_rate": 1.652020202020202e-05,
      "loss": 0.5343,
      "step": 6729
    },
    {
      "epoch": 1.3014000000000001,
      "grad_norm": 0.3558070659637451,
      "learning_rate": 1.6515151515151517e-05,
      "loss": 0.0032,
      "step": 6730
    },
    {
      "epoch": 1.3014999999999999,
      "grad_norm": 6.3100361824035645,
      "learning_rate": 1.6510101010101013e-05,
      "loss": 0.0572,
      "step": 6731
    },
    {
      "epoch": 1.3016,
      "grad_norm": 7.354369640350342,
      "learning_rate": 1.6505050505050505e-05,
      "loss": 0.2636,
      "step": 6732
    },
    {
      "epoch": 1.3017,
      "grad_norm": 2.6981148719787598,
      "learning_rate": 1.65e-05,
      "loss": 0.0413,
      "step": 6733
    },
    {
      "epoch": 1.3018,
      "grad_norm": 5.605031490325928,
      "learning_rate": 1.6494949494949497e-05,
      "loss": 0.22,
      "step": 6734
    },
    {
      "epoch": 1.3019,
      "grad_norm": 4.074124336242676,
      "learning_rate": 1.648989898989899e-05,
      "loss": 0.1805,
      "step": 6735
    },
    {
      "epoch": 1.302,
      "grad_norm": 3.6198413372039795,
      "learning_rate": 1.6484848484848486e-05,
      "loss": 0.1087,
      "step": 6736
    },
    {
      "epoch": 1.3021,
      "grad_norm": 12.802278518676758,
      "learning_rate": 1.6479797979797982e-05,
      "loss": 0.2189,
      "step": 6737
    },
    {
      "epoch": 1.3022,
      "grad_norm": 4.180410861968994,
      "learning_rate": 1.6474747474747478e-05,
      "loss": 0.1308,
      "step": 6738
    },
    {
      "epoch": 1.3023,
      "grad_norm": 1.7197654247283936,
      "learning_rate": 1.646969696969697e-05,
      "loss": 0.0527,
      "step": 6739
    },
    {
      "epoch": 1.3024,
      "grad_norm": 17.02336311340332,
      "learning_rate": 1.6464646464646466e-05,
      "loss": 0.2042,
      "step": 6740
    },
    {
      "epoch": 1.3025,
      "grad_norm": 2.069221019744873,
      "learning_rate": 1.6459595959595962e-05,
      "loss": 0.051,
      "step": 6741
    },
    {
      "epoch": 1.3026,
      "grad_norm": 9.10206413269043,
      "learning_rate": 1.6454545454545455e-05,
      "loss": 0.0667,
      "step": 6742
    },
    {
      "epoch": 1.3027,
      "grad_norm": 1.690280795097351,
      "learning_rate": 1.644949494949495e-05,
      "loss": 0.0416,
      "step": 6743
    },
    {
      "epoch": 1.3028,
      "grad_norm": 5.779150485992432,
      "learning_rate": 1.6444444444444447e-05,
      "loss": 0.2201,
      "step": 6744
    },
    {
      "epoch": 1.3029,
      "grad_norm": 4.357267379760742,
      "learning_rate": 1.643939393939394e-05,
      "loss": 0.1254,
      "step": 6745
    },
    {
      "epoch": 1.303,
      "grad_norm": 4.725241184234619,
      "learning_rate": 1.6434343434343436e-05,
      "loss": 0.0883,
      "step": 6746
    },
    {
      "epoch": 1.3031,
      "grad_norm": 10.406817436218262,
      "learning_rate": 1.642929292929293e-05,
      "loss": 0.1023,
      "step": 6747
    },
    {
      "epoch": 1.3032,
      "grad_norm": 12.20217514038086,
      "learning_rate": 1.6424242424242424e-05,
      "loss": 0.3262,
      "step": 6748
    },
    {
      "epoch": 1.3033000000000001,
      "grad_norm": 3.6077041625976562,
      "learning_rate": 1.641919191919192e-05,
      "loss": 0.1432,
      "step": 6749
    },
    {
      "epoch": 1.3034,
      "grad_norm": 6.219247817993164,
      "learning_rate": 1.6414141414141416e-05,
      "loss": 0.0863,
      "step": 6750
    },
    {
      "epoch": 1.3035,
      "grad_norm": 4.760560035705566,
      "learning_rate": 1.640909090909091e-05,
      "loss": 0.1297,
      "step": 6751
    },
    {
      "epoch": 1.3035999999999999,
      "grad_norm": 2.9668891429901123,
      "learning_rate": 1.6404040404040405e-05,
      "loss": 0.0551,
      "step": 6752
    },
    {
      "epoch": 1.3037,
      "grad_norm": 2.510004997253418,
      "learning_rate": 1.63989898989899e-05,
      "loss": 0.053,
      "step": 6753
    },
    {
      "epoch": 1.3038,
      "grad_norm": 9.858717918395996,
      "learning_rate": 1.6393939393939393e-05,
      "loss": 0.274,
      "step": 6754
    },
    {
      "epoch": 1.3039,
      "grad_norm": 6.709403038024902,
      "learning_rate": 1.638888888888889e-05,
      "loss": 0.2607,
      "step": 6755
    },
    {
      "epoch": 1.304,
      "grad_norm": 1.2353253364562988,
      "learning_rate": 1.6383838383838385e-05,
      "loss": 0.0108,
      "step": 6756
    },
    {
      "epoch": 1.3041,
      "grad_norm": 22.401241302490234,
      "learning_rate": 1.6378787878787878e-05,
      "loss": 0.2391,
      "step": 6757
    },
    {
      "epoch": 1.3042,
      "grad_norm": 3.8037424087524414,
      "learning_rate": 1.6373737373737374e-05,
      "loss": 0.0589,
      "step": 6758
    },
    {
      "epoch": 1.3043,
      "grad_norm": 6.742501258850098,
      "learning_rate": 1.636868686868687e-05,
      "loss": 0.1276,
      "step": 6759
    },
    {
      "epoch": 1.3044,
      "grad_norm": 5.188708782196045,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.1873,
      "step": 6760
    },
    {
      "epoch": 1.3045,
      "grad_norm": 9.619400978088379,
      "learning_rate": 1.635858585858586e-05,
      "loss": 0.1252,
      "step": 6761
    },
    {
      "epoch": 1.3046,
      "grad_norm": 2.7811152935028076,
      "learning_rate": 1.6353535353535354e-05,
      "loss": 0.0962,
      "step": 6762
    },
    {
      "epoch": 1.3047,
      "grad_norm": 8.56663703918457,
      "learning_rate": 1.634848484848485e-05,
      "loss": 0.3592,
      "step": 6763
    },
    {
      "epoch": 1.3048,
      "grad_norm": 4.737478733062744,
      "learning_rate": 1.6343434343434343e-05,
      "loss": 0.3799,
      "step": 6764
    },
    {
      "epoch": 1.3049,
      "grad_norm": 9.925914764404297,
      "learning_rate": 1.633838383838384e-05,
      "loss": 0.0408,
      "step": 6765
    },
    {
      "epoch": 1.305,
      "grad_norm": 2.645451784133911,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 0.0715,
      "step": 6766
    },
    {
      "epoch": 1.3051,
      "grad_norm": 4.39331579208374,
      "learning_rate": 1.6328282828282828e-05,
      "loss": 0.1853,
      "step": 6767
    },
    {
      "epoch": 1.3052000000000001,
      "grad_norm": 1.05507230758667,
      "learning_rate": 1.6323232323232324e-05,
      "loss": 0.0159,
      "step": 6768
    },
    {
      "epoch": 1.3053,
      "grad_norm": 4.206623077392578,
      "learning_rate": 1.631818181818182e-05,
      "loss": 0.2161,
      "step": 6769
    },
    {
      "epoch": 1.3054000000000001,
      "grad_norm": 2.4337692260742188,
      "learning_rate": 1.6313131313131312e-05,
      "loss": 0.0302,
      "step": 6770
    },
    {
      "epoch": 1.3054999999999999,
      "grad_norm": 4.120059967041016,
      "learning_rate": 1.6308080808080808e-05,
      "loss": 0.086,
      "step": 6771
    },
    {
      "epoch": 1.3056,
      "grad_norm": 5.180163860321045,
      "learning_rate": 1.6303030303030304e-05,
      "loss": 0.0656,
      "step": 6772
    },
    {
      "epoch": 1.3057,
      "grad_norm": 10.810601234436035,
      "learning_rate": 1.6297979797979797e-05,
      "loss": 0.7304,
      "step": 6773
    },
    {
      "epoch": 1.3058,
      "grad_norm": 2.191126823425293,
      "learning_rate": 1.6292929292929293e-05,
      "loss": 0.0706,
      "step": 6774
    },
    {
      "epoch": 1.3059,
      "grad_norm": 4.275767803192139,
      "learning_rate": 1.628787878787879e-05,
      "loss": 0.1905,
      "step": 6775
    },
    {
      "epoch": 1.306,
      "grad_norm": 2.536360502243042,
      "learning_rate": 1.628282828282828e-05,
      "loss": 0.018,
      "step": 6776
    },
    {
      "epoch": 1.3061,
      "grad_norm": 2.2760443687438965,
      "learning_rate": 1.6277777777777777e-05,
      "loss": 0.046,
      "step": 6777
    },
    {
      "epoch": 1.3062,
      "grad_norm": 8.170455932617188,
      "learning_rate": 1.6272727272727273e-05,
      "loss": 0.0853,
      "step": 6778
    },
    {
      "epoch": 1.3063,
      "grad_norm": 2.418593645095825,
      "learning_rate": 1.626767676767677e-05,
      "loss": 0.0344,
      "step": 6779
    },
    {
      "epoch": 1.3064,
      "grad_norm": 1.874571442604065,
      "learning_rate": 1.6262626262626262e-05,
      "loss": 0.0224,
      "step": 6780
    },
    {
      "epoch": 1.3065,
      "grad_norm": 4.771717071533203,
      "learning_rate": 1.6257575757575758e-05,
      "loss": 0.1923,
      "step": 6781
    },
    {
      "epoch": 1.3066,
      "grad_norm": 1.833077311515808,
      "learning_rate": 1.6252525252525254e-05,
      "loss": 0.0285,
      "step": 6782
    },
    {
      "epoch": 1.3067,
      "grad_norm": 4.002261161804199,
      "learning_rate": 1.6247474747474747e-05,
      "loss": 0.2036,
      "step": 6783
    },
    {
      "epoch": 1.3068,
      "grad_norm": 10.419198989868164,
      "learning_rate": 1.6242424242424243e-05,
      "loss": 0.0983,
      "step": 6784
    },
    {
      "epoch": 1.3069,
      "grad_norm": 68.96037292480469,
      "learning_rate": 1.623737373737374e-05,
      "loss": 0.1118,
      "step": 6785
    },
    {
      "epoch": 1.307,
      "grad_norm": 2.380129098892212,
      "learning_rate": 1.623232323232323e-05,
      "loss": 0.0416,
      "step": 6786
    },
    {
      "epoch": 1.3071,
      "grad_norm": 4.715398788452148,
      "learning_rate": 1.6227272727272727e-05,
      "loss": 0.1952,
      "step": 6787
    },
    {
      "epoch": 1.3072,
      "grad_norm": 8.28405475616455,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 0.3828,
      "step": 6788
    },
    {
      "epoch": 1.3073000000000001,
      "grad_norm": 11.609987258911133,
      "learning_rate": 1.6217171717171716e-05,
      "loss": 0.1255,
      "step": 6789
    },
    {
      "epoch": 1.3074,
      "grad_norm": 4.461781978607178,
      "learning_rate": 1.6212121212121212e-05,
      "loss": 0.1711,
      "step": 6790
    },
    {
      "epoch": 1.3075,
      "grad_norm": 6.518438816070557,
      "learning_rate": 1.6207070707070708e-05,
      "loss": 0.3642,
      "step": 6791
    },
    {
      "epoch": 1.3075999999999999,
      "grad_norm": 4.311072826385498,
      "learning_rate": 1.62020202020202e-05,
      "loss": 0.1476,
      "step": 6792
    },
    {
      "epoch": 1.3077,
      "grad_norm": 4.7429585456848145,
      "learning_rate": 1.6196969696969696e-05,
      "loss": 0.1487,
      "step": 6793
    },
    {
      "epoch": 1.3078,
      "grad_norm": 12.074556350708008,
      "learning_rate": 1.6191919191919192e-05,
      "loss": 0.159,
      "step": 6794
    },
    {
      "epoch": 1.3079,
      "grad_norm": 5.897863864898682,
      "learning_rate": 1.6186868686868685e-05,
      "loss": 0.1654,
      "step": 6795
    },
    {
      "epoch": 1.308,
      "grad_norm": 22.694183349609375,
      "learning_rate": 1.618181818181818e-05,
      "loss": 0.2967,
      "step": 6796
    },
    {
      "epoch": 1.3081,
      "grad_norm": 4.384032726287842,
      "learning_rate": 1.6176767676767677e-05,
      "loss": 0.2315,
      "step": 6797
    },
    {
      "epoch": 1.3082,
      "grad_norm": 3.9618186950683594,
      "learning_rate": 1.617171717171717e-05,
      "loss": 0.113,
      "step": 6798
    },
    {
      "epoch": 1.3083,
      "grad_norm": 6.021419048309326,
      "learning_rate": 1.6166666666666665e-05,
      "loss": 0.125,
      "step": 6799
    },
    {
      "epoch": 1.3084,
      "grad_norm": 6.768368244171143,
      "learning_rate": 1.6161616161616165e-05,
      "loss": 0.1453,
      "step": 6800
    },
    {
      "epoch": 1.3085,
      "grad_norm": 2.7346088886260986,
      "learning_rate": 1.6156565656565657e-05,
      "loss": 0.0607,
      "step": 6801
    },
    {
      "epoch": 1.3086,
      "grad_norm": 4.921850204467773,
      "learning_rate": 1.6151515151515153e-05,
      "loss": 0.2688,
      "step": 6802
    },
    {
      "epoch": 1.3087,
      "grad_norm": 7.449355125427246,
      "learning_rate": 1.614646464646465e-05,
      "loss": 0.2789,
      "step": 6803
    },
    {
      "epoch": 1.3088,
      "grad_norm": 1.5598862171173096,
      "learning_rate": 1.6141414141414142e-05,
      "loss": 0.0343,
      "step": 6804
    },
    {
      "epoch": 1.3089,
      "grad_norm": 2.031101703643799,
      "learning_rate": 1.6136363636363638e-05,
      "loss": 0.0691,
      "step": 6805
    },
    {
      "epoch": 1.309,
      "grad_norm": 5.927606105804443,
      "learning_rate": 1.6131313131313134e-05,
      "loss": 0.2297,
      "step": 6806
    },
    {
      "epoch": 1.3091,
      "grad_norm": 7.019954681396484,
      "learning_rate": 1.6126262626262627e-05,
      "loss": 0.3481,
      "step": 6807
    },
    {
      "epoch": 1.3092,
      "grad_norm": 11.68343734741211,
      "learning_rate": 1.6121212121212123e-05,
      "loss": 0.6127,
      "step": 6808
    },
    {
      "epoch": 1.3093,
      "grad_norm": 5.094001770019531,
      "learning_rate": 1.611616161616162e-05,
      "loss": 0.2698,
      "step": 6809
    },
    {
      "epoch": 1.3094000000000001,
      "grad_norm": 4.738152027130127,
      "learning_rate": 1.6111111111111115e-05,
      "loss": 0.085,
      "step": 6810
    },
    {
      "epoch": 1.3094999999999999,
      "grad_norm": 26.90009307861328,
      "learning_rate": 1.6106060606060607e-05,
      "loss": 0.17,
      "step": 6811
    },
    {
      "epoch": 1.3096,
      "grad_norm": 5.430023670196533,
      "learning_rate": 1.6101010101010103e-05,
      "loss": 0.1299,
      "step": 6812
    },
    {
      "epoch": 1.3096999999999999,
      "grad_norm": 5.151377201080322,
      "learning_rate": 1.60959595959596e-05,
      "loss": 0.257,
      "step": 6813
    },
    {
      "epoch": 1.3098,
      "grad_norm": 2.2432632446289062,
      "learning_rate": 1.6090909090909092e-05,
      "loss": 0.1068,
      "step": 6814
    },
    {
      "epoch": 1.3099,
      "grad_norm": 6.231196880340576,
      "learning_rate": 1.6085858585858588e-05,
      "loss": 0.5817,
      "step": 6815
    },
    {
      "epoch": 1.31,
      "grad_norm": 3.8012924194335938,
      "learning_rate": 1.6080808080808084e-05,
      "loss": 0.1213,
      "step": 6816
    },
    {
      "epoch": 1.3101,
      "grad_norm": 3.1970527172088623,
      "learning_rate": 1.6075757575757576e-05,
      "loss": 0.1161,
      "step": 6817
    },
    {
      "epoch": 1.3102,
      "grad_norm": 3.8190860748291016,
      "learning_rate": 1.6070707070707072e-05,
      "loss": 0.0853,
      "step": 6818
    },
    {
      "epoch": 1.3103,
      "grad_norm": 6.837596416473389,
      "learning_rate": 1.606565656565657e-05,
      "loss": 0.2329,
      "step": 6819
    },
    {
      "epoch": 1.3104,
      "grad_norm": 10.17619514465332,
      "learning_rate": 1.606060606060606e-05,
      "loss": 0.2454,
      "step": 6820
    },
    {
      "epoch": 1.3105,
      "grad_norm": 3.824507713317871,
      "learning_rate": 1.6055555555555557e-05,
      "loss": 0.1222,
      "step": 6821
    },
    {
      "epoch": 1.3106,
      "grad_norm": 2.660576343536377,
      "learning_rate": 1.6050505050505053e-05,
      "loss": 0.0961,
      "step": 6822
    },
    {
      "epoch": 1.3107,
      "grad_norm": 3.0914409160614014,
      "learning_rate": 1.6045454545454546e-05,
      "loss": 0.0949,
      "step": 6823
    },
    {
      "epoch": 1.3108,
      "grad_norm": 6.785470485687256,
      "learning_rate": 1.604040404040404e-05,
      "loss": 0.222,
      "step": 6824
    },
    {
      "epoch": 1.3109,
      "grad_norm": 6.679121494293213,
      "learning_rate": 1.6035353535353538e-05,
      "loss": 0.4798,
      "step": 6825
    },
    {
      "epoch": 1.311,
      "grad_norm": 4.815081596374512,
      "learning_rate": 1.603030303030303e-05,
      "loss": 0.2111,
      "step": 6826
    },
    {
      "epoch": 1.3111,
      "grad_norm": 5.957653522491455,
      "learning_rate": 1.6025252525252526e-05,
      "loss": 0.3306,
      "step": 6827
    },
    {
      "epoch": 1.3112,
      "grad_norm": 6.9760236740112305,
      "learning_rate": 1.6020202020202022e-05,
      "loss": 0.2779,
      "step": 6828
    },
    {
      "epoch": 1.3113000000000001,
      "grad_norm": 6.711704254150391,
      "learning_rate": 1.6015151515151515e-05,
      "loss": 0.2838,
      "step": 6829
    },
    {
      "epoch": 1.3114,
      "grad_norm": 6.436984539031982,
      "learning_rate": 1.601010101010101e-05,
      "loss": 0.3299,
      "step": 6830
    },
    {
      "epoch": 1.3115,
      "grad_norm": 12.274857521057129,
      "learning_rate": 1.6005050505050507e-05,
      "loss": 0.3279,
      "step": 6831
    },
    {
      "epoch": 1.3115999999999999,
      "grad_norm": 15.80693531036377,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.5643,
      "step": 6832
    },
    {
      "epoch": 1.3117,
      "grad_norm": 11.65902328491211,
      "learning_rate": 1.5994949494949495e-05,
      "loss": 0.531,
      "step": 6833
    },
    {
      "epoch": 1.3118,
      "grad_norm": 12.732671737670898,
      "learning_rate": 1.598989898989899e-05,
      "loss": 0.5628,
      "step": 6834
    },
    {
      "epoch": 1.3119,
      "grad_norm": 92.08818817138672,
      "learning_rate": 1.5984848484848487e-05,
      "loss": 0.5045,
      "step": 6835
    },
    {
      "epoch": 1.312,
      "grad_norm": 7.463220119476318,
      "learning_rate": 1.597979797979798e-05,
      "loss": 0.399,
      "step": 6836
    },
    {
      "epoch": 1.3121,
      "grad_norm": 6.834704875946045,
      "learning_rate": 1.5974747474747476e-05,
      "loss": 0.4104,
      "step": 6837
    },
    {
      "epoch": 1.3122,
      "grad_norm": 5.6300129890441895,
      "learning_rate": 1.5969696969696972e-05,
      "loss": 0.5641,
      "step": 6838
    },
    {
      "epoch": 1.3123,
      "grad_norm": 6.579987525939941,
      "learning_rate": 1.5964646464646464e-05,
      "loss": 0.3289,
      "step": 6839
    },
    {
      "epoch": 1.3124,
      "grad_norm": 5.309232234954834,
      "learning_rate": 1.595959595959596e-05,
      "loss": 0.2802,
      "step": 6840
    },
    {
      "epoch": 1.3125,
      "grad_norm": 5.526000499725342,
      "learning_rate": 1.5954545454545456e-05,
      "loss": 0.3385,
      "step": 6841
    },
    {
      "epoch": 1.3126,
      "grad_norm": 15.555458068847656,
      "learning_rate": 1.594949494949495e-05,
      "loss": 0.4835,
      "step": 6842
    },
    {
      "epoch": 1.3127,
      "grad_norm": 9.611800193786621,
      "learning_rate": 1.5944444444444445e-05,
      "loss": 0.4254,
      "step": 6843
    },
    {
      "epoch": 1.3128,
      "grad_norm": 19.3100528717041,
      "learning_rate": 1.593939393939394e-05,
      "loss": 0.2525,
      "step": 6844
    },
    {
      "epoch": 1.3129,
      "grad_norm": 6.585668563842773,
      "learning_rate": 1.5934343434343434e-05,
      "loss": 0.1874,
      "step": 6845
    },
    {
      "epoch": 1.313,
      "grad_norm": 7.263988971710205,
      "learning_rate": 1.592929292929293e-05,
      "loss": 0.4832,
      "step": 6846
    },
    {
      "epoch": 1.3131,
      "grad_norm": 6.425844192504883,
      "learning_rate": 1.5924242424242426e-05,
      "loss": 0.2704,
      "step": 6847
    },
    {
      "epoch": 1.3132,
      "grad_norm": 12.582218170166016,
      "learning_rate": 1.5919191919191918e-05,
      "loss": 0.2278,
      "step": 6848
    },
    {
      "epoch": 1.3133,
      "grad_norm": 11.540337562561035,
      "learning_rate": 1.5914141414141414e-05,
      "loss": 0.3019,
      "step": 6849
    },
    {
      "epoch": 1.3134000000000001,
      "grad_norm": 7.164450168609619,
      "learning_rate": 1.590909090909091e-05,
      "loss": 0.3516,
      "step": 6850
    },
    {
      "epoch": 1.3135,
      "grad_norm": 8.768320083618164,
      "learning_rate": 1.5904040404040406e-05,
      "loss": 0.3733,
      "step": 6851
    },
    {
      "epoch": 1.3136,
      "grad_norm": 12.911639213562012,
      "learning_rate": 1.58989898989899e-05,
      "loss": 0.7266,
      "step": 6852
    },
    {
      "epoch": 1.3136999999999999,
      "grad_norm": 6.521647930145264,
      "learning_rate": 1.5893939393939395e-05,
      "loss": 0.2841,
      "step": 6853
    },
    {
      "epoch": 1.3138,
      "grad_norm": 12.826120376586914,
      "learning_rate": 1.588888888888889e-05,
      "loss": 0.5369,
      "step": 6854
    },
    {
      "epoch": 1.3139,
      "grad_norm": 29.29589080810547,
      "learning_rate": 1.5883838383838383e-05,
      "loss": 0.438,
      "step": 6855
    },
    {
      "epoch": 1.314,
      "grad_norm": 7.267152309417725,
      "learning_rate": 1.587878787878788e-05,
      "loss": 0.4035,
      "step": 6856
    },
    {
      "epoch": 1.3141,
      "grad_norm": 22.415390014648438,
      "learning_rate": 1.5873737373737375e-05,
      "loss": 0.4179,
      "step": 6857
    },
    {
      "epoch": 1.3142,
      "grad_norm": 7.563183307647705,
      "learning_rate": 1.5868686868686868e-05,
      "loss": 0.6456,
      "step": 6858
    },
    {
      "epoch": 1.3143,
      "grad_norm": 5.567509651184082,
      "learning_rate": 1.5863636363636364e-05,
      "loss": 0.3825,
      "step": 6859
    },
    {
      "epoch": 1.3144,
      "grad_norm": 49.94884490966797,
      "learning_rate": 1.585858585858586e-05,
      "loss": 0.5066,
      "step": 6860
    },
    {
      "epoch": 1.3145,
      "grad_norm": 6.864365577697754,
      "learning_rate": 1.5853535353535353e-05,
      "loss": 0.3625,
      "step": 6861
    },
    {
      "epoch": 1.3146,
      "grad_norm": 11.421664237976074,
      "learning_rate": 1.584848484848485e-05,
      "loss": 0.503,
      "step": 6862
    },
    {
      "epoch": 1.3147,
      "grad_norm": 16.073999404907227,
      "learning_rate": 1.5843434343434344e-05,
      "loss": 0.6071,
      "step": 6863
    },
    {
      "epoch": 1.3148,
      "grad_norm": 8.016093254089355,
      "learning_rate": 1.5838383838383837e-05,
      "loss": 0.2985,
      "step": 6864
    },
    {
      "epoch": 1.3149,
      "grad_norm": 9.451783180236816,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.4121,
      "step": 6865
    },
    {
      "epoch": 1.315,
      "grad_norm": 11.014639854431152,
      "learning_rate": 1.582828282828283e-05,
      "loss": 0.3484,
      "step": 6866
    },
    {
      "epoch": 1.3151,
      "grad_norm": 19.107851028442383,
      "learning_rate": 1.582323232323232e-05,
      "loss": 0.6088,
      "step": 6867
    },
    {
      "epoch": 1.3152,
      "grad_norm": 7.884775161743164,
      "learning_rate": 1.5818181818181818e-05,
      "loss": 0.296,
      "step": 6868
    },
    {
      "epoch": 1.3153000000000001,
      "grad_norm": 10.868391036987305,
      "learning_rate": 1.5813131313131314e-05,
      "loss": 0.6529,
      "step": 6869
    },
    {
      "epoch": 1.3154,
      "grad_norm": 12.204669952392578,
      "learning_rate": 1.5808080808080806e-05,
      "loss": 0.3025,
      "step": 6870
    },
    {
      "epoch": 1.3155000000000001,
      "grad_norm": 5.617972373962402,
      "learning_rate": 1.5803030303030302e-05,
      "loss": 0.2463,
      "step": 6871
    },
    {
      "epoch": 1.3155999999999999,
      "grad_norm": 6.307064533233643,
      "learning_rate": 1.5797979797979798e-05,
      "loss": 0.3374,
      "step": 6872
    },
    {
      "epoch": 1.3157,
      "grad_norm": 5.0693793296813965,
      "learning_rate": 1.5792929292929294e-05,
      "loss": 0.1832,
      "step": 6873
    },
    {
      "epoch": 1.3158,
      "grad_norm": 10.983179092407227,
      "learning_rate": 1.5787878787878787e-05,
      "loss": 0.1528,
      "step": 6874
    },
    {
      "epoch": 1.3159,
      "grad_norm": 5.985663890838623,
      "learning_rate": 1.5782828282828283e-05,
      "loss": 0.2528,
      "step": 6875
    },
    {
      "epoch": 1.316,
      "grad_norm": 6.694779396057129,
      "learning_rate": 1.577777777777778e-05,
      "loss": 0.3868,
      "step": 6876
    },
    {
      "epoch": 1.3161,
      "grad_norm": 5.967103958129883,
      "learning_rate": 1.577272727272727e-05,
      "loss": 0.4378,
      "step": 6877
    },
    {
      "epoch": 1.3162,
      "grad_norm": 10.644890785217285,
      "learning_rate": 1.5767676767676767e-05,
      "loss": 0.319,
      "step": 6878
    },
    {
      "epoch": 1.3163,
      "grad_norm": 7.011590480804443,
      "learning_rate": 1.5762626262626263e-05,
      "loss": 0.337,
      "step": 6879
    },
    {
      "epoch": 1.3164,
      "grad_norm": 11.391036033630371,
      "learning_rate": 1.5757575757575756e-05,
      "loss": 0.3885,
      "step": 6880
    },
    {
      "epoch": 1.3165,
      "grad_norm": 4.883376121520996,
      "learning_rate": 1.5752525252525252e-05,
      "loss": 0.3381,
      "step": 6881
    },
    {
      "epoch": 1.3166,
      "grad_norm": 6.949097633361816,
      "learning_rate": 1.5747474747474748e-05,
      "loss": 0.2423,
      "step": 6882
    },
    {
      "epoch": 1.3167,
      "grad_norm": 7.441651821136475,
      "learning_rate": 1.574242424242424e-05,
      "loss": 0.3212,
      "step": 6883
    },
    {
      "epoch": 1.3168,
      "grad_norm": 6.1817169189453125,
      "learning_rate": 1.5737373737373737e-05,
      "loss": 0.2445,
      "step": 6884
    },
    {
      "epoch": 1.3169,
      "grad_norm": 7.769890785217285,
      "learning_rate": 1.5732323232323236e-05,
      "loss": 0.3553,
      "step": 6885
    },
    {
      "epoch": 1.317,
      "grad_norm": 24.696855545043945,
      "learning_rate": 1.572727272727273e-05,
      "loss": 0.3333,
      "step": 6886
    },
    {
      "epoch": 1.3171,
      "grad_norm": 7.905683994293213,
      "learning_rate": 1.5722222222222225e-05,
      "loss": 0.3458,
      "step": 6887
    },
    {
      "epoch": 1.3172,
      "grad_norm": 9.141756057739258,
      "learning_rate": 1.571717171717172e-05,
      "loss": 0.4694,
      "step": 6888
    },
    {
      "epoch": 1.3173,
      "grad_norm": 11.600042343139648,
      "learning_rate": 1.5712121212121213e-05,
      "loss": 0.3797,
      "step": 6889
    },
    {
      "epoch": 1.3174000000000001,
      "grad_norm": 32.25870132446289,
      "learning_rate": 1.570707070707071e-05,
      "loss": 0.4534,
      "step": 6890
    },
    {
      "epoch": 1.3175,
      "grad_norm": 12.437358856201172,
      "learning_rate": 1.5702020202020205e-05,
      "loss": 0.578,
      "step": 6891
    },
    {
      "epoch": 1.3176,
      "grad_norm": 8.089607238769531,
      "learning_rate": 1.5696969696969698e-05,
      "loss": 0.284,
      "step": 6892
    },
    {
      "epoch": 1.3176999999999999,
      "grad_norm": 11.7155179977417,
      "learning_rate": 1.5691919191919194e-05,
      "loss": 0.4136,
      "step": 6893
    },
    {
      "epoch": 1.3178,
      "grad_norm": 16.538686752319336,
      "learning_rate": 1.568686868686869e-05,
      "loss": 0.6045,
      "step": 6894
    },
    {
      "epoch": 1.3179,
      "grad_norm": 5.741827011108398,
      "learning_rate": 1.5681818181818182e-05,
      "loss": 0.3509,
      "step": 6895
    },
    {
      "epoch": 1.318,
      "grad_norm": 7.526790142059326,
      "learning_rate": 1.5676767676767678e-05,
      "loss": 0.7293,
      "step": 6896
    },
    {
      "epoch": 1.3181,
      "grad_norm": 10.481663703918457,
      "learning_rate": 1.5671717171717174e-05,
      "loss": 0.6031,
      "step": 6897
    },
    {
      "epoch": 1.3182,
      "grad_norm": 10.032752990722656,
      "learning_rate": 1.5666666666666667e-05,
      "loss": 0.7069,
      "step": 6898
    },
    {
      "epoch": 1.3183,
      "grad_norm": 7.6582350730896,
      "learning_rate": 1.5661616161616163e-05,
      "loss": 0.2278,
      "step": 6899
    },
    {
      "epoch": 1.3184,
      "grad_norm": 18.112632751464844,
      "learning_rate": 1.565656565656566e-05,
      "loss": 0.5688,
      "step": 6900
    },
    {
      "epoch": 1.3185,
      "grad_norm": 15.237825393676758,
      "learning_rate": 1.565151515151515e-05,
      "loss": 0.5426,
      "step": 6901
    },
    {
      "epoch": 1.3186,
      "grad_norm": 11.788360595703125,
      "learning_rate": 1.5646464646464647e-05,
      "loss": 0.3502,
      "step": 6902
    },
    {
      "epoch": 1.3187,
      "grad_norm": 10.479864120483398,
      "learning_rate": 1.5641414141414143e-05,
      "loss": 0.4156,
      "step": 6903
    },
    {
      "epoch": 1.3188,
      "grad_norm": 7.488223075866699,
      "learning_rate": 1.563636363636364e-05,
      "loss": 0.5362,
      "step": 6904
    },
    {
      "epoch": 1.3189,
      "grad_norm": 5.762332916259766,
      "learning_rate": 1.5631313131313132e-05,
      "loss": 0.4913,
      "step": 6905
    },
    {
      "epoch": 1.319,
      "grad_norm": 11.002002716064453,
      "learning_rate": 1.5626262626262628e-05,
      "loss": 0.6292,
      "step": 6906
    },
    {
      "epoch": 1.3191,
      "grad_norm": 35.64567947387695,
      "learning_rate": 1.5621212121212124e-05,
      "loss": 0.7251,
      "step": 6907
    },
    {
      "epoch": 1.3192,
      "grad_norm": 13.735912322998047,
      "learning_rate": 1.5616161616161617e-05,
      "loss": 0.3998,
      "step": 6908
    },
    {
      "epoch": 1.3193,
      "grad_norm": 12.070637702941895,
      "learning_rate": 1.5611111111111113e-05,
      "loss": 0.7579,
      "step": 6909
    },
    {
      "epoch": 1.3194,
      "grad_norm": 11.21473217010498,
      "learning_rate": 1.560606060606061e-05,
      "loss": 0.8832,
      "step": 6910
    },
    {
      "epoch": 1.3195000000000001,
      "grad_norm": 7.760034561157227,
      "learning_rate": 1.56010101010101e-05,
      "loss": 0.3709,
      "step": 6911
    },
    {
      "epoch": 1.3195999999999999,
      "grad_norm": 19.50244140625,
      "learning_rate": 1.5595959595959597e-05,
      "loss": 0.6042,
      "step": 6912
    },
    {
      "epoch": 1.3197,
      "grad_norm": 6.783799648284912,
      "learning_rate": 1.5590909090909093e-05,
      "loss": 0.2188,
      "step": 6913
    },
    {
      "epoch": 1.3197999999999999,
      "grad_norm": 8.605473518371582,
      "learning_rate": 1.5585858585858586e-05,
      "loss": 0.2699,
      "step": 6914
    },
    {
      "epoch": 1.3199,
      "grad_norm": 10.74212646484375,
      "learning_rate": 1.5580808080808082e-05,
      "loss": 0.3963,
      "step": 6915
    },
    {
      "epoch": 1.32,
      "grad_norm": 14.826930046081543,
      "learning_rate": 1.5575757575757578e-05,
      "loss": 0.6097,
      "step": 6916
    },
    {
      "epoch": 1.3201,
      "grad_norm": 19.367231369018555,
      "learning_rate": 1.557070707070707e-05,
      "loss": 1.1954,
      "step": 6917
    },
    {
      "epoch": 1.3202,
      "grad_norm": 11.016959190368652,
      "learning_rate": 1.5565656565656566e-05,
      "loss": 0.3711,
      "step": 6918
    },
    {
      "epoch": 1.3203,
      "grad_norm": 12.254639625549316,
      "learning_rate": 1.5560606060606062e-05,
      "loss": 0.3384,
      "step": 6919
    },
    {
      "epoch": 1.3204,
      "grad_norm": 8.689528465270996,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.3067,
      "step": 6920
    },
    {
      "epoch": 1.3205,
      "grad_norm": 7.498450756072998,
      "learning_rate": 1.555050505050505e-05,
      "loss": 0.5237,
      "step": 6921
    },
    {
      "epoch": 1.3206,
      "grad_norm": 10.165263175964355,
      "learning_rate": 1.5545454545454547e-05,
      "loss": 0.322,
      "step": 6922
    },
    {
      "epoch": 1.3207,
      "grad_norm": 7.435948848724365,
      "learning_rate": 1.5540404040404043e-05,
      "loss": 0.2497,
      "step": 6923
    },
    {
      "epoch": 1.3208,
      "grad_norm": 7.822391986846924,
      "learning_rate": 1.5535353535353536e-05,
      "loss": 0.3097,
      "step": 6924
    },
    {
      "epoch": 1.3209,
      "grad_norm": 11.451165199279785,
      "learning_rate": 1.553030303030303e-05,
      "loss": 0.5617,
      "step": 6925
    },
    {
      "epoch": 1.321,
      "grad_norm": 6.2169270515441895,
      "learning_rate": 1.5525252525252527e-05,
      "loss": 0.4139,
      "step": 6926
    },
    {
      "epoch": 1.3211,
      "grad_norm": 8.907374382019043,
      "learning_rate": 1.552020202020202e-05,
      "loss": 0.3899,
      "step": 6927
    },
    {
      "epoch": 1.3212,
      "grad_norm": 13.258091926574707,
      "learning_rate": 1.5515151515151516e-05,
      "loss": 0.4752,
      "step": 6928
    },
    {
      "epoch": 1.3213,
      "grad_norm": 6.219761848449707,
      "learning_rate": 1.5510101010101012e-05,
      "loss": 0.2367,
      "step": 6929
    },
    {
      "epoch": 1.3214000000000001,
      "grad_norm": 7.699864387512207,
      "learning_rate": 1.5505050505050505e-05,
      "loss": 0.3542,
      "step": 6930
    },
    {
      "epoch": 1.3215,
      "grad_norm": 16.037235260009766,
      "learning_rate": 1.55e-05,
      "loss": 0.2634,
      "step": 6931
    },
    {
      "epoch": 1.3216,
      "grad_norm": 12.823858261108398,
      "learning_rate": 1.5494949494949497e-05,
      "loss": 0.2743,
      "step": 6932
    },
    {
      "epoch": 1.3216999999999999,
      "grad_norm": 6.373583793640137,
      "learning_rate": 1.548989898989899e-05,
      "loss": 0.2488,
      "step": 6933
    },
    {
      "epoch": 1.3218,
      "grad_norm": 11.156543731689453,
      "learning_rate": 1.5484848484848485e-05,
      "loss": 0.3856,
      "step": 6934
    },
    {
      "epoch": 1.3219,
      "grad_norm": 7.416475772857666,
      "learning_rate": 1.547979797979798e-05,
      "loss": 0.3863,
      "step": 6935
    },
    {
      "epoch": 1.322,
      "grad_norm": 11.666166305541992,
      "learning_rate": 1.5474747474747474e-05,
      "loss": 0.7559,
      "step": 6936
    },
    {
      "epoch": 1.3221,
      "grad_norm": 7.48766565322876,
      "learning_rate": 1.546969696969697e-05,
      "loss": 0.3436,
      "step": 6937
    },
    {
      "epoch": 1.3222,
      "grad_norm": 8.316155433654785,
      "learning_rate": 1.5464646464646466e-05,
      "loss": 0.366,
      "step": 6938
    },
    {
      "epoch": 1.3223,
      "grad_norm": 5.862684726715088,
      "learning_rate": 1.545959595959596e-05,
      "loss": 0.1427,
      "step": 6939
    },
    {
      "epoch": 1.3224,
      "grad_norm": 7.319891929626465,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.3365,
      "step": 6940
    },
    {
      "epoch": 1.3225,
      "grad_norm": 18.95326042175293,
      "learning_rate": 1.544949494949495e-05,
      "loss": 0.6689,
      "step": 6941
    },
    {
      "epoch": 1.3226,
      "grad_norm": 11.289694786071777,
      "learning_rate": 1.5444444444444446e-05,
      "loss": 0.2808,
      "step": 6942
    },
    {
      "epoch": 1.3227,
      "grad_norm": 10.24682903289795,
      "learning_rate": 1.543939393939394e-05,
      "loss": 0.4026,
      "step": 6943
    },
    {
      "epoch": 1.3228,
      "grad_norm": 17.04966926574707,
      "learning_rate": 1.5434343434343435e-05,
      "loss": 0.6824,
      "step": 6944
    },
    {
      "epoch": 1.3229,
      "grad_norm": 26.325603485107422,
      "learning_rate": 1.542929292929293e-05,
      "loss": 0.3604,
      "step": 6945
    },
    {
      "epoch": 1.323,
      "grad_norm": 9.0492582321167,
      "learning_rate": 1.5424242424242424e-05,
      "loss": 0.4816,
      "step": 6946
    },
    {
      "epoch": 1.3231,
      "grad_norm": 28.14038848876953,
      "learning_rate": 1.541919191919192e-05,
      "loss": 0.5582,
      "step": 6947
    },
    {
      "epoch": 1.3232,
      "grad_norm": 7.029725074768066,
      "learning_rate": 1.5414141414141416e-05,
      "loss": 0.2565,
      "step": 6948
    },
    {
      "epoch": 1.3233,
      "grad_norm": 43.09664535522461,
      "learning_rate": 1.5409090909090908e-05,
      "loss": 0.435,
      "step": 6949
    },
    {
      "epoch": 1.3234,
      "grad_norm": 6.067574977874756,
      "learning_rate": 1.5404040404040404e-05,
      "loss": 0.4315,
      "step": 6950
    },
    {
      "epoch": 1.3235000000000001,
      "grad_norm": 11.499271392822266,
      "learning_rate": 1.53989898989899e-05,
      "loss": 0.3193,
      "step": 6951
    },
    {
      "epoch": 1.3235999999999999,
      "grad_norm": 8.634714126586914,
      "learning_rate": 1.5393939393939393e-05,
      "loss": 0.272,
      "step": 6952
    },
    {
      "epoch": 1.3237,
      "grad_norm": 6.811501979827881,
      "learning_rate": 1.538888888888889e-05,
      "loss": 0.2612,
      "step": 6953
    },
    {
      "epoch": 1.3237999999999999,
      "grad_norm": 7.461156368255615,
      "learning_rate": 1.5383838383838385e-05,
      "loss": 0.3898,
      "step": 6954
    },
    {
      "epoch": 1.3239,
      "grad_norm": 9.737293243408203,
      "learning_rate": 1.5378787878787877e-05,
      "loss": 0.4035,
      "step": 6955
    },
    {
      "epoch": 1.324,
      "grad_norm": 10.880868911743164,
      "learning_rate": 1.5373737373737373e-05,
      "loss": 0.2338,
      "step": 6956
    },
    {
      "epoch": 1.3241,
      "grad_norm": 4.879293918609619,
      "learning_rate": 1.536868686868687e-05,
      "loss": 0.2449,
      "step": 6957
    },
    {
      "epoch": 1.3242,
      "grad_norm": 9.621926307678223,
      "learning_rate": 1.5363636363636362e-05,
      "loss": 0.3792,
      "step": 6958
    },
    {
      "epoch": 1.3243,
      "grad_norm": 11.543575286865234,
      "learning_rate": 1.5358585858585858e-05,
      "loss": 0.3066,
      "step": 6959
    },
    {
      "epoch": 1.3244,
      "grad_norm": 8.255009651184082,
      "learning_rate": 1.5353535353535354e-05,
      "loss": 0.3058,
      "step": 6960
    },
    {
      "epoch": 1.3245,
      "grad_norm": 7.611567974090576,
      "learning_rate": 1.5348484848484847e-05,
      "loss": 0.3458,
      "step": 6961
    },
    {
      "epoch": 1.3246,
      "grad_norm": 11.870185852050781,
      "learning_rate": 1.5343434343434342e-05,
      "loss": 0.2058,
      "step": 6962
    },
    {
      "epoch": 1.3247,
      "grad_norm": 7.310112476348877,
      "learning_rate": 1.533838383838384e-05,
      "loss": 0.3889,
      "step": 6963
    },
    {
      "epoch": 1.3248,
      "grad_norm": 16.451339721679688,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 0.5061,
      "step": 6964
    },
    {
      "epoch": 1.3249,
      "grad_norm": 6.072748184204102,
      "learning_rate": 1.5328282828282827e-05,
      "loss": 0.2396,
      "step": 6965
    },
    {
      "epoch": 1.325,
      "grad_norm": 7.136501789093018,
      "learning_rate": 1.5323232323232323e-05,
      "loss": 0.3123,
      "step": 6966
    },
    {
      "epoch": 1.3251,
      "grad_norm": 8.860017776489258,
      "learning_rate": 1.531818181818182e-05,
      "loss": 0.3318,
      "step": 6967
    },
    {
      "epoch": 1.3252,
      "grad_norm": 8.653266906738281,
      "learning_rate": 1.531313131313131e-05,
      "loss": 0.3574,
      "step": 6968
    },
    {
      "epoch": 1.3253,
      "grad_norm": 18.63471794128418,
      "learning_rate": 1.5308080808080808e-05,
      "loss": 0.7207,
      "step": 6969
    },
    {
      "epoch": 1.3254000000000001,
      "grad_norm": 74.73147583007812,
      "learning_rate": 1.5303030303030304e-05,
      "loss": 0.3896,
      "step": 6970
    },
    {
      "epoch": 1.3255,
      "grad_norm": 9.430298805236816,
      "learning_rate": 1.52979797979798e-05,
      "loss": 0.3247,
      "step": 6971
    },
    {
      "epoch": 1.3256000000000001,
      "grad_norm": 17.4604434967041,
      "learning_rate": 1.5292929292929296e-05,
      "loss": 0.3316,
      "step": 6972
    },
    {
      "epoch": 1.3256999999999999,
      "grad_norm": 8.841449737548828,
      "learning_rate": 1.5287878787878788e-05,
      "loss": 0.568,
      "step": 6973
    },
    {
      "epoch": 1.3258,
      "grad_norm": 11.975850105285645,
      "learning_rate": 1.5282828282828284e-05,
      "loss": 0.3738,
      "step": 6974
    },
    {
      "epoch": 1.3259,
      "grad_norm": 4.980841636657715,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.2606,
      "step": 6975
    },
    {
      "epoch": 1.326,
      "grad_norm": 38.98058319091797,
      "learning_rate": 1.5272727272727276e-05,
      "loss": 0.42,
      "step": 6976
    },
    {
      "epoch": 1.3261,
      "grad_norm": 9.678878784179688,
      "learning_rate": 1.526767676767677e-05,
      "loss": 0.4661,
      "step": 6977
    },
    {
      "epoch": 1.3262,
      "grad_norm": 13.423295974731445,
      "learning_rate": 1.5262626262626265e-05,
      "loss": 0.3314,
      "step": 6978
    },
    {
      "epoch": 1.3263,
      "grad_norm": 26.99964141845703,
      "learning_rate": 1.5257575757575759e-05,
      "loss": 0.8333,
      "step": 6979
    },
    {
      "epoch": 1.3264,
      "grad_norm": 8.262094497680664,
      "learning_rate": 1.5252525252525255e-05,
      "loss": 0.4591,
      "step": 6980
    },
    {
      "epoch": 1.3265,
      "grad_norm": 9.900596618652344,
      "learning_rate": 1.524747474747475e-05,
      "loss": 0.4286,
      "step": 6981
    },
    {
      "epoch": 1.3266,
      "grad_norm": 5.685843467712402,
      "learning_rate": 1.5242424242424244e-05,
      "loss": 0.2957,
      "step": 6982
    },
    {
      "epoch": 1.3267,
      "grad_norm": 17.035274505615234,
      "learning_rate": 1.523737373737374e-05,
      "loss": 0.3482,
      "step": 6983
    },
    {
      "epoch": 1.3268,
      "grad_norm": 9.083319664001465,
      "learning_rate": 1.5232323232323234e-05,
      "loss": 0.3087,
      "step": 6984
    },
    {
      "epoch": 1.3269,
      "grad_norm": 5.871020793914795,
      "learning_rate": 1.5227272727272728e-05,
      "loss": 0.2464,
      "step": 6985
    },
    {
      "epoch": 1.327,
      "grad_norm": 11.147168159484863,
      "learning_rate": 1.5222222222222224e-05,
      "loss": 0.2461,
      "step": 6986
    },
    {
      "epoch": 1.3271,
      "grad_norm": 10.456708908081055,
      "learning_rate": 1.5217171717171719e-05,
      "loss": 0.3528,
      "step": 6987
    },
    {
      "epoch": 1.3272,
      "grad_norm": 8.780186653137207,
      "learning_rate": 1.5212121212121213e-05,
      "loss": 0.3028,
      "step": 6988
    },
    {
      "epoch": 1.3273,
      "grad_norm": 5.155318260192871,
      "learning_rate": 1.5207070707070709e-05,
      "loss": 0.2996,
      "step": 6989
    },
    {
      "epoch": 1.3274,
      "grad_norm": 14.766261100769043,
      "learning_rate": 1.5202020202020203e-05,
      "loss": 0.4488,
      "step": 6990
    },
    {
      "epoch": 1.3275000000000001,
      "grad_norm": 7.049752712249756,
      "learning_rate": 1.5196969696969699e-05,
      "loss": 0.2682,
      "step": 6991
    },
    {
      "epoch": 1.3276,
      "grad_norm": 13.129105567932129,
      "learning_rate": 1.5191919191919193e-05,
      "loss": 0.5227,
      "step": 6992
    },
    {
      "epoch": 1.3277,
      "grad_norm": 8.8102445602417,
      "learning_rate": 1.5186868686868688e-05,
      "loss": 0.3639,
      "step": 6993
    },
    {
      "epoch": 1.3277999999999999,
      "grad_norm": 5.676969528198242,
      "learning_rate": 1.5181818181818184e-05,
      "loss": 0.2441,
      "step": 6994
    },
    {
      "epoch": 1.3279,
      "grad_norm": 8.403861045837402,
      "learning_rate": 1.5176767676767678e-05,
      "loss": 0.3632,
      "step": 6995
    },
    {
      "epoch": 1.328,
      "grad_norm": 5.114412784576416,
      "learning_rate": 1.5171717171717172e-05,
      "loss": 0.3254,
      "step": 6996
    },
    {
      "epoch": 1.3281,
      "grad_norm": 29.888439178466797,
      "learning_rate": 1.5166666666666668e-05,
      "loss": 0.4544,
      "step": 6997
    },
    {
      "epoch": 1.3282,
      "grad_norm": 12.728256225585938,
      "learning_rate": 1.5161616161616163e-05,
      "loss": 0.55,
      "step": 6998
    },
    {
      "epoch": 1.3283,
      "grad_norm": 10.039166450500488,
      "learning_rate": 1.5156565656565659e-05,
      "loss": 0.3584,
      "step": 6999
    },
    {
      "epoch": 1.3284,
      "grad_norm": 8.052762031555176,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.3756,
      "step": 7000
    },
    {
      "epoch": 1.3285,
      "grad_norm": 11.360738754272461,
      "learning_rate": 1.5146464646464647e-05,
      "loss": 0.3615,
      "step": 7001
    },
    {
      "epoch": 1.3286,
      "grad_norm": 17.30747413635254,
      "learning_rate": 1.5141414141414143e-05,
      "loss": 0.3469,
      "step": 7002
    },
    {
      "epoch": 1.3287,
      "grad_norm": 10.565643310546875,
      "learning_rate": 1.5136363636363637e-05,
      "loss": 0.2711,
      "step": 7003
    },
    {
      "epoch": 1.3288,
      "grad_norm": 16.967573165893555,
      "learning_rate": 1.5131313131313132e-05,
      "loss": 0.6103,
      "step": 7004
    },
    {
      "epoch": 1.3289,
      "grad_norm": 9.857536315917969,
      "learning_rate": 1.5126262626262628e-05,
      "loss": 0.4448,
      "step": 7005
    },
    {
      "epoch": 1.329,
      "grad_norm": 8.723343849182129,
      "learning_rate": 1.5121212121212122e-05,
      "loss": 0.4756,
      "step": 7006
    },
    {
      "epoch": 1.3291,
      "grad_norm": 15.661434173583984,
      "learning_rate": 1.5116161616161616e-05,
      "loss": 0.4996,
      "step": 7007
    },
    {
      "epoch": 1.3292,
      "grad_norm": 11.513143539428711,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 0.5652,
      "step": 7008
    },
    {
      "epoch": 1.3293,
      "grad_norm": 9.52767276763916,
      "learning_rate": 1.5106060606060607e-05,
      "loss": 0.3824,
      "step": 7009
    },
    {
      "epoch": 1.3294000000000001,
      "grad_norm": 21.515727996826172,
      "learning_rate": 1.5101010101010103e-05,
      "loss": 0.3158,
      "step": 7010
    },
    {
      "epoch": 1.3295,
      "grad_norm": 7.2036356925964355,
      "learning_rate": 1.5095959595959597e-05,
      "loss": 0.253,
      "step": 7011
    },
    {
      "epoch": 1.3296000000000001,
      "grad_norm": 7.792147636413574,
      "learning_rate": 1.5090909090909091e-05,
      "loss": 0.4134,
      "step": 7012
    },
    {
      "epoch": 1.3296999999999999,
      "grad_norm": 15.279303550720215,
      "learning_rate": 1.5085858585858587e-05,
      "loss": 0.4322,
      "step": 7013
    },
    {
      "epoch": 1.3298,
      "grad_norm": 8.304242134094238,
      "learning_rate": 1.5080808080808081e-05,
      "loss": 0.318,
      "step": 7014
    },
    {
      "epoch": 1.3299,
      "grad_norm": 7.149102687835693,
      "learning_rate": 1.5075757575757576e-05,
      "loss": 0.2364,
      "step": 7015
    },
    {
      "epoch": 1.33,
      "grad_norm": 81.13360595703125,
      "learning_rate": 1.5070707070707072e-05,
      "loss": 0.3957,
      "step": 7016
    },
    {
      "epoch": 1.3301,
      "grad_norm": 6.772754669189453,
      "learning_rate": 1.5065656565656566e-05,
      "loss": 0.2827,
      "step": 7017
    },
    {
      "epoch": 1.3302,
      "grad_norm": 7.213513374328613,
      "learning_rate": 1.506060606060606e-05,
      "loss": 0.208,
      "step": 7018
    },
    {
      "epoch": 1.3303,
      "grad_norm": 13.128242492675781,
      "learning_rate": 1.5055555555555556e-05,
      "loss": 0.537,
      "step": 7019
    },
    {
      "epoch": 1.3304,
      "grad_norm": 6.145074844360352,
      "learning_rate": 1.505050505050505e-05,
      "loss": 0.3089,
      "step": 7020
    },
    {
      "epoch": 1.3305,
      "grad_norm": 13.052054405212402,
      "learning_rate": 1.5045454545454547e-05,
      "loss": 0.932,
      "step": 7021
    },
    {
      "epoch": 1.3306,
      "grad_norm": 10.329078674316406,
      "learning_rate": 1.5040404040404041e-05,
      "loss": 0.5348,
      "step": 7022
    },
    {
      "epoch": 1.3307,
      "grad_norm": 8.976799964904785,
      "learning_rate": 1.5035353535353535e-05,
      "loss": 0.6157,
      "step": 7023
    },
    {
      "epoch": 1.3308,
      "grad_norm": 11.812976837158203,
      "learning_rate": 1.5030303030303031e-05,
      "loss": 0.5228,
      "step": 7024
    },
    {
      "epoch": 1.3309,
      "grad_norm": 6.647338390350342,
      "learning_rate": 1.5025252525252526e-05,
      "loss": 0.299,
      "step": 7025
    },
    {
      "epoch": 1.331,
      "grad_norm": 6.595182418823242,
      "learning_rate": 1.502020202020202e-05,
      "loss": 0.219,
      "step": 7026
    },
    {
      "epoch": 1.3311,
      "grad_norm": 7.369240760803223,
      "learning_rate": 1.5015151515151516e-05,
      "loss": 0.3283,
      "step": 7027
    },
    {
      "epoch": 1.3312,
      "grad_norm": 21.277551651000977,
      "learning_rate": 1.501010101010101e-05,
      "loss": 0.4222,
      "step": 7028
    },
    {
      "epoch": 1.3313,
      "grad_norm": 15.383933067321777,
      "learning_rate": 1.5005050505050506e-05,
      "loss": 0.3328,
      "step": 7029
    },
    {
      "epoch": 1.3314,
      "grad_norm": 19.0538330078125,
      "learning_rate": 1.5e-05,
      "loss": 1.1112,
      "step": 7030
    },
    {
      "epoch": 1.3315000000000001,
      "grad_norm": 7.614649295806885,
      "learning_rate": 1.4994949494949495e-05,
      "loss": 0.3264,
      "step": 7031
    },
    {
      "epoch": 1.3316,
      "grad_norm": 15.862857818603516,
      "learning_rate": 1.498989898989899e-05,
      "loss": 0.7104,
      "step": 7032
    },
    {
      "epoch": 1.3317,
      "grad_norm": 28.222763061523438,
      "learning_rate": 1.4984848484848485e-05,
      "loss": 0.5231,
      "step": 7033
    },
    {
      "epoch": 1.3317999999999999,
      "grad_norm": 9.448585510253906,
      "learning_rate": 1.497979797979798e-05,
      "loss": 0.2308,
      "step": 7034
    },
    {
      "epoch": 1.3319,
      "grad_norm": 9.720707893371582,
      "learning_rate": 1.4974747474747475e-05,
      "loss": 0.3076,
      "step": 7035
    },
    {
      "epoch": 1.332,
      "grad_norm": 11.52583122253418,
      "learning_rate": 1.496969696969697e-05,
      "loss": 0.358,
      "step": 7036
    },
    {
      "epoch": 1.3321,
      "grad_norm": 7.281243324279785,
      "learning_rate": 1.4964646464646464e-05,
      "loss": 0.3781,
      "step": 7037
    },
    {
      "epoch": 1.3322,
      "grad_norm": 4.499144077301025,
      "learning_rate": 1.495959595959596e-05,
      "loss": 0.2584,
      "step": 7038
    },
    {
      "epoch": 1.3323,
      "grad_norm": 7.483189582824707,
      "learning_rate": 1.4954545454545454e-05,
      "loss": 0.3219,
      "step": 7039
    },
    {
      "epoch": 1.3324,
      "grad_norm": 8.231513977050781,
      "learning_rate": 1.494949494949495e-05,
      "loss": 0.3761,
      "step": 7040
    },
    {
      "epoch": 1.3325,
      "grad_norm": 5.513689994812012,
      "learning_rate": 1.4944444444444444e-05,
      "loss": 0.3165,
      "step": 7041
    },
    {
      "epoch": 1.3326,
      "grad_norm": 4.85574197769165,
      "learning_rate": 1.4939393939393939e-05,
      "loss": 0.2183,
      "step": 7042
    },
    {
      "epoch": 1.3327,
      "grad_norm": 12.4844331741333,
      "learning_rate": 1.4934343434343435e-05,
      "loss": 0.3322,
      "step": 7043
    },
    {
      "epoch": 1.3328,
      "grad_norm": 5.4588093757629395,
      "learning_rate": 1.4929292929292929e-05,
      "loss": 0.1452,
      "step": 7044
    },
    {
      "epoch": 1.3329,
      "grad_norm": 13.687373161315918,
      "learning_rate": 1.4924242424242423e-05,
      "loss": 0.2568,
      "step": 7045
    },
    {
      "epoch": 1.333,
      "grad_norm": 10.980308532714844,
      "learning_rate": 1.491919191919192e-05,
      "loss": 0.5149,
      "step": 7046
    },
    {
      "epoch": 1.3331,
      "grad_norm": 7.108087539672852,
      "learning_rate": 1.4914141414141414e-05,
      "loss": 0.4159,
      "step": 7047
    },
    {
      "epoch": 1.3332,
      "grad_norm": 6.79282283782959,
      "learning_rate": 1.4909090909090908e-05,
      "loss": 0.3078,
      "step": 7048
    },
    {
      "epoch": 1.3333,
      "grad_norm": 23.29188346862793,
      "learning_rate": 1.4904040404040404e-05,
      "loss": 0.7037,
      "step": 7049
    },
    {
      "epoch": 1.3334,
      "grad_norm": 7.949595928192139,
      "learning_rate": 1.4898989898989898e-05,
      "loss": 0.366,
      "step": 7050
    },
    {
      "epoch": 1.3335,
      "grad_norm": 11.763601303100586,
      "learning_rate": 1.4893939393939394e-05,
      "loss": 0.3427,
      "step": 7051
    },
    {
      "epoch": 1.3336000000000001,
      "grad_norm": 9.650727272033691,
      "learning_rate": 1.4888888888888888e-05,
      "loss": 0.3033,
      "step": 7052
    },
    {
      "epoch": 1.3336999999999999,
      "grad_norm": 7.447687149047852,
      "learning_rate": 1.4883838383838383e-05,
      "loss": 0.4298,
      "step": 7053
    },
    {
      "epoch": 1.3338,
      "grad_norm": 6.745293617248535,
      "learning_rate": 1.4878787878787879e-05,
      "loss": 0.3673,
      "step": 7054
    },
    {
      "epoch": 1.3338999999999999,
      "grad_norm": 18.108579635620117,
      "learning_rate": 1.4873737373737373e-05,
      "loss": 0.396,
      "step": 7055
    },
    {
      "epoch": 1.334,
      "grad_norm": 7.617610454559326,
      "learning_rate": 1.486868686868687e-05,
      "loss": 0.2335,
      "step": 7056
    },
    {
      "epoch": 1.3341,
      "grad_norm": 9.16643238067627,
      "learning_rate": 1.4863636363636365e-05,
      "loss": 0.1789,
      "step": 7057
    },
    {
      "epoch": 1.3342,
      "grad_norm": 6.459829807281494,
      "learning_rate": 1.4858585858585861e-05,
      "loss": 0.2031,
      "step": 7058
    },
    {
      "epoch": 1.3343,
      "grad_norm": 19.315433502197266,
      "learning_rate": 1.4853535353535355e-05,
      "loss": 1.2374,
      "step": 7059
    },
    {
      "epoch": 1.3344,
      "grad_norm": 14.12862777709961,
      "learning_rate": 1.484848484848485e-05,
      "loss": 0.819,
      "step": 7060
    },
    {
      "epoch": 1.3345,
      "grad_norm": 11.95677661895752,
      "learning_rate": 1.4843434343434346e-05,
      "loss": 0.4511,
      "step": 7061
    },
    {
      "epoch": 1.3346,
      "grad_norm": 6.735055446624756,
      "learning_rate": 1.483838383838384e-05,
      "loss": 0.3094,
      "step": 7062
    },
    {
      "epoch": 1.3347,
      "grad_norm": 9.74269962310791,
      "learning_rate": 1.4833333333333336e-05,
      "loss": 0.1602,
      "step": 7063
    },
    {
      "epoch": 1.3348,
      "grad_norm": 10.205225944519043,
      "learning_rate": 1.482828282828283e-05,
      "loss": 0.2049,
      "step": 7064
    },
    {
      "epoch": 1.3349,
      "grad_norm": 10.29328727722168,
      "learning_rate": 1.4823232323232324e-05,
      "loss": 0.6651,
      "step": 7065
    },
    {
      "epoch": 1.335,
      "grad_norm": 11.852298736572266,
      "learning_rate": 1.481818181818182e-05,
      "loss": 0.245,
      "step": 7066
    },
    {
      "epoch": 1.3351,
      "grad_norm": 7.601961135864258,
      "learning_rate": 1.4813131313131315e-05,
      "loss": 0.2192,
      "step": 7067
    },
    {
      "epoch": 1.3352,
      "grad_norm": 7.522674083709717,
      "learning_rate": 1.4808080808080809e-05,
      "loss": 0.4946,
      "step": 7068
    },
    {
      "epoch": 1.3353,
      "grad_norm": 4.629284381866455,
      "learning_rate": 1.4803030303030305e-05,
      "loss": 0.2191,
      "step": 7069
    },
    {
      "epoch": 1.3354,
      "grad_norm": 9.47840404510498,
      "learning_rate": 1.47979797979798e-05,
      "loss": 0.2546,
      "step": 7070
    },
    {
      "epoch": 1.3355000000000001,
      "grad_norm": 7.415660381317139,
      "learning_rate": 1.4792929292929295e-05,
      "loss": 0.5006,
      "step": 7071
    },
    {
      "epoch": 1.3356,
      "grad_norm": 3.651665687561035,
      "learning_rate": 1.478787878787879e-05,
      "loss": 0.1634,
      "step": 7072
    },
    {
      "epoch": 1.3357,
      "grad_norm": 3.9932851791381836,
      "learning_rate": 1.4782828282828284e-05,
      "loss": 0.1177,
      "step": 7073
    },
    {
      "epoch": 1.3357999999999999,
      "grad_norm": 17.11578941345215,
      "learning_rate": 1.477777777777778e-05,
      "loss": 1.2939,
      "step": 7074
    },
    {
      "epoch": 1.3359,
      "grad_norm": 12.1766939163208,
      "learning_rate": 1.4772727272727274e-05,
      "loss": 0.5322,
      "step": 7075
    },
    {
      "epoch": 1.336,
      "grad_norm": 7.5638909339904785,
      "learning_rate": 1.4767676767676769e-05,
      "loss": 0.5426,
      "step": 7076
    },
    {
      "epoch": 1.3361,
      "grad_norm": 10.499726295471191,
      "learning_rate": 1.4762626262626264e-05,
      "loss": 0.8691,
      "step": 7077
    },
    {
      "epoch": 1.3362,
      "grad_norm": 13.197108268737793,
      "learning_rate": 1.4757575757575759e-05,
      "loss": 0.1891,
      "step": 7078
    },
    {
      "epoch": 1.3363,
      "grad_norm": 6.745004653930664,
      "learning_rate": 1.4752525252525253e-05,
      "loss": 0.151,
      "step": 7079
    },
    {
      "epoch": 1.3364,
      "grad_norm": 9.82958698272705,
      "learning_rate": 1.4747474747474749e-05,
      "loss": 0.2771,
      "step": 7080
    },
    {
      "epoch": 1.3365,
      "grad_norm": 3.46368670463562,
      "learning_rate": 1.4742424242424243e-05,
      "loss": 0.1862,
      "step": 7081
    },
    {
      "epoch": 1.3366,
      "grad_norm": 5.193665027618408,
      "learning_rate": 1.473737373737374e-05,
      "loss": 0.3243,
      "step": 7082
    },
    {
      "epoch": 1.3367,
      "grad_norm": 54.16807556152344,
      "learning_rate": 1.4732323232323234e-05,
      "loss": 0.251,
      "step": 7083
    },
    {
      "epoch": 1.3368,
      "grad_norm": 6.304310321807861,
      "learning_rate": 1.4727272727272728e-05,
      "loss": 0.2372,
      "step": 7084
    },
    {
      "epoch": 1.3369,
      "grad_norm": 12.067259788513184,
      "learning_rate": 1.4722222222222224e-05,
      "loss": 0.8767,
      "step": 7085
    },
    {
      "epoch": 1.337,
      "grad_norm": 9.478843688964844,
      "learning_rate": 1.4717171717171718e-05,
      "loss": 0.2952,
      "step": 7086
    },
    {
      "epoch": 1.3371,
      "grad_norm": 7.056292533874512,
      "learning_rate": 1.4712121212121213e-05,
      "loss": 0.2524,
      "step": 7087
    },
    {
      "epoch": 1.3372,
      "grad_norm": 5.730135917663574,
      "learning_rate": 1.4707070707070709e-05,
      "loss": 0.3243,
      "step": 7088
    },
    {
      "epoch": 1.3373,
      "grad_norm": 12.462244033813477,
      "learning_rate": 1.4702020202020203e-05,
      "loss": 0.2509,
      "step": 7089
    },
    {
      "epoch": 1.3374,
      "grad_norm": 11.290921211242676,
      "learning_rate": 1.4696969696969697e-05,
      "loss": 0.2873,
      "step": 7090
    },
    {
      "epoch": 1.3375,
      "grad_norm": 18.202932357788086,
      "learning_rate": 1.4691919191919193e-05,
      "loss": 1.1337,
      "step": 7091
    },
    {
      "epoch": 1.3376000000000001,
      "grad_norm": 6.745841979980469,
      "learning_rate": 1.4686868686868687e-05,
      "loss": 0.1924,
      "step": 7092
    },
    {
      "epoch": 1.3377,
      "grad_norm": 7.692257881164551,
      "learning_rate": 1.4681818181818183e-05,
      "loss": 0.1374,
      "step": 7093
    },
    {
      "epoch": 1.3378,
      "grad_norm": 5.943739891052246,
      "learning_rate": 1.4676767676767678e-05,
      "loss": 0.2891,
      "step": 7094
    },
    {
      "epoch": 1.3378999999999999,
      "grad_norm": 5.213412761688232,
      "learning_rate": 1.4671717171717172e-05,
      "loss": 0.2907,
      "step": 7095
    },
    {
      "epoch": 1.338,
      "grad_norm": 9.123315811157227,
      "learning_rate": 1.4666666666666668e-05,
      "loss": 0.3103,
      "step": 7096
    },
    {
      "epoch": 1.3381,
      "grad_norm": 7.308578014373779,
      "learning_rate": 1.4661616161616162e-05,
      "loss": 0.2709,
      "step": 7097
    },
    {
      "epoch": 1.3382,
      "grad_norm": 6.402060031890869,
      "learning_rate": 1.4656565656565657e-05,
      "loss": 0.2905,
      "step": 7098
    },
    {
      "epoch": 1.3383,
      "grad_norm": 14.999517440795898,
      "learning_rate": 1.4651515151515153e-05,
      "loss": 0.2658,
      "step": 7099
    },
    {
      "epoch": 1.3384,
      "grad_norm": 4.872513771057129,
      "learning_rate": 1.4646464646464647e-05,
      "loss": 0.1232,
      "step": 7100
    },
    {
      "epoch": 1.3385,
      "grad_norm": 3.8031537532806396,
      "learning_rate": 1.4641414141414143e-05,
      "loss": 0.1585,
      "step": 7101
    },
    {
      "epoch": 1.3386,
      "grad_norm": 9.843697547912598,
      "learning_rate": 1.4636363636363637e-05,
      "loss": 0.1945,
      "step": 7102
    },
    {
      "epoch": 1.3387,
      "grad_norm": 7.637648582458496,
      "learning_rate": 1.4631313131313131e-05,
      "loss": 0.2764,
      "step": 7103
    },
    {
      "epoch": 1.3388,
      "grad_norm": 6.7872772216796875,
      "learning_rate": 1.4626262626262627e-05,
      "loss": 0.3322,
      "step": 7104
    },
    {
      "epoch": 1.3389,
      "grad_norm": 5.568493843078613,
      "learning_rate": 1.4621212121212122e-05,
      "loss": 0.2193,
      "step": 7105
    },
    {
      "epoch": 1.339,
      "grad_norm": 11.099014282226562,
      "learning_rate": 1.4616161616161616e-05,
      "loss": 0.2998,
      "step": 7106
    },
    {
      "epoch": 1.3391,
      "grad_norm": 6.719240665435791,
      "learning_rate": 1.4611111111111112e-05,
      "loss": 0.1626,
      "step": 7107
    },
    {
      "epoch": 1.3392,
      "grad_norm": 5.809096336364746,
      "learning_rate": 1.4606060606060606e-05,
      "loss": 0.1829,
      "step": 7108
    },
    {
      "epoch": 1.3393,
      "grad_norm": 8.160956382751465,
      "learning_rate": 1.46010101010101e-05,
      "loss": 0.2157,
      "step": 7109
    },
    {
      "epoch": 1.3394,
      "grad_norm": 15.535470962524414,
      "learning_rate": 1.4595959595959597e-05,
      "loss": 0.1318,
      "step": 7110
    },
    {
      "epoch": 1.3395000000000001,
      "grad_norm": 8.005062103271484,
      "learning_rate": 1.4590909090909091e-05,
      "loss": 0.2538,
      "step": 7111
    },
    {
      "epoch": 1.3396,
      "grad_norm": 7.36880350112915,
      "learning_rate": 1.4585858585858587e-05,
      "loss": 0.2432,
      "step": 7112
    },
    {
      "epoch": 1.3397000000000001,
      "grad_norm": 9.055438995361328,
      "learning_rate": 1.4580808080808081e-05,
      "loss": 0.5221,
      "step": 7113
    },
    {
      "epoch": 1.3397999999999999,
      "grad_norm": 5.109979152679443,
      "learning_rate": 1.4575757575757575e-05,
      "loss": 0.3105,
      "step": 7114
    },
    {
      "epoch": 1.3399,
      "grad_norm": 10.717833518981934,
      "learning_rate": 1.4570707070707071e-05,
      "loss": 0.4465,
      "step": 7115
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.279204368591309,
      "learning_rate": 1.4565656565656566e-05,
      "loss": 0.5714,
      "step": 7116
    },
    {
      "epoch": 1.3401,
      "grad_norm": 6.98174524307251,
      "learning_rate": 1.456060606060606e-05,
      "loss": 0.1883,
      "step": 7117
    },
    {
      "epoch": 1.3402,
      "grad_norm": 12.684432983398438,
      "learning_rate": 1.4555555555555556e-05,
      "loss": 0.2871,
      "step": 7118
    },
    {
      "epoch": 1.3403,
      "grad_norm": 5.302495002746582,
      "learning_rate": 1.455050505050505e-05,
      "loss": 0.2229,
      "step": 7119
    },
    {
      "epoch": 1.3404,
      "grad_norm": 23.788475036621094,
      "learning_rate": 1.4545454545454545e-05,
      "loss": 0.5334,
      "step": 7120
    },
    {
      "epoch": 1.3405,
      "grad_norm": 12.949740409851074,
      "learning_rate": 1.454040404040404e-05,
      "loss": 0.4917,
      "step": 7121
    },
    {
      "epoch": 1.3406,
      "grad_norm": 5.4880266189575195,
      "learning_rate": 1.4535353535353535e-05,
      "loss": 0.1143,
      "step": 7122
    },
    {
      "epoch": 1.3407,
      "grad_norm": 6.739587306976318,
      "learning_rate": 1.4530303030303031e-05,
      "loss": 0.385,
      "step": 7123
    },
    {
      "epoch": 1.3408,
      "grad_norm": 5.790350914001465,
      "learning_rate": 1.4525252525252525e-05,
      "loss": 0.1596,
      "step": 7124
    },
    {
      "epoch": 1.3409,
      "grad_norm": 9.187888145446777,
      "learning_rate": 1.452020202020202e-05,
      "loss": 0.5921,
      "step": 7125
    },
    {
      "epoch": 1.341,
      "grad_norm": 6.92056131362915,
      "learning_rate": 1.4515151515151516e-05,
      "loss": 0.3331,
      "step": 7126
    },
    {
      "epoch": 1.3411,
      "grad_norm": 4.707560062408447,
      "learning_rate": 1.451010101010101e-05,
      "loss": 0.1824,
      "step": 7127
    },
    {
      "epoch": 1.3412,
      "grad_norm": 8.243234634399414,
      "learning_rate": 1.4505050505050504e-05,
      "loss": 0.1684,
      "step": 7128
    },
    {
      "epoch": 1.3413,
      "grad_norm": 25.616840362548828,
      "learning_rate": 1.45e-05,
      "loss": 0.5529,
      "step": 7129
    },
    {
      "epoch": 1.3414,
      "grad_norm": 5.42565393447876,
      "learning_rate": 1.4494949494949494e-05,
      "loss": 0.2011,
      "step": 7130
    },
    {
      "epoch": 1.3415,
      "grad_norm": 13.243587493896484,
      "learning_rate": 1.448989898989899e-05,
      "loss": 0.3173,
      "step": 7131
    },
    {
      "epoch": 1.3416000000000001,
      "grad_norm": 7.376512050628662,
      "learning_rate": 1.4484848484848485e-05,
      "loss": 0.2604,
      "step": 7132
    },
    {
      "epoch": 1.3417,
      "grad_norm": 14.533218383789062,
      "learning_rate": 1.4479797979797979e-05,
      "loss": 0.7703,
      "step": 7133
    },
    {
      "epoch": 1.3418,
      "grad_norm": 8.298497200012207,
      "learning_rate": 1.4474747474747475e-05,
      "loss": 0.2159,
      "step": 7134
    },
    {
      "epoch": 1.3418999999999999,
      "grad_norm": 6.52913761138916,
      "learning_rate": 1.446969696969697e-05,
      "loss": 0.2728,
      "step": 7135
    },
    {
      "epoch": 1.342,
      "grad_norm": 8.00589370727539,
      "learning_rate": 1.4464646464646464e-05,
      "loss": 0.3559,
      "step": 7136
    },
    {
      "epoch": 1.3421,
      "grad_norm": 10.639691352844238,
      "learning_rate": 1.445959595959596e-05,
      "loss": 0.3568,
      "step": 7137
    },
    {
      "epoch": 1.3422,
      "grad_norm": 7.290460109710693,
      "learning_rate": 1.4454545454545454e-05,
      "loss": 0.3678,
      "step": 7138
    },
    {
      "epoch": 1.3423,
      "grad_norm": 4.8558549880981445,
      "learning_rate": 1.4449494949494948e-05,
      "loss": 0.1647,
      "step": 7139
    },
    {
      "epoch": 1.3424,
      "grad_norm": 15.969025611877441,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 0.4274,
      "step": 7140
    },
    {
      "epoch": 1.3425,
      "grad_norm": 10.885160446166992,
      "learning_rate": 1.4439393939393942e-05,
      "loss": 0.7641,
      "step": 7141
    },
    {
      "epoch": 1.3426,
      "grad_norm": 10.30473518371582,
      "learning_rate": 1.4434343434343436e-05,
      "loss": 0.5403,
      "step": 7142
    },
    {
      "epoch": 1.3427,
      "grad_norm": 8.033065795898438,
      "learning_rate": 1.4429292929292932e-05,
      "loss": 0.4732,
      "step": 7143
    },
    {
      "epoch": 1.3428,
      "grad_norm": 7.39136266708374,
      "learning_rate": 1.4424242424242426e-05,
      "loss": 0.3306,
      "step": 7144
    },
    {
      "epoch": 1.3429,
      "grad_norm": 14.669373512268066,
      "learning_rate": 1.441919191919192e-05,
      "loss": 0.3746,
      "step": 7145
    },
    {
      "epoch": 1.343,
      "grad_norm": 4.920650482177734,
      "learning_rate": 1.4414141414141417e-05,
      "loss": 0.2058,
      "step": 7146
    },
    {
      "epoch": 1.3431,
      "grad_norm": 7.702996730804443,
      "learning_rate": 1.4409090909090911e-05,
      "loss": 0.2128,
      "step": 7147
    },
    {
      "epoch": 1.3432,
      "grad_norm": 9.105953216552734,
      "learning_rate": 1.4404040404040405e-05,
      "loss": 0.1959,
      "step": 7148
    },
    {
      "epoch": 1.3433,
      "grad_norm": 5.357209205627441,
      "learning_rate": 1.4398989898989901e-05,
      "loss": 0.177,
      "step": 7149
    },
    {
      "epoch": 1.3434,
      "grad_norm": 12.02001667022705,
      "learning_rate": 1.4393939393939396e-05,
      "loss": 0.5319,
      "step": 7150
    },
    {
      "epoch": 1.3435000000000001,
      "grad_norm": 11.293535232543945,
      "learning_rate": 1.438888888888889e-05,
      "loss": 0.5069,
      "step": 7151
    },
    {
      "epoch": 1.3436,
      "grad_norm": 12.646852493286133,
      "learning_rate": 1.4383838383838386e-05,
      "loss": 0.2519,
      "step": 7152
    },
    {
      "epoch": 1.3437000000000001,
      "grad_norm": 11.6351957321167,
      "learning_rate": 1.437878787878788e-05,
      "loss": 0.2613,
      "step": 7153
    },
    {
      "epoch": 1.3437999999999999,
      "grad_norm": 11.004632949829102,
      "learning_rate": 1.4373737373737376e-05,
      "loss": 0.5686,
      "step": 7154
    },
    {
      "epoch": 1.3439,
      "grad_norm": 11.560479164123535,
      "learning_rate": 1.436868686868687e-05,
      "loss": 0.8042,
      "step": 7155
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 6.08201265335083,
      "learning_rate": 1.4363636363636365e-05,
      "loss": 0.1894,
      "step": 7156
    },
    {
      "epoch": 1.3441,
      "grad_norm": 6.1250762939453125,
      "learning_rate": 1.435858585858586e-05,
      "loss": 0.2897,
      "step": 7157
    },
    {
      "epoch": 1.3442,
      "grad_norm": 5.787818431854248,
      "learning_rate": 1.4353535353535355e-05,
      "loss": 0.2912,
      "step": 7158
    },
    {
      "epoch": 1.3443,
      "grad_norm": 7.72598123550415,
      "learning_rate": 1.434848484848485e-05,
      "loss": 0.2515,
      "step": 7159
    },
    {
      "epoch": 1.3444,
      "grad_norm": 8.073678970336914,
      "learning_rate": 1.4343434343434345e-05,
      "loss": 0.2851,
      "step": 7160
    },
    {
      "epoch": 1.3445,
      "grad_norm": 23.742111206054688,
      "learning_rate": 1.433838383838384e-05,
      "loss": 0.2906,
      "step": 7161
    },
    {
      "epoch": 1.3446,
      "grad_norm": 6.619199752807617,
      "learning_rate": 1.4333333333333334e-05,
      "loss": 0.1712,
      "step": 7162
    },
    {
      "epoch": 1.3447,
      "grad_norm": 13.357319831848145,
      "learning_rate": 1.432828282828283e-05,
      "loss": 0.4321,
      "step": 7163
    },
    {
      "epoch": 1.3448,
      "grad_norm": 10.914114952087402,
      "learning_rate": 1.4323232323232324e-05,
      "loss": 0.2056,
      "step": 7164
    },
    {
      "epoch": 1.3449,
      "grad_norm": 79.9100570678711,
      "learning_rate": 1.431818181818182e-05,
      "loss": 0.4672,
      "step": 7165
    },
    {
      "epoch": 1.345,
      "grad_norm": 14.629908561706543,
      "learning_rate": 1.4313131313131314e-05,
      "loss": 0.5136,
      "step": 7166
    },
    {
      "epoch": 1.3451,
      "grad_norm": 11.963672637939453,
      "learning_rate": 1.4308080808080809e-05,
      "loss": 0.5031,
      "step": 7167
    },
    {
      "epoch": 1.3452,
      "grad_norm": 7.961976051330566,
      "learning_rate": 1.4303030303030305e-05,
      "loss": 0.2787,
      "step": 7168
    },
    {
      "epoch": 1.3453,
      "grad_norm": 9.10389232635498,
      "learning_rate": 1.4297979797979799e-05,
      "loss": 0.1969,
      "step": 7169
    },
    {
      "epoch": 1.3454,
      "grad_norm": 7.766971111297607,
      "learning_rate": 1.4292929292929293e-05,
      "loss": 0.2839,
      "step": 7170
    },
    {
      "epoch": 1.3455,
      "grad_norm": 8.12436580657959,
      "learning_rate": 1.428787878787879e-05,
      "loss": 0.4405,
      "step": 7171
    },
    {
      "epoch": 1.3456000000000001,
      "grad_norm": 18.326934814453125,
      "learning_rate": 1.4282828282828284e-05,
      "loss": 0.4003,
      "step": 7172
    },
    {
      "epoch": 1.3457,
      "grad_norm": 16.904144287109375,
      "learning_rate": 1.427777777777778e-05,
      "loss": 0.547,
      "step": 7173
    },
    {
      "epoch": 1.3458,
      "grad_norm": 14.950281143188477,
      "learning_rate": 1.4272727272727274e-05,
      "loss": 0.6,
      "step": 7174
    },
    {
      "epoch": 1.3458999999999999,
      "grad_norm": 5.822946548461914,
      "learning_rate": 1.4267676767676768e-05,
      "loss": 0.2495,
      "step": 7175
    },
    {
      "epoch": 1.346,
      "grad_norm": 16.53923988342285,
      "learning_rate": 1.4262626262626264e-05,
      "loss": 0.3592,
      "step": 7176
    },
    {
      "epoch": 1.3461,
      "grad_norm": 6.850736141204834,
      "learning_rate": 1.4257575757575758e-05,
      "loss": 0.3424,
      "step": 7177
    },
    {
      "epoch": 1.3462,
      "grad_norm": 4.764069557189941,
      "learning_rate": 1.4252525252525253e-05,
      "loss": 0.1548,
      "step": 7178
    },
    {
      "epoch": 1.3463,
      "grad_norm": 2.8875598907470703,
      "learning_rate": 1.4247474747474749e-05,
      "loss": 0.115,
      "step": 7179
    },
    {
      "epoch": 1.3464,
      "grad_norm": 20.699562072753906,
      "learning_rate": 1.4242424242424243e-05,
      "loss": 0.1269,
      "step": 7180
    },
    {
      "epoch": 1.3465,
      "grad_norm": 11.951380729675293,
      "learning_rate": 1.4237373737373737e-05,
      "loss": 0.1949,
      "step": 7181
    },
    {
      "epoch": 1.3466,
      "grad_norm": 4.371339797973633,
      "learning_rate": 1.4232323232323233e-05,
      "loss": 0.1538,
      "step": 7182
    },
    {
      "epoch": 1.3467,
      "grad_norm": 6.868022441864014,
      "learning_rate": 1.4227272727272728e-05,
      "loss": 0.4505,
      "step": 7183
    },
    {
      "epoch": 1.3468,
      "grad_norm": 8.634530067443848,
      "learning_rate": 1.4222222222222224e-05,
      "loss": 0.4556,
      "step": 7184
    },
    {
      "epoch": 1.3469,
      "grad_norm": 8.367551803588867,
      "learning_rate": 1.4217171717171718e-05,
      "loss": 0.2581,
      "step": 7185
    },
    {
      "epoch": 1.347,
      "grad_norm": 5.766191482543945,
      "learning_rate": 1.4212121212121212e-05,
      "loss": 0.1347,
      "step": 7186
    },
    {
      "epoch": 1.3471,
      "grad_norm": 5.294175148010254,
      "learning_rate": 1.4207070707070708e-05,
      "loss": 0.0778,
      "step": 7187
    },
    {
      "epoch": 1.3472,
      "grad_norm": 2.682513475418091,
      "learning_rate": 1.4202020202020203e-05,
      "loss": 0.1168,
      "step": 7188
    },
    {
      "epoch": 1.3473,
      "grad_norm": 7.148698329925537,
      "learning_rate": 1.4196969696969697e-05,
      "loss": 0.1913,
      "step": 7189
    },
    {
      "epoch": 1.3474,
      "grad_norm": 5.051438808441162,
      "learning_rate": 1.4191919191919193e-05,
      "loss": 0.2416,
      "step": 7190
    },
    {
      "epoch": 1.3475,
      "grad_norm": 9.03092098236084,
      "learning_rate": 1.4186868686868687e-05,
      "loss": 0.3248,
      "step": 7191
    },
    {
      "epoch": 1.3476,
      "grad_norm": 8.937836647033691,
      "learning_rate": 1.4181818181818181e-05,
      "loss": 0.3323,
      "step": 7192
    },
    {
      "epoch": 1.3477000000000001,
      "grad_norm": 3.2591934204101562,
      "learning_rate": 1.4176767676767677e-05,
      "loss": 0.1514,
      "step": 7193
    },
    {
      "epoch": 1.3477999999999999,
      "grad_norm": 5.656079292297363,
      "learning_rate": 1.4171717171717172e-05,
      "loss": 0.1876,
      "step": 7194
    },
    {
      "epoch": 1.3479,
      "grad_norm": 5.745997905731201,
      "learning_rate": 1.4166666666666668e-05,
      "loss": 0.3088,
      "step": 7195
    },
    {
      "epoch": 1.3479999999999999,
      "grad_norm": 10.830232620239258,
      "learning_rate": 1.4161616161616162e-05,
      "loss": 0.2874,
      "step": 7196
    },
    {
      "epoch": 1.3481,
      "grad_norm": 4.284721851348877,
      "learning_rate": 1.4156565656565656e-05,
      "loss": 0.1313,
      "step": 7197
    },
    {
      "epoch": 1.3482,
      "grad_norm": 3.9272758960723877,
      "learning_rate": 1.4151515151515152e-05,
      "loss": 0.3088,
      "step": 7198
    },
    {
      "epoch": 1.3483,
      "grad_norm": 22.376840591430664,
      "learning_rate": 1.4146464646464647e-05,
      "loss": 0.2573,
      "step": 7199
    },
    {
      "epoch": 1.3484,
      "grad_norm": 8.162372589111328,
      "learning_rate": 1.4141414141414141e-05,
      "loss": 0.4953,
      "step": 7200
    },
    {
      "epoch": 1.3485,
      "grad_norm": 6.622065544128418,
      "learning_rate": 1.4136363636363637e-05,
      "loss": 0.3006,
      "step": 7201
    },
    {
      "epoch": 1.3486,
      "grad_norm": 7.081238269805908,
      "learning_rate": 1.4131313131313131e-05,
      "loss": 0.265,
      "step": 7202
    },
    {
      "epoch": 1.3487,
      "grad_norm": 5.662996292114258,
      "learning_rate": 1.4126262626262627e-05,
      "loss": 0.2,
      "step": 7203
    },
    {
      "epoch": 1.3488,
      "grad_norm": 5.9389472007751465,
      "learning_rate": 1.4121212121212121e-05,
      "loss": 0.2026,
      "step": 7204
    },
    {
      "epoch": 1.3489,
      "grad_norm": 7.125166416168213,
      "learning_rate": 1.4116161616161616e-05,
      "loss": 0.4332,
      "step": 7205
    },
    {
      "epoch": 1.349,
      "grad_norm": 9.512670516967773,
      "learning_rate": 1.4111111111111112e-05,
      "loss": 0.2287,
      "step": 7206
    },
    {
      "epoch": 1.3491,
      "grad_norm": 19.92790412902832,
      "learning_rate": 1.4106060606060606e-05,
      "loss": 0.1694,
      "step": 7207
    },
    {
      "epoch": 1.3492,
      "grad_norm": 12.899961471557617,
      "learning_rate": 1.41010101010101e-05,
      "loss": 0.7995,
      "step": 7208
    },
    {
      "epoch": 1.3493,
      "grad_norm": 4.736947059631348,
      "learning_rate": 1.4095959595959596e-05,
      "loss": 0.2078,
      "step": 7209
    },
    {
      "epoch": 1.3494,
      "grad_norm": 10.14119815826416,
      "learning_rate": 1.409090909090909e-05,
      "loss": 0.13,
      "step": 7210
    },
    {
      "epoch": 1.3495,
      "grad_norm": 6.547591209411621,
      "learning_rate": 1.4085858585858585e-05,
      "loss": 0.204,
      "step": 7211
    },
    {
      "epoch": 1.3496000000000001,
      "grad_norm": 10.165964126586914,
      "learning_rate": 1.4080808080808081e-05,
      "loss": 0.3044,
      "step": 7212
    },
    {
      "epoch": 1.3497,
      "grad_norm": 16.701557159423828,
      "learning_rate": 1.4075757575757575e-05,
      "loss": 0.2643,
      "step": 7213
    },
    {
      "epoch": 1.3498,
      "grad_norm": 12.476609230041504,
      "learning_rate": 1.4070707070707071e-05,
      "loss": 0.5058,
      "step": 7214
    },
    {
      "epoch": 1.3498999999999999,
      "grad_norm": 6.481112003326416,
      "learning_rate": 1.4065656565656565e-05,
      "loss": 0.2349,
      "step": 7215
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.585615158081055,
      "learning_rate": 1.406060606060606e-05,
      "loss": 0.2028,
      "step": 7216
    },
    {
      "epoch": 1.3501,
      "grad_norm": 8.084648132324219,
      "learning_rate": 1.4055555555555556e-05,
      "loss": 0.2685,
      "step": 7217
    },
    {
      "epoch": 1.3502,
      "grad_norm": 4.955648422241211,
      "learning_rate": 1.405050505050505e-05,
      "loss": 0.1541,
      "step": 7218
    },
    {
      "epoch": 1.3503,
      "grad_norm": 9.82510757446289,
      "learning_rate": 1.4045454545454544e-05,
      "loss": 0.4546,
      "step": 7219
    },
    {
      "epoch": 1.3504,
      "grad_norm": 5.223247528076172,
      "learning_rate": 1.404040404040404e-05,
      "loss": 0.2229,
      "step": 7220
    },
    {
      "epoch": 1.3505,
      "grad_norm": 6.778182506561279,
      "learning_rate": 1.4035353535353535e-05,
      "loss": 0.2414,
      "step": 7221
    },
    {
      "epoch": 1.3506,
      "grad_norm": 12.179394721984863,
      "learning_rate": 1.4030303030303029e-05,
      "loss": 0.2432,
      "step": 7222
    },
    {
      "epoch": 1.3507,
      "grad_norm": 6.925106525421143,
      "learning_rate": 1.4025252525252525e-05,
      "loss": 0.2298,
      "step": 7223
    },
    {
      "epoch": 1.3508,
      "grad_norm": 5.919551849365234,
      "learning_rate": 1.402020202020202e-05,
      "loss": 0.2579,
      "step": 7224
    },
    {
      "epoch": 1.3509,
      "grad_norm": 7.238466739654541,
      "learning_rate": 1.4015151515151515e-05,
      "loss": 0.1807,
      "step": 7225
    },
    {
      "epoch": 1.351,
      "grad_norm": 2.9524598121643066,
      "learning_rate": 1.401010101010101e-05,
      "loss": 0.0926,
      "step": 7226
    },
    {
      "epoch": 1.3511,
      "grad_norm": 9.154086112976074,
      "learning_rate": 1.4005050505050507e-05,
      "loss": 0.1288,
      "step": 7227
    },
    {
      "epoch": 1.3512,
      "grad_norm": 7.588508605957031,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.2128,
      "step": 7228
    },
    {
      "epoch": 1.3513,
      "grad_norm": 8.186918258666992,
      "learning_rate": 1.3994949494949497e-05,
      "loss": 0.3627,
      "step": 7229
    },
    {
      "epoch": 1.3514,
      "grad_norm": 8.227171897888184,
      "learning_rate": 1.3989898989898992e-05,
      "loss": 0.3399,
      "step": 7230
    },
    {
      "epoch": 1.3515,
      "grad_norm": 13.229992866516113,
      "learning_rate": 1.3984848484848486e-05,
      "loss": 0.7395,
      "step": 7231
    },
    {
      "epoch": 1.3516,
      "grad_norm": 8.180464744567871,
      "learning_rate": 1.3979797979797982e-05,
      "loss": 0.1777,
      "step": 7232
    },
    {
      "epoch": 1.3517000000000001,
      "grad_norm": 21.736595153808594,
      "learning_rate": 1.3974747474747476e-05,
      "loss": 0.3285,
      "step": 7233
    },
    {
      "epoch": 1.3518,
      "grad_norm": 3.3786110877990723,
      "learning_rate": 1.3969696969696972e-05,
      "loss": 0.1087,
      "step": 7234
    },
    {
      "epoch": 1.3519,
      "grad_norm": 14.25472354888916,
      "learning_rate": 1.3964646464646467e-05,
      "loss": 0.2525,
      "step": 7235
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 6.136772155761719,
      "learning_rate": 1.3959595959595961e-05,
      "loss": 0.2116,
      "step": 7236
    },
    {
      "epoch": 1.3521,
      "grad_norm": 4.059474945068359,
      "learning_rate": 1.3954545454545457e-05,
      "loss": 0.2312,
      "step": 7237
    },
    {
      "epoch": 1.3522,
      "grad_norm": 106.70281982421875,
      "learning_rate": 1.3949494949494951e-05,
      "loss": 0.2451,
      "step": 7238
    },
    {
      "epoch": 1.3523,
      "grad_norm": 5.43370246887207,
      "learning_rate": 1.3944444444444446e-05,
      "loss": 0.2928,
      "step": 7239
    },
    {
      "epoch": 1.3524,
      "grad_norm": 11.253753662109375,
      "learning_rate": 1.3939393939393942e-05,
      "loss": 0.586,
      "step": 7240
    },
    {
      "epoch": 1.3525,
      "grad_norm": 4.611744403839111,
      "learning_rate": 1.3934343434343436e-05,
      "loss": 0.1615,
      "step": 7241
    },
    {
      "epoch": 1.3526,
      "grad_norm": 7.507802486419678,
      "learning_rate": 1.392929292929293e-05,
      "loss": 0.2582,
      "step": 7242
    },
    {
      "epoch": 1.3527,
      "grad_norm": 4.622288703918457,
      "learning_rate": 1.3924242424242426e-05,
      "loss": 0.1736,
      "step": 7243
    },
    {
      "epoch": 1.3528,
      "grad_norm": 8.369211196899414,
      "learning_rate": 1.391919191919192e-05,
      "loss": 0.2735,
      "step": 7244
    },
    {
      "epoch": 1.3529,
      "grad_norm": 10.077502250671387,
      "learning_rate": 1.3914141414141416e-05,
      "loss": 0.234,
      "step": 7245
    },
    {
      "epoch": 1.353,
      "grad_norm": 9.531304359436035,
      "learning_rate": 1.390909090909091e-05,
      "loss": 0.713,
      "step": 7246
    },
    {
      "epoch": 1.3531,
      "grad_norm": 7.3392510414123535,
      "learning_rate": 1.3904040404040405e-05,
      "loss": 0.585,
      "step": 7247
    },
    {
      "epoch": 1.3532,
      "grad_norm": 4.762653827667236,
      "learning_rate": 1.3898989898989901e-05,
      "loss": 0.2014,
      "step": 7248
    },
    {
      "epoch": 1.3533,
      "grad_norm": 10.87932300567627,
      "learning_rate": 1.3893939393939395e-05,
      "loss": 0.1704,
      "step": 7249
    },
    {
      "epoch": 1.3534,
      "grad_norm": 9.845534324645996,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.3073,
      "step": 7250
    },
    {
      "epoch": 1.3535,
      "grad_norm": 11.106229782104492,
      "learning_rate": 1.3883838383838386e-05,
      "loss": 0.2795,
      "step": 7251
    },
    {
      "epoch": 1.3536000000000001,
      "grad_norm": 7.463860511779785,
      "learning_rate": 1.387878787878788e-05,
      "loss": 0.3248,
      "step": 7252
    },
    {
      "epoch": 1.3537,
      "grad_norm": 16.049333572387695,
      "learning_rate": 1.3873737373737374e-05,
      "loss": 0.3829,
      "step": 7253
    },
    {
      "epoch": 1.3538000000000001,
      "grad_norm": 6.862534523010254,
      "learning_rate": 1.386868686868687e-05,
      "loss": 0.2563,
      "step": 7254
    },
    {
      "epoch": 1.3538999999999999,
      "grad_norm": 9.799659729003906,
      "learning_rate": 1.3863636363636364e-05,
      "loss": 0.3196,
      "step": 7255
    },
    {
      "epoch": 1.354,
      "grad_norm": 16.081401824951172,
      "learning_rate": 1.385858585858586e-05,
      "loss": 0.2845,
      "step": 7256
    },
    {
      "epoch": 1.3541,
      "grad_norm": 8.3316650390625,
      "learning_rate": 1.3853535353535355e-05,
      "loss": 0.5107,
      "step": 7257
    },
    {
      "epoch": 1.3542,
      "grad_norm": 4.901011943817139,
      "learning_rate": 1.3848484848484849e-05,
      "loss": 0.268,
      "step": 7258
    },
    {
      "epoch": 1.3543,
      "grad_norm": 4.1084418296813965,
      "learning_rate": 1.3843434343434345e-05,
      "loss": 0.1427,
      "step": 7259
    },
    {
      "epoch": 1.3544,
      "grad_norm": 7.609618663787842,
      "learning_rate": 1.383838383838384e-05,
      "loss": 0.1893,
      "step": 7260
    },
    {
      "epoch": 1.3545,
      "grad_norm": 5.388002872467041,
      "learning_rate": 1.3833333333333334e-05,
      "loss": 0.2194,
      "step": 7261
    },
    {
      "epoch": 1.3546,
      "grad_norm": 12.124622344970703,
      "learning_rate": 1.382828282828283e-05,
      "loss": 0.2581,
      "step": 7262
    },
    {
      "epoch": 1.3547,
      "grad_norm": 20.580297470092773,
      "learning_rate": 1.3823232323232324e-05,
      "loss": 0.4134,
      "step": 7263
    },
    {
      "epoch": 1.3548,
      "grad_norm": 7.911673069000244,
      "learning_rate": 1.3818181818181818e-05,
      "loss": 0.1643,
      "step": 7264
    },
    {
      "epoch": 1.3549,
      "grad_norm": 7.204929828643799,
      "learning_rate": 1.3813131313131314e-05,
      "loss": 0.1721,
      "step": 7265
    },
    {
      "epoch": 1.355,
      "grad_norm": 5.101128578186035,
      "learning_rate": 1.3808080808080808e-05,
      "loss": 0.2435,
      "step": 7266
    },
    {
      "epoch": 1.3551,
      "grad_norm": 9.198049545288086,
      "learning_rate": 1.3803030303030304e-05,
      "loss": 0.3454,
      "step": 7267
    },
    {
      "epoch": 1.3552,
      "grad_norm": 6.586385250091553,
      "learning_rate": 1.3797979797979799e-05,
      "loss": 0.1401,
      "step": 7268
    },
    {
      "epoch": 1.3553,
      "grad_norm": 5.056763648986816,
      "learning_rate": 1.3792929292929293e-05,
      "loss": 0.2754,
      "step": 7269
    },
    {
      "epoch": 1.3554,
      "grad_norm": 5.300357818603516,
      "learning_rate": 1.3787878787878789e-05,
      "loss": 0.2046,
      "step": 7270
    },
    {
      "epoch": 1.3555,
      "grad_norm": 18.572803497314453,
      "learning_rate": 1.3782828282828283e-05,
      "loss": 0.3579,
      "step": 7271
    },
    {
      "epoch": 1.3556,
      "grad_norm": 6.9961628913879395,
      "learning_rate": 1.3777777777777778e-05,
      "loss": 0.2586,
      "step": 7272
    },
    {
      "epoch": 1.3557000000000001,
      "grad_norm": 6.507721900939941,
      "learning_rate": 1.3772727272727274e-05,
      "loss": 0.2623,
      "step": 7273
    },
    {
      "epoch": 1.3558,
      "grad_norm": 9.107076644897461,
      "learning_rate": 1.3767676767676768e-05,
      "loss": 0.2171,
      "step": 7274
    },
    {
      "epoch": 1.3559,
      "grad_norm": 7.0891594886779785,
      "learning_rate": 1.3762626262626264e-05,
      "loss": 0.3255,
      "step": 7275
    },
    {
      "epoch": 1.3559999999999999,
      "grad_norm": 6.087429523468018,
      "learning_rate": 1.3757575757575758e-05,
      "loss": 0.1544,
      "step": 7276
    },
    {
      "epoch": 1.3561,
      "grad_norm": 7.387682914733887,
      "learning_rate": 1.3752525252525252e-05,
      "loss": 0.3354,
      "step": 7277
    },
    {
      "epoch": 1.3562,
      "grad_norm": 10.20298957824707,
      "learning_rate": 1.3747474747474748e-05,
      "loss": 0.3869,
      "step": 7278
    },
    {
      "epoch": 1.3563,
      "grad_norm": 19.895811080932617,
      "learning_rate": 1.3742424242424243e-05,
      "loss": 0.2739,
      "step": 7279
    },
    {
      "epoch": 1.3564,
      "grad_norm": 3.096534252166748,
      "learning_rate": 1.3737373737373737e-05,
      "loss": 0.1009,
      "step": 7280
    },
    {
      "epoch": 1.3565,
      "grad_norm": 4.792577266693115,
      "learning_rate": 1.3732323232323233e-05,
      "loss": 0.1481,
      "step": 7281
    },
    {
      "epoch": 1.3566,
      "grad_norm": 37.42909622192383,
      "learning_rate": 1.3727272727272727e-05,
      "loss": 0.3487,
      "step": 7282
    },
    {
      "epoch": 1.3567,
      "grad_norm": 7.92329740524292,
      "learning_rate": 1.3722222222222222e-05,
      "loss": 0.3257,
      "step": 7283
    },
    {
      "epoch": 1.3568,
      "grad_norm": 8.492177963256836,
      "learning_rate": 1.3717171717171718e-05,
      "loss": 0.2709,
      "step": 7284
    },
    {
      "epoch": 1.3569,
      "grad_norm": 42.99041748046875,
      "learning_rate": 1.3712121212121212e-05,
      "loss": 0.3099,
      "step": 7285
    },
    {
      "epoch": 1.357,
      "grad_norm": 8.197415351867676,
      "learning_rate": 1.3707070707070708e-05,
      "loss": 0.4083,
      "step": 7286
    },
    {
      "epoch": 1.3571,
      "grad_norm": 7.708413124084473,
      "learning_rate": 1.3702020202020202e-05,
      "loss": 0.5169,
      "step": 7287
    },
    {
      "epoch": 1.3572,
      "grad_norm": 11.943949699401855,
      "learning_rate": 1.3696969696969697e-05,
      "loss": 0.4361,
      "step": 7288
    },
    {
      "epoch": 1.3573,
      "grad_norm": 12.299735069274902,
      "learning_rate": 1.3691919191919193e-05,
      "loss": 0.2319,
      "step": 7289
    },
    {
      "epoch": 1.3574,
      "grad_norm": 4.608478546142578,
      "learning_rate": 1.3686868686868687e-05,
      "loss": 0.131,
      "step": 7290
    },
    {
      "epoch": 1.3575,
      "grad_norm": 6.2787089347839355,
      "learning_rate": 1.3681818181818181e-05,
      "loss": 0.2095,
      "step": 7291
    },
    {
      "epoch": 1.3576,
      "grad_norm": 4.372931480407715,
      "learning_rate": 1.3676767676767677e-05,
      "loss": 0.1972,
      "step": 7292
    },
    {
      "epoch": 1.3577,
      "grad_norm": 3.741279363632202,
      "learning_rate": 1.3671717171717171e-05,
      "loss": 0.1537,
      "step": 7293
    },
    {
      "epoch": 1.3578000000000001,
      "grad_norm": 4.2275309562683105,
      "learning_rate": 1.3666666666666666e-05,
      "loss": 0.1385,
      "step": 7294
    },
    {
      "epoch": 1.3578999999999999,
      "grad_norm": 5.046774864196777,
      "learning_rate": 1.3661616161616162e-05,
      "loss": 0.1738,
      "step": 7295
    },
    {
      "epoch": 1.358,
      "grad_norm": 17.04853630065918,
      "learning_rate": 1.3656565656565656e-05,
      "loss": 0.2132,
      "step": 7296
    },
    {
      "epoch": 1.3580999999999999,
      "grad_norm": 4.3664469718933105,
      "learning_rate": 1.3651515151515152e-05,
      "loss": 0.1721,
      "step": 7297
    },
    {
      "epoch": 1.3582,
      "grad_norm": 6.283840656280518,
      "learning_rate": 1.3646464646464646e-05,
      "loss": 0.0992,
      "step": 7298
    },
    {
      "epoch": 1.3583,
      "grad_norm": 8.413589477539062,
      "learning_rate": 1.364141414141414e-05,
      "loss": 0.1606,
      "step": 7299
    },
    {
      "epoch": 1.3584,
      "grad_norm": 6.849740505218506,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.1331,
      "step": 7300
    },
    {
      "epoch": 1.3585,
      "grad_norm": 5.500327110290527,
      "learning_rate": 1.3631313131313131e-05,
      "loss": 0.1525,
      "step": 7301
    },
    {
      "epoch": 1.3586,
      "grad_norm": 6.621022701263428,
      "learning_rate": 1.3626262626262625e-05,
      "loss": 0.1261,
      "step": 7302
    },
    {
      "epoch": 1.3587,
      "grad_norm": 6.053813457489014,
      "learning_rate": 1.3621212121212121e-05,
      "loss": 0.1184,
      "step": 7303
    },
    {
      "epoch": 1.3588,
      "grad_norm": 5.133228302001953,
      "learning_rate": 1.3616161616161615e-05,
      "loss": 0.2017,
      "step": 7304
    },
    {
      "epoch": 1.3589,
      "grad_norm": 6.342249870300293,
      "learning_rate": 1.3611111111111111e-05,
      "loss": 0.2286,
      "step": 7305
    },
    {
      "epoch": 1.359,
      "grad_norm": 5.388069152832031,
      "learning_rate": 1.3606060606060606e-05,
      "loss": 0.1708,
      "step": 7306
    },
    {
      "epoch": 1.3591,
      "grad_norm": 15.207536697387695,
      "learning_rate": 1.36010101010101e-05,
      "loss": 0.1713,
      "step": 7307
    },
    {
      "epoch": 1.3592,
      "grad_norm": 4.806139945983887,
      "learning_rate": 1.3595959595959596e-05,
      "loss": 0.1489,
      "step": 7308
    },
    {
      "epoch": 1.3593,
      "grad_norm": 9.78294849395752,
      "learning_rate": 1.359090909090909e-05,
      "loss": 0.3124,
      "step": 7309
    },
    {
      "epoch": 1.3594,
      "grad_norm": 12.378129959106445,
      "learning_rate": 1.3585858585858585e-05,
      "loss": 0.2507,
      "step": 7310
    },
    {
      "epoch": 1.3595,
      "grad_norm": 9.079397201538086,
      "learning_rate": 1.358080808080808e-05,
      "loss": 0.2246,
      "step": 7311
    },
    {
      "epoch": 1.3596,
      "grad_norm": 7.09236478805542,
      "learning_rate": 1.3575757575757578e-05,
      "loss": 0.2615,
      "step": 7312
    },
    {
      "epoch": 1.3597000000000001,
      "grad_norm": 5.660340785980225,
      "learning_rate": 1.3570707070707073e-05,
      "loss": 0.3088,
      "step": 7313
    },
    {
      "epoch": 1.3598,
      "grad_norm": 4.709356784820557,
      "learning_rate": 1.3565656565656567e-05,
      "loss": 0.2469,
      "step": 7314
    },
    {
      "epoch": 1.3599,
      "grad_norm": 5.692628860473633,
      "learning_rate": 1.3560606060606063e-05,
      "loss": 0.2459,
      "step": 7315
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 8.896580696105957,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 0.2255,
      "step": 7316
    },
    {
      "epoch": 1.3601,
      "grad_norm": 6.949743747711182,
      "learning_rate": 1.3550505050505053e-05,
      "loss": 0.2616,
      "step": 7317
    },
    {
      "epoch": 1.3602,
      "grad_norm": 8.303091049194336,
      "learning_rate": 1.3545454545454547e-05,
      "loss": 0.3423,
      "step": 7318
    },
    {
      "epoch": 1.3603,
      "grad_norm": 4.757632255554199,
      "learning_rate": 1.3540404040404042e-05,
      "loss": 0.267,
      "step": 7319
    },
    {
      "epoch": 1.3604,
      "grad_norm": 26.43181800842285,
      "learning_rate": 1.3535353535353538e-05,
      "loss": 0.1303,
      "step": 7320
    },
    {
      "epoch": 1.3605,
      "grad_norm": 8.211459159851074,
      "learning_rate": 1.3530303030303032e-05,
      "loss": 0.3781,
      "step": 7321
    },
    {
      "epoch": 1.3606,
      "grad_norm": 4.71828031539917,
      "learning_rate": 1.3525252525252526e-05,
      "loss": 0.2439,
      "step": 7322
    },
    {
      "epoch": 1.3607,
      "grad_norm": 13.23886775970459,
      "learning_rate": 1.3520202020202022e-05,
      "loss": 0.2067,
      "step": 7323
    },
    {
      "epoch": 1.3608,
      "grad_norm": 12.199446678161621,
      "learning_rate": 1.3515151515151517e-05,
      "loss": 0.1315,
      "step": 7324
    },
    {
      "epoch": 1.3609,
      "grad_norm": 6.63808536529541,
      "learning_rate": 1.3510101010101011e-05,
      "loss": 0.2005,
      "step": 7325
    },
    {
      "epoch": 1.361,
      "grad_norm": 8.08220100402832,
      "learning_rate": 1.3505050505050507e-05,
      "loss": 0.1718,
      "step": 7326
    },
    {
      "epoch": 1.3611,
      "grad_norm": 2.835512638092041,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.0727,
      "step": 7327
    },
    {
      "epoch": 1.3612,
      "grad_norm": 4.611718654632568,
      "learning_rate": 1.3494949494949497e-05,
      "loss": 0.2235,
      "step": 7328
    },
    {
      "epoch": 1.3613,
      "grad_norm": 4.212069034576416,
      "learning_rate": 1.3489898989898991e-05,
      "loss": 0.1621,
      "step": 7329
    },
    {
      "epoch": 1.3614,
      "grad_norm": 5.538906097412109,
      "learning_rate": 1.3484848484848486e-05,
      "loss": 0.1723,
      "step": 7330
    },
    {
      "epoch": 1.3615,
      "grad_norm": 6.138006687164307,
      "learning_rate": 1.3479797979797982e-05,
      "loss": 0.2502,
      "step": 7331
    },
    {
      "epoch": 1.3616,
      "grad_norm": 5.361143589019775,
      "learning_rate": 1.3474747474747476e-05,
      "loss": 0.1889,
      "step": 7332
    },
    {
      "epoch": 1.3617,
      "grad_norm": 17.877456665039062,
      "learning_rate": 1.346969696969697e-05,
      "loss": 0.2408,
      "step": 7333
    },
    {
      "epoch": 1.3618000000000001,
      "grad_norm": 6.850100040435791,
      "learning_rate": 1.3464646464646466e-05,
      "loss": 0.2366,
      "step": 7334
    },
    {
      "epoch": 1.3618999999999999,
      "grad_norm": 10.904051780700684,
      "learning_rate": 1.345959595959596e-05,
      "loss": 0.258,
      "step": 7335
    },
    {
      "epoch": 1.362,
      "grad_norm": 5.776535987854004,
      "learning_rate": 1.3454545454545457e-05,
      "loss": 0.2071,
      "step": 7336
    },
    {
      "epoch": 1.3620999999999999,
      "grad_norm": 7.439198017120361,
      "learning_rate": 1.3449494949494951e-05,
      "loss": 0.2011,
      "step": 7337
    },
    {
      "epoch": 1.3622,
      "grad_norm": 7.317779064178467,
      "learning_rate": 1.3444444444444445e-05,
      "loss": 0.2688,
      "step": 7338
    },
    {
      "epoch": 1.3623,
      "grad_norm": 6.90136194229126,
      "learning_rate": 1.3439393939393941e-05,
      "loss": 0.1631,
      "step": 7339
    },
    {
      "epoch": 1.3624,
      "grad_norm": 5.593254089355469,
      "learning_rate": 1.3434343434343436e-05,
      "loss": 0.2347,
      "step": 7340
    },
    {
      "epoch": 1.3625,
      "grad_norm": 4.826152324676514,
      "learning_rate": 1.342929292929293e-05,
      "loss": 0.199,
      "step": 7341
    },
    {
      "epoch": 1.3626,
      "grad_norm": 6.488945007324219,
      "learning_rate": 1.3424242424242426e-05,
      "loss": 0.1934,
      "step": 7342
    },
    {
      "epoch": 1.3627,
      "grad_norm": 10.668092727661133,
      "learning_rate": 1.341919191919192e-05,
      "loss": 0.2168,
      "step": 7343
    },
    {
      "epoch": 1.3628,
      "grad_norm": 6.237236976623535,
      "learning_rate": 1.3414141414141414e-05,
      "loss": 0.2587,
      "step": 7344
    },
    {
      "epoch": 1.3629,
      "grad_norm": 9.008034706115723,
      "learning_rate": 1.340909090909091e-05,
      "loss": 0.1586,
      "step": 7345
    },
    {
      "epoch": 1.363,
      "grad_norm": 4.5607733726501465,
      "learning_rate": 1.3404040404040405e-05,
      "loss": 0.1582,
      "step": 7346
    },
    {
      "epoch": 1.3631,
      "grad_norm": 21.89518928527832,
      "learning_rate": 1.33989898989899e-05,
      "loss": 0.1732,
      "step": 7347
    },
    {
      "epoch": 1.3632,
      "grad_norm": 11.201786041259766,
      "learning_rate": 1.3393939393939395e-05,
      "loss": 0.3085,
      "step": 7348
    },
    {
      "epoch": 1.3633,
      "grad_norm": 14.54820728302002,
      "learning_rate": 1.338888888888889e-05,
      "loss": 0.2082,
      "step": 7349
    },
    {
      "epoch": 1.3634,
      "grad_norm": 47.42871856689453,
      "learning_rate": 1.3383838383838385e-05,
      "loss": 0.3006,
      "step": 7350
    },
    {
      "epoch": 1.3635,
      "grad_norm": 7.025759696960449,
      "learning_rate": 1.337878787878788e-05,
      "loss": 0.4042,
      "step": 7351
    },
    {
      "epoch": 1.3636,
      "grad_norm": 4.5487775802612305,
      "learning_rate": 1.3373737373737374e-05,
      "loss": 0.174,
      "step": 7352
    },
    {
      "epoch": 1.3637000000000001,
      "grad_norm": 32.15953063964844,
      "learning_rate": 1.336868686868687e-05,
      "loss": 0.2133,
      "step": 7353
    },
    {
      "epoch": 1.3638,
      "grad_norm": 21.139450073242188,
      "learning_rate": 1.3363636363636364e-05,
      "loss": 0.6728,
      "step": 7354
    },
    {
      "epoch": 1.3639000000000001,
      "grad_norm": 7.328402042388916,
      "learning_rate": 1.3358585858585858e-05,
      "loss": 0.459,
      "step": 7355
    },
    {
      "epoch": 1.3639999999999999,
      "grad_norm": 8.201581954956055,
      "learning_rate": 1.3353535353535354e-05,
      "loss": 0.3845,
      "step": 7356
    },
    {
      "epoch": 1.3641,
      "grad_norm": 18.9661865234375,
      "learning_rate": 1.3348484848484849e-05,
      "loss": 0.3857,
      "step": 7357
    },
    {
      "epoch": 1.3642,
      "grad_norm": 13.297045707702637,
      "learning_rate": 1.3343434343434345e-05,
      "loss": 0.2522,
      "step": 7358
    },
    {
      "epoch": 1.3643,
      "grad_norm": 8.221348762512207,
      "learning_rate": 1.3338383838383839e-05,
      "loss": 0.2622,
      "step": 7359
    },
    {
      "epoch": 1.3644,
      "grad_norm": 4.324363708496094,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.1969,
      "step": 7360
    },
    {
      "epoch": 1.3645,
      "grad_norm": 28.523195266723633,
      "learning_rate": 1.332828282828283e-05,
      "loss": 0.2909,
      "step": 7361
    },
    {
      "epoch": 1.3646,
      "grad_norm": 6.753603935241699,
      "learning_rate": 1.3323232323232324e-05,
      "loss": 0.2624,
      "step": 7362
    },
    {
      "epoch": 1.3647,
      "grad_norm": 127.6070556640625,
      "learning_rate": 1.3318181818181818e-05,
      "loss": 0.1828,
      "step": 7363
    },
    {
      "epoch": 1.3648,
      "grad_norm": 9.916587829589844,
      "learning_rate": 1.3313131313131314e-05,
      "loss": 0.2336,
      "step": 7364
    },
    {
      "epoch": 1.3649,
      "grad_norm": 8.824202537536621,
      "learning_rate": 1.3308080808080808e-05,
      "loss": 0.2594,
      "step": 7365
    },
    {
      "epoch": 1.365,
      "grad_norm": 4.732639789581299,
      "learning_rate": 1.3303030303030302e-05,
      "loss": 0.252,
      "step": 7366
    },
    {
      "epoch": 1.3651,
      "grad_norm": 6.09342098236084,
      "learning_rate": 1.3297979797979798e-05,
      "loss": 0.2452,
      "step": 7367
    },
    {
      "epoch": 1.3652,
      "grad_norm": 9.09153938293457,
      "learning_rate": 1.3292929292929293e-05,
      "loss": 0.2801,
      "step": 7368
    },
    {
      "epoch": 1.3653,
      "grad_norm": 11.385300636291504,
      "learning_rate": 1.3287878787878789e-05,
      "loss": 0.3045,
      "step": 7369
    },
    {
      "epoch": 1.3654,
      "grad_norm": 41.575523376464844,
      "learning_rate": 1.3282828282828283e-05,
      "loss": 0.1688,
      "step": 7370
    },
    {
      "epoch": 1.3655,
      "grad_norm": 210.78012084960938,
      "learning_rate": 1.3277777777777777e-05,
      "loss": 0.2983,
      "step": 7371
    },
    {
      "epoch": 1.3656,
      "grad_norm": 5.658142566680908,
      "learning_rate": 1.3272727272727273e-05,
      "loss": 0.284,
      "step": 7372
    },
    {
      "epoch": 1.3657,
      "grad_norm": 12.56075668334961,
      "learning_rate": 1.3267676767676768e-05,
      "loss": 0.2687,
      "step": 7373
    },
    {
      "epoch": 1.3658000000000001,
      "grad_norm": 11.60124683380127,
      "learning_rate": 1.3262626262626262e-05,
      "loss": 0.3347,
      "step": 7374
    },
    {
      "epoch": 1.3659,
      "grad_norm": 13.527230262756348,
      "learning_rate": 1.3257575757575758e-05,
      "loss": 0.3676,
      "step": 7375
    },
    {
      "epoch": 1.366,
      "grad_norm": 6.561531066894531,
      "learning_rate": 1.3252525252525252e-05,
      "loss": 0.1947,
      "step": 7376
    },
    {
      "epoch": 1.3660999999999999,
      "grad_norm": 12.043034553527832,
      "learning_rate": 1.3247474747474748e-05,
      "loss": 0.3795,
      "step": 7377
    },
    {
      "epoch": 1.3662,
      "grad_norm": 5.997438907623291,
      "learning_rate": 1.3242424242424242e-05,
      "loss": 0.2372,
      "step": 7378
    },
    {
      "epoch": 1.3663,
      "grad_norm": 46.09365463256836,
      "learning_rate": 1.3237373737373737e-05,
      "loss": 0.3561,
      "step": 7379
    },
    {
      "epoch": 1.3664,
      "grad_norm": 9.295849800109863,
      "learning_rate": 1.3232323232323233e-05,
      "loss": 0.2607,
      "step": 7380
    },
    {
      "epoch": 1.3665,
      "grad_norm": 8.306079864501953,
      "learning_rate": 1.3227272727272727e-05,
      "loss": 0.2328,
      "step": 7381
    },
    {
      "epoch": 1.3666,
      "grad_norm": 9.36131763458252,
      "learning_rate": 1.3222222222222221e-05,
      "loss": 0.4191,
      "step": 7382
    },
    {
      "epoch": 1.3667,
      "grad_norm": 4.868494510650635,
      "learning_rate": 1.3217171717171717e-05,
      "loss": 0.3432,
      "step": 7383
    },
    {
      "epoch": 1.3668,
      "grad_norm": 10.880635261535645,
      "learning_rate": 1.3212121212121212e-05,
      "loss": 0.3302,
      "step": 7384
    },
    {
      "epoch": 1.3669,
      "grad_norm": 7.310423374176025,
      "learning_rate": 1.3207070707070706e-05,
      "loss": 0.4118,
      "step": 7385
    },
    {
      "epoch": 1.367,
      "grad_norm": 7.215883731842041,
      "learning_rate": 1.3202020202020202e-05,
      "loss": 0.3473,
      "step": 7386
    },
    {
      "epoch": 1.3671,
      "grad_norm": 7.270547389984131,
      "learning_rate": 1.3196969696969696e-05,
      "loss": 0.337,
      "step": 7387
    },
    {
      "epoch": 1.3672,
      "grad_norm": 7.890033721923828,
      "learning_rate": 1.3191919191919192e-05,
      "loss": 0.3683,
      "step": 7388
    },
    {
      "epoch": 1.3673,
      "grad_norm": 10.466940879821777,
      "learning_rate": 1.3186868686868687e-05,
      "loss": 0.3526,
      "step": 7389
    },
    {
      "epoch": 1.3674,
      "grad_norm": 5.722336769104004,
      "learning_rate": 1.318181818181818e-05,
      "loss": 0.3641,
      "step": 7390
    },
    {
      "epoch": 1.3675,
      "grad_norm": 9.498021125793457,
      "learning_rate": 1.3176767676767677e-05,
      "loss": 0.3614,
      "step": 7391
    },
    {
      "epoch": 1.3676,
      "grad_norm": 8.05918025970459,
      "learning_rate": 1.3171717171717171e-05,
      "loss": 0.3973,
      "step": 7392
    },
    {
      "epoch": 1.3677000000000001,
      "grad_norm": 10.006900787353516,
      "learning_rate": 1.3166666666666665e-05,
      "loss": 0.3634,
      "step": 7393
    },
    {
      "epoch": 1.3678,
      "grad_norm": 7.409060955047607,
      "learning_rate": 1.3161616161616161e-05,
      "loss": 0.2661,
      "step": 7394
    },
    {
      "epoch": 1.3679000000000001,
      "grad_norm": 8.935864448547363,
      "learning_rate": 1.3156565656565656e-05,
      "loss": 0.4514,
      "step": 7395
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 8.044816017150879,
      "learning_rate": 1.315151515151515e-05,
      "loss": 0.3184,
      "step": 7396
    },
    {
      "epoch": 1.3681,
      "grad_norm": 15.903424263000488,
      "learning_rate": 1.3146464646464648e-05,
      "loss": 0.4252,
      "step": 7397
    },
    {
      "epoch": 1.3682,
      "grad_norm": 7.801822185516357,
      "learning_rate": 1.3141414141414144e-05,
      "loss": 0.3085,
      "step": 7398
    },
    {
      "epoch": 1.3683,
      "grad_norm": 5.62069845199585,
      "learning_rate": 1.3136363636363638e-05,
      "loss": 0.2951,
      "step": 7399
    },
    {
      "epoch": 1.3684,
      "grad_norm": 12.350367546081543,
      "learning_rate": 1.3131313131313134e-05,
      "loss": 0.2905,
      "step": 7400
    },
    {
      "epoch": 1.3685,
      "grad_norm": 8.085210800170898,
      "learning_rate": 1.3126262626262628e-05,
      "loss": 0.3043,
      "step": 7401
    },
    {
      "epoch": 1.3686,
      "grad_norm": 8.482155799865723,
      "learning_rate": 1.3121212121212123e-05,
      "loss": 0.347,
      "step": 7402
    },
    {
      "epoch": 1.3687,
      "grad_norm": 5.908836841583252,
      "learning_rate": 1.3116161616161619e-05,
      "loss": 0.2639,
      "step": 7403
    },
    {
      "epoch": 1.3688,
      "grad_norm": 18.810089111328125,
      "learning_rate": 1.3111111111111113e-05,
      "loss": 0.4872,
      "step": 7404
    },
    {
      "epoch": 1.3689,
      "grad_norm": 16.656494140625,
      "learning_rate": 1.3106060606060607e-05,
      "loss": 0.4649,
      "step": 7405
    },
    {
      "epoch": 1.369,
      "grad_norm": 12.443870544433594,
      "learning_rate": 1.3101010101010103e-05,
      "loss": 0.4089,
      "step": 7406
    },
    {
      "epoch": 1.3691,
      "grad_norm": 11.817103385925293,
      "learning_rate": 1.3095959595959597e-05,
      "loss": 0.3933,
      "step": 7407
    },
    {
      "epoch": 1.3692,
      "grad_norm": 12.83425235748291,
      "learning_rate": 1.3090909090909093e-05,
      "loss": 0.3898,
      "step": 7408
    },
    {
      "epoch": 1.3693,
      "grad_norm": 10.635514259338379,
      "learning_rate": 1.3085858585858588e-05,
      "loss": 0.3813,
      "step": 7409
    },
    {
      "epoch": 1.3694,
      "grad_norm": 21.68118667602539,
      "learning_rate": 1.3080808080808082e-05,
      "loss": 0.3631,
      "step": 7410
    },
    {
      "epoch": 1.3695,
      "grad_norm": 13.755393028259277,
      "learning_rate": 1.3075757575757578e-05,
      "loss": 0.34,
      "step": 7411
    },
    {
      "epoch": 1.3696,
      "grad_norm": 13.838836669921875,
      "learning_rate": 1.3070707070707072e-05,
      "loss": 0.477,
      "step": 7412
    },
    {
      "epoch": 1.3697,
      "grad_norm": 17.64360809326172,
      "learning_rate": 1.3065656565656567e-05,
      "loss": 0.4115,
      "step": 7413
    },
    {
      "epoch": 1.3698000000000001,
      "grad_norm": 18.16439437866211,
      "learning_rate": 1.3060606060606063e-05,
      "loss": 0.4043,
      "step": 7414
    },
    {
      "epoch": 1.3699,
      "grad_norm": 16.635456085205078,
      "learning_rate": 1.3055555555555557e-05,
      "loss": 0.4285,
      "step": 7415
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.528682708740234,
      "learning_rate": 1.3050505050505051e-05,
      "loss": 0.4487,
      "step": 7416
    },
    {
      "epoch": 1.3700999999999999,
      "grad_norm": 23.81866455078125,
      "learning_rate": 1.3045454545454547e-05,
      "loss": 0.3836,
      "step": 7417
    },
    {
      "epoch": 1.3702,
      "grad_norm": 19.81641960144043,
      "learning_rate": 1.3040404040404041e-05,
      "loss": 0.5234,
      "step": 7418
    },
    {
      "epoch": 1.3703,
      "grad_norm": 9.73354434967041,
      "learning_rate": 1.3035353535353537e-05,
      "loss": 0.4154,
      "step": 7419
    },
    {
      "epoch": 1.3704,
      "grad_norm": 12.810078620910645,
      "learning_rate": 1.3030303030303032e-05,
      "loss": 0.3832,
      "step": 7420
    },
    {
      "epoch": 1.3705,
      "grad_norm": 8.613332748413086,
      "learning_rate": 1.3025252525252526e-05,
      "loss": 0.351,
      "step": 7421
    },
    {
      "epoch": 1.3706,
      "grad_norm": 10.458340644836426,
      "learning_rate": 1.3020202020202022e-05,
      "loss": 0.4177,
      "step": 7422
    },
    {
      "epoch": 1.3707,
      "grad_norm": 14.231201171875,
      "learning_rate": 1.3015151515151516e-05,
      "loss": 0.3578,
      "step": 7423
    },
    {
      "epoch": 1.3708,
      "grad_norm": 8.818279266357422,
      "learning_rate": 1.301010101010101e-05,
      "loss": 0.4153,
      "step": 7424
    },
    {
      "epoch": 1.3709,
      "grad_norm": 20.86869239807129,
      "learning_rate": 1.3005050505050507e-05,
      "loss": 0.3128,
      "step": 7425
    },
    {
      "epoch": 1.371,
      "grad_norm": 12.943617820739746,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.4115,
      "step": 7426
    },
    {
      "epoch": 1.3711,
      "grad_norm": 23.996679306030273,
      "learning_rate": 1.2994949494949495e-05,
      "loss": 0.3977,
      "step": 7427
    },
    {
      "epoch": 1.3712,
      "grad_norm": 13.174825668334961,
      "learning_rate": 1.2989898989898991e-05,
      "loss": 0.3875,
      "step": 7428
    },
    {
      "epoch": 1.3713,
      "grad_norm": 18.733308792114258,
      "learning_rate": 1.2984848484848485e-05,
      "loss": 0.4328,
      "step": 7429
    },
    {
      "epoch": 1.3714,
      "grad_norm": 20.44736099243164,
      "learning_rate": 1.2979797979797981e-05,
      "loss": 0.468,
      "step": 7430
    },
    {
      "epoch": 1.3715,
      "grad_norm": 14.648477554321289,
      "learning_rate": 1.2974747474747476e-05,
      "loss": 0.5359,
      "step": 7431
    },
    {
      "epoch": 1.3716,
      "grad_norm": 49.60794448852539,
      "learning_rate": 1.296969696969697e-05,
      "loss": 0.3751,
      "step": 7432
    },
    {
      "epoch": 2.0001,
      "grad_norm": 11.067770957946777,
      "learning_rate": 1.2964646464646466e-05,
      "loss": 0.5738,
      "step": 7433
    },
    {
      "epoch": 2.0002,
      "grad_norm": 6.534769058227539,
      "learning_rate": 1.295959595959596e-05,
      "loss": 0.3284,
      "step": 7434
    },
    {
      "epoch": 2.0003,
      "grad_norm": 27.10575294494629,
      "learning_rate": 1.2954545454545455e-05,
      "loss": 0.6017,
      "step": 7435
    },
    {
      "epoch": 2.0004,
      "grad_norm": 10.019984245300293,
      "learning_rate": 1.294949494949495e-05,
      "loss": 0.5729,
      "step": 7436
    },
    {
      "epoch": 2.0005,
      "grad_norm": 9.773133277893066,
      "learning_rate": 1.2944444444444445e-05,
      "loss": 0.6408,
      "step": 7437
    },
    {
      "epoch": 2.0006,
      "grad_norm": 7.424463272094727,
      "learning_rate": 1.293939393939394e-05,
      "loss": 0.3881,
      "step": 7438
    },
    {
      "epoch": 2.0007,
      "grad_norm": 12.544411659240723,
      "learning_rate": 1.2934343434343435e-05,
      "loss": 0.6236,
      "step": 7439
    },
    {
      "epoch": 2.0008,
      "grad_norm": 6.948230266571045,
      "learning_rate": 1.292929292929293e-05,
      "loss": 0.9936,
      "step": 7440
    },
    {
      "epoch": 2.0009,
      "grad_norm": 13.439285278320312,
      "learning_rate": 1.2924242424242426e-05,
      "loss": 0.4695,
      "step": 7441
    },
    {
      "epoch": 2.001,
      "grad_norm": 6.689608097076416,
      "learning_rate": 1.291919191919192e-05,
      "loss": 0.814,
      "step": 7442
    },
    {
      "epoch": 2.0011,
      "grad_norm": 12.433146476745605,
      "learning_rate": 1.2914141414141414e-05,
      "loss": 0.5923,
      "step": 7443
    },
    {
      "epoch": 2.0012,
      "grad_norm": 12.174345016479492,
      "learning_rate": 1.290909090909091e-05,
      "loss": 0.3397,
      "step": 7444
    },
    {
      "epoch": 2.0013,
      "grad_norm": 6.4756669998168945,
      "learning_rate": 1.2904040404040404e-05,
      "loss": 0.6228,
      "step": 7445
    },
    {
      "epoch": 2.0014,
      "grad_norm": 5.661073207855225,
      "learning_rate": 1.2898989898989899e-05,
      "loss": 0.5778,
      "step": 7446
    },
    {
      "epoch": 2.0015,
      "grad_norm": 6.128524303436279,
      "learning_rate": 1.2893939393939395e-05,
      "loss": 0.4074,
      "step": 7447
    },
    {
      "epoch": 2.0016,
      "grad_norm": 10.11327075958252,
      "learning_rate": 1.2888888888888889e-05,
      "loss": 1.0723,
      "step": 7448
    },
    {
      "epoch": 2.0017,
      "grad_norm": 5.530300617218018,
      "learning_rate": 1.2883838383838385e-05,
      "loss": 0.8128,
      "step": 7449
    },
    {
      "epoch": 2.0018,
      "grad_norm": 8.044440269470215,
      "learning_rate": 1.287878787878788e-05,
      "loss": 1.0236,
      "step": 7450
    },
    {
      "epoch": 2.0019,
      "grad_norm": 7.598689556121826,
      "learning_rate": 1.2873737373737374e-05,
      "loss": 0.2876,
      "step": 7451
    },
    {
      "epoch": 2.002,
      "grad_norm": 10.18588638305664,
      "learning_rate": 1.286868686868687e-05,
      "loss": 0.9261,
      "step": 7452
    },
    {
      "epoch": 2.0021,
      "grad_norm": 6.480282306671143,
      "learning_rate": 1.2863636363636364e-05,
      "loss": 1.0445,
      "step": 7453
    },
    {
      "epoch": 2.0022,
      "grad_norm": 3.903812885284424,
      "learning_rate": 1.2858585858585858e-05,
      "loss": 0.3251,
      "step": 7454
    },
    {
      "epoch": 2.0023,
      "grad_norm": 6.04302978515625,
      "learning_rate": 1.2853535353535354e-05,
      "loss": 0.1691,
      "step": 7455
    },
    {
      "epoch": 2.0024,
      "grad_norm": 5.280454635620117,
      "learning_rate": 1.2848484848484848e-05,
      "loss": 0.5799,
      "step": 7456
    },
    {
      "epoch": 2.0025,
      "grad_norm": 24.237531661987305,
      "learning_rate": 1.2843434343434343e-05,
      "loss": 0.4352,
      "step": 7457
    },
    {
      "epoch": 2.0026,
      "grad_norm": 6.6389384269714355,
      "learning_rate": 1.2838383838383839e-05,
      "loss": 0.4477,
      "step": 7458
    },
    {
      "epoch": 2.0027,
      "grad_norm": 12.442008018493652,
      "learning_rate": 1.2833333333333333e-05,
      "loss": 0.6982,
      "step": 7459
    },
    {
      "epoch": 2.0028,
      "grad_norm": 7.432277679443359,
      "learning_rate": 1.2828282828282829e-05,
      "loss": 0.3122,
      "step": 7460
    },
    {
      "epoch": 2.0029,
      "grad_norm": 4.25301456451416,
      "learning_rate": 1.2823232323232323e-05,
      "loss": 0.6632,
      "step": 7461
    },
    {
      "epoch": 2.003,
      "grad_norm": 4.820925712585449,
      "learning_rate": 1.2818181818181818e-05,
      "loss": 0.6762,
      "step": 7462
    },
    {
      "epoch": 2.0031,
      "grad_norm": 7.532407760620117,
      "learning_rate": 1.2813131313131314e-05,
      "loss": 0.4629,
      "step": 7463
    },
    {
      "epoch": 2.0032,
      "grad_norm": 8.892746925354004,
      "learning_rate": 1.2808080808080808e-05,
      "loss": 0.464,
      "step": 7464
    },
    {
      "epoch": 2.0033,
      "grad_norm": 15.651168823242188,
      "learning_rate": 1.2803030303030302e-05,
      "loss": 1.0429,
      "step": 7465
    },
    {
      "epoch": 2.0034,
      "grad_norm": 6.512859344482422,
      "learning_rate": 1.2797979797979798e-05,
      "loss": 0.7784,
      "step": 7466
    },
    {
      "epoch": 2.0035,
      "grad_norm": 5.247759819030762,
      "learning_rate": 1.2792929292929292e-05,
      "loss": 0.4397,
      "step": 7467
    },
    {
      "epoch": 2.0036,
      "grad_norm": 7.799246788024902,
      "learning_rate": 1.2787878787878787e-05,
      "loss": 0.9093,
      "step": 7468
    },
    {
      "epoch": 2.0037,
      "grad_norm": 5.9853596687316895,
      "learning_rate": 1.2782828282828283e-05,
      "loss": 1.1402,
      "step": 7469
    },
    {
      "epoch": 2.0038,
      "grad_norm": 8.222250938415527,
      "learning_rate": 1.2777777777777777e-05,
      "loss": 0.5901,
      "step": 7470
    },
    {
      "epoch": 2.0039,
      "grad_norm": 5.30120849609375,
      "learning_rate": 1.2772727272727273e-05,
      "loss": 0.2898,
      "step": 7471
    },
    {
      "epoch": 2.004,
      "grad_norm": 12.381537437438965,
      "learning_rate": 1.2767676767676767e-05,
      "loss": 0.2819,
      "step": 7472
    },
    {
      "epoch": 2.0041,
      "grad_norm": 5.207355976104736,
      "learning_rate": 1.2762626262626262e-05,
      "loss": 0.4923,
      "step": 7473
    },
    {
      "epoch": 2.0042,
      "grad_norm": 7.453055381774902,
      "learning_rate": 1.2757575757575758e-05,
      "loss": 0.5831,
      "step": 7474
    },
    {
      "epoch": 2.0043,
      "grad_norm": 10.600326538085938,
      "learning_rate": 1.2752525252525252e-05,
      "loss": 1.082,
      "step": 7475
    },
    {
      "epoch": 2.0044,
      "grad_norm": 11.97979736328125,
      "learning_rate": 1.2747474747474746e-05,
      "loss": 0.6045,
      "step": 7476
    },
    {
      "epoch": 2.0045,
      "grad_norm": 11.58600902557373,
      "learning_rate": 1.2742424242424242e-05,
      "loss": 0.6275,
      "step": 7477
    },
    {
      "epoch": 2.0046,
      "grad_norm": 19.97001075744629,
      "learning_rate": 1.2737373737373736e-05,
      "loss": 0.5151,
      "step": 7478
    },
    {
      "epoch": 2.0047,
      "grad_norm": 6.609193801879883,
      "learning_rate": 1.2732323232323232e-05,
      "loss": 0.265,
      "step": 7479
    },
    {
      "epoch": 2.0048,
      "grad_norm": 53.80645751953125,
      "learning_rate": 1.2727272727272727e-05,
      "loss": 0.4618,
      "step": 7480
    },
    {
      "epoch": 2.0049,
      "grad_norm": 9.497522354125977,
      "learning_rate": 1.2722222222222221e-05,
      "loss": 0.2446,
      "step": 7481
    },
    {
      "epoch": 2.005,
      "grad_norm": 5.536461353302002,
      "learning_rate": 1.2717171717171717e-05,
      "loss": 0.2438,
      "step": 7482
    },
    {
      "epoch": 2.0051,
      "grad_norm": 2.806544542312622,
      "learning_rate": 1.2712121212121215e-05,
      "loss": 0.5572,
      "step": 7483
    },
    {
      "epoch": 2.0052,
      "grad_norm": 5.041750907897949,
      "learning_rate": 1.2707070707070709e-05,
      "loss": 0.9875,
      "step": 7484
    },
    {
      "epoch": 2.0053,
      "grad_norm": 14.288222312927246,
      "learning_rate": 1.2702020202020203e-05,
      "loss": 0.89,
      "step": 7485
    },
    {
      "epoch": 2.0054,
      "grad_norm": 8.126935005187988,
      "learning_rate": 1.26969696969697e-05,
      "loss": 0.5868,
      "step": 7486
    },
    {
      "epoch": 2.0055,
      "grad_norm": 6.937923431396484,
      "learning_rate": 1.2691919191919194e-05,
      "loss": 0.6375,
      "step": 7487
    },
    {
      "epoch": 2.0056,
      "grad_norm": 39.820987701416016,
      "learning_rate": 1.2686868686868688e-05,
      "loss": 0.4519,
      "step": 7488
    },
    {
      "epoch": 2.0057,
      "grad_norm": 16.770347595214844,
      "learning_rate": 1.2681818181818184e-05,
      "loss": 1.0459,
      "step": 7489
    },
    {
      "epoch": 2.0058,
      "grad_norm": 15.585384368896484,
      "learning_rate": 1.2676767676767678e-05,
      "loss": 0.3925,
      "step": 7490
    },
    {
      "epoch": 2.0059,
      "grad_norm": 9.102439880371094,
      "learning_rate": 1.2671717171717174e-05,
      "loss": 0.5719,
      "step": 7491
    },
    {
      "epoch": 2.006,
      "grad_norm": 9.479997634887695,
      "learning_rate": 1.2666666666666668e-05,
      "loss": 0.3222,
      "step": 7492
    },
    {
      "epoch": 2.0061,
      "grad_norm": 12.419371604919434,
      "learning_rate": 1.2661616161616163e-05,
      "loss": 0.6608,
      "step": 7493
    },
    {
      "epoch": 2.0062,
      "grad_norm": 9.052971839904785,
      "learning_rate": 1.2656565656565659e-05,
      "loss": 0.5591,
      "step": 7494
    },
    {
      "epoch": 2.0063,
      "grad_norm": 6.024280548095703,
      "learning_rate": 1.2651515151515153e-05,
      "loss": 0.3364,
      "step": 7495
    },
    {
      "epoch": 2.0064,
      "grad_norm": 8.686830520629883,
      "learning_rate": 1.2646464646464647e-05,
      "loss": 1.2473,
      "step": 7496
    },
    {
      "epoch": 2.0065,
      "grad_norm": 4.80410623550415,
      "learning_rate": 1.2641414141414143e-05,
      "loss": 0.4308,
      "step": 7497
    },
    {
      "epoch": 2.0066,
      "grad_norm": 4.29066801071167,
      "learning_rate": 1.2636363636363638e-05,
      "loss": 0.2006,
      "step": 7498
    },
    {
      "epoch": 2.0067,
      "grad_norm": 8.845946311950684,
      "learning_rate": 1.2631313131313132e-05,
      "loss": 0.4287,
      "step": 7499
    },
    {
      "epoch": 2.0068,
      "grad_norm": 6.051027297973633,
      "learning_rate": 1.2626262626262628e-05,
      "loss": 0.1774,
      "step": 7500
    },
    {
      "epoch": 2.0069,
      "grad_norm": 6.9337358474731445,
      "learning_rate": 1.2621212121212122e-05,
      "loss": 0.8596,
      "step": 7501
    },
    {
      "epoch": 2.007,
      "grad_norm": 10.618175506591797,
      "learning_rate": 1.2616161616161618e-05,
      "loss": 0.829,
      "step": 7502
    },
    {
      "epoch": 2.0071,
      "grad_norm": 5.994649887084961,
      "learning_rate": 1.2611111111111113e-05,
      "loss": 0.8033,
      "step": 7503
    },
    {
      "epoch": 2.0072,
      "grad_norm": 16.134706497192383,
      "learning_rate": 1.2606060606060607e-05,
      "loss": 0.4895,
      "step": 7504
    },
    {
      "epoch": 2.0073,
      "grad_norm": 5.958493232727051,
      "learning_rate": 1.2601010101010103e-05,
      "loss": 0.8415,
      "step": 7505
    },
    {
      "epoch": 2.0074,
      "grad_norm": 4.522373676300049,
      "learning_rate": 1.2595959595959597e-05,
      "loss": 0.1932,
      "step": 7506
    },
    {
      "epoch": 2.0075,
      "grad_norm": 11.670877456665039,
      "learning_rate": 1.2590909090909091e-05,
      "loss": 0.3938,
      "step": 7507
    },
    {
      "epoch": 2.0076,
      "grad_norm": 25.850955963134766,
      "learning_rate": 1.2585858585858587e-05,
      "loss": 0.8806,
      "step": 7508
    },
    {
      "epoch": 2.0077,
      "grad_norm": 4.861108303070068,
      "learning_rate": 1.2580808080808082e-05,
      "loss": 0.2855,
      "step": 7509
    },
    {
      "epoch": 2.0078,
      "grad_norm": 5.991410732269287,
      "learning_rate": 1.2575757575757578e-05,
      "loss": 0.6704,
      "step": 7510
    },
    {
      "epoch": 2.0079,
      "grad_norm": 5.8068461418151855,
      "learning_rate": 1.2570707070707072e-05,
      "loss": 0.7586,
      "step": 7511
    },
    {
      "epoch": 2.008,
      "grad_norm": 44.81625747680664,
      "learning_rate": 1.2565656565656566e-05,
      "loss": 1.2766,
      "step": 7512
    },
    {
      "epoch": 2.0081,
      "grad_norm": 9.13131332397461,
      "learning_rate": 1.2560606060606062e-05,
      "loss": 0.7171,
      "step": 7513
    },
    {
      "epoch": 2.0082,
      "grad_norm": 5.927697658538818,
      "learning_rate": 1.2555555555555557e-05,
      "loss": 0.1887,
      "step": 7514
    },
    {
      "epoch": 2.0083,
      "grad_norm": 5.310041904449463,
      "learning_rate": 1.2550505050505051e-05,
      "loss": 0.4272,
      "step": 7515
    },
    {
      "epoch": 2.0084,
      "grad_norm": 7.670498847961426,
      "learning_rate": 1.2545454545454547e-05,
      "loss": 0.3278,
      "step": 7516
    },
    {
      "epoch": 2.0085,
      "grad_norm": 8.428955078125,
      "learning_rate": 1.2540404040404041e-05,
      "loss": 0.4056,
      "step": 7517
    },
    {
      "epoch": 2.0086,
      "grad_norm": 10.200645446777344,
      "learning_rate": 1.2535353535353535e-05,
      "loss": 0.8718,
      "step": 7518
    },
    {
      "epoch": 2.0087,
      "grad_norm": 8.069005966186523,
      "learning_rate": 1.2530303030303031e-05,
      "loss": 0.3119,
      "step": 7519
    },
    {
      "epoch": 2.0088,
      "grad_norm": 7.480169773101807,
      "learning_rate": 1.2525252525252526e-05,
      "loss": 0.2922,
      "step": 7520
    },
    {
      "epoch": 2.0089,
      "grad_norm": 8.024491310119629,
      "learning_rate": 1.2520202020202022e-05,
      "loss": 0.4855,
      "step": 7521
    },
    {
      "epoch": 2.009,
      "grad_norm": 4.800903797149658,
      "learning_rate": 1.2515151515151516e-05,
      "loss": 0.7937,
      "step": 7522
    },
    {
      "epoch": 2.0091,
      "grad_norm": 6.599670886993408,
      "learning_rate": 1.251010101010101e-05,
      "loss": 1.5748,
      "step": 7523
    },
    {
      "epoch": 2.0092,
      "grad_norm": 25.564294815063477,
      "learning_rate": 1.2505050505050506e-05,
      "loss": 1.2201,
      "step": 7524
    },
    {
      "epoch": 2.0093,
      "grad_norm": 9.775410652160645,
      "learning_rate": 1.25e-05,
      "loss": 0.8373,
      "step": 7525
    },
    {
      "epoch": 2.0094,
      "grad_norm": 8.979830741882324,
      "learning_rate": 1.2494949494949495e-05,
      "loss": 0.4967,
      "step": 7526
    },
    {
      "epoch": 2.0095,
      "grad_norm": 7.396960258483887,
      "learning_rate": 1.2489898989898991e-05,
      "loss": 0.6858,
      "step": 7527
    },
    {
      "epoch": 2.0096,
      "grad_norm": 11.498488426208496,
      "learning_rate": 1.2484848484848485e-05,
      "loss": 0.5592,
      "step": 7528
    },
    {
      "epoch": 2.0097,
      "grad_norm": 7.424448490142822,
      "learning_rate": 1.247979797979798e-05,
      "loss": 0.6811,
      "step": 7529
    },
    {
      "epoch": 2.0098,
      "grad_norm": 5.997277736663818,
      "learning_rate": 1.2474747474747475e-05,
      "loss": 0.2903,
      "step": 7530
    },
    {
      "epoch": 2.0099,
      "grad_norm": 8.194046020507812,
      "learning_rate": 1.246969696969697e-05,
      "loss": 0.5551,
      "step": 7531
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.260828018188477,
      "learning_rate": 1.2464646464646466e-05,
      "loss": 0.7071,
      "step": 7532
    },
    {
      "epoch": 2.0101,
      "grad_norm": 7.789144515991211,
      "learning_rate": 1.245959595959596e-05,
      "loss": 0.4537,
      "step": 7533
    },
    {
      "epoch": 2.0102,
      "grad_norm": 10.506525993347168,
      "learning_rate": 1.2454545454545454e-05,
      "loss": 0.3475,
      "step": 7534
    },
    {
      "epoch": 2.0103,
      "grad_norm": 4.725184917449951,
      "learning_rate": 1.244949494949495e-05,
      "loss": 0.4794,
      "step": 7535
    },
    {
      "epoch": 2.0104,
      "grad_norm": 18.916690826416016,
      "learning_rate": 1.2444444444444445e-05,
      "loss": 0.2961,
      "step": 7536
    },
    {
      "epoch": 2.0105,
      "grad_norm": 12.679047584533691,
      "learning_rate": 1.2439393939393939e-05,
      "loss": 0.3879,
      "step": 7537
    },
    {
      "epoch": 2.0106,
      "grad_norm": 6.6373677253723145,
      "learning_rate": 1.2434343434343435e-05,
      "loss": 0.5687,
      "step": 7538
    },
    {
      "epoch": 2.0107,
      "grad_norm": 10.552349090576172,
      "learning_rate": 1.242929292929293e-05,
      "loss": 0.5288,
      "step": 7539
    },
    {
      "epoch": 2.0108,
      "grad_norm": 13.288294792175293,
      "learning_rate": 1.2424242424242424e-05,
      "loss": 0.3449,
      "step": 7540
    },
    {
      "epoch": 2.0109,
      "grad_norm": 7.585926055908203,
      "learning_rate": 1.241919191919192e-05,
      "loss": 1.1246,
      "step": 7541
    },
    {
      "epoch": 2.011,
      "grad_norm": 7.630847454071045,
      "learning_rate": 1.2414141414141414e-05,
      "loss": 0.2838,
      "step": 7542
    },
    {
      "epoch": 2.0111,
      "grad_norm": 17.53959846496582,
      "learning_rate": 1.240909090909091e-05,
      "loss": 0.3753,
      "step": 7543
    },
    {
      "epoch": 2.0112,
      "grad_norm": 5.543396949768066,
      "learning_rate": 1.2404040404040404e-05,
      "loss": 0.5631,
      "step": 7544
    },
    {
      "epoch": 2.0113,
      "grad_norm": 21.721364974975586,
      "learning_rate": 1.2398989898989898e-05,
      "loss": 0.7879,
      "step": 7545
    },
    {
      "epoch": 2.0114,
      "grad_norm": 8.47199535369873,
      "learning_rate": 1.2393939393939394e-05,
      "loss": 0.696,
      "step": 7546
    },
    {
      "epoch": 2.0115,
      "grad_norm": 5.511487007141113,
      "learning_rate": 1.238888888888889e-05,
      "loss": 0.7087,
      "step": 7547
    },
    {
      "epoch": 2.0116,
      "grad_norm": 4.642251491546631,
      "learning_rate": 1.2383838383838385e-05,
      "loss": 0.4547,
      "step": 7548
    },
    {
      "epoch": 2.0117,
      "grad_norm": 8.975547790527344,
      "learning_rate": 1.237878787878788e-05,
      "loss": 0.5447,
      "step": 7549
    },
    {
      "epoch": 2.0118,
      "grad_norm": 2.59200119972229,
      "learning_rate": 1.2373737373737375e-05,
      "loss": 0.5191,
      "step": 7550
    },
    {
      "epoch": 2.0119,
      "grad_norm": 4.6559672355651855,
      "learning_rate": 1.236868686868687e-05,
      "loss": 0.267,
      "step": 7551
    },
    {
      "epoch": 2.012,
      "grad_norm": 7.013993263244629,
      "learning_rate": 1.2363636363636365e-05,
      "loss": 0.342,
      "step": 7552
    },
    {
      "epoch": 2.0121,
      "grad_norm": 6.771530628204346,
      "learning_rate": 1.235858585858586e-05,
      "loss": 0.2263,
      "step": 7553
    },
    {
      "epoch": 2.0122,
      "grad_norm": 8.161633491516113,
      "learning_rate": 1.2353535353535354e-05,
      "loss": 0.4426,
      "step": 7554
    },
    {
      "epoch": 2.0123,
      "grad_norm": 10.495365142822266,
      "learning_rate": 1.234848484848485e-05,
      "loss": 0.5062,
      "step": 7555
    },
    {
      "epoch": 2.0124,
      "grad_norm": 15.599736213684082,
      "learning_rate": 1.2343434343434344e-05,
      "loss": 0.5478,
      "step": 7556
    },
    {
      "epoch": 2.0125,
      "grad_norm": 7.5674967765808105,
      "learning_rate": 1.233838383838384e-05,
      "loss": 0.3776,
      "step": 7557
    },
    {
      "epoch": 2.0126,
      "grad_norm": 6.550327301025391,
      "learning_rate": 1.2333333333333334e-05,
      "loss": 0.4482,
      "step": 7558
    },
    {
      "epoch": 2.0127,
      "grad_norm": 9.436637878417969,
      "learning_rate": 1.2328282828282829e-05,
      "loss": 0.3905,
      "step": 7559
    },
    {
      "epoch": 2.0128,
      "grad_norm": 5.886406898498535,
      "learning_rate": 1.2323232323232325e-05,
      "loss": 0.5001,
      "step": 7560
    },
    {
      "epoch": 2.0129,
      "grad_norm": 91.49087524414062,
      "learning_rate": 1.2318181818181819e-05,
      "loss": 0.4305,
      "step": 7561
    },
    {
      "epoch": 2.013,
      "grad_norm": 4.2274556159973145,
      "learning_rate": 1.2313131313131313e-05,
      "loss": 0.2517,
      "step": 7562
    },
    {
      "epoch": 2.0131,
      "grad_norm": 8.681233406066895,
      "learning_rate": 1.230808080808081e-05,
      "loss": 0.7816,
      "step": 7563
    },
    {
      "epoch": 2.0132,
      "grad_norm": 6.418660640716553,
      "learning_rate": 1.2303030303030304e-05,
      "loss": 0.2899,
      "step": 7564
    },
    {
      "epoch": 2.0133,
      "grad_norm": 5.817200183868408,
      "learning_rate": 1.2297979797979798e-05,
      "loss": 0.4548,
      "step": 7565
    },
    {
      "epoch": 2.0134,
      "grad_norm": 7.738798141479492,
      "learning_rate": 1.2292929292929294e-05,
      "loss": 0.1911,
      "step": 7566
    },
    {
      "epoch": 2.0135,
      "grad_norm": 4.936166286468506,
      "learning_rate": 1.2287878787878788e-05,
      "loss": 0.2209,
      "step": 7567
    },
    {
      "epoch": 2.0136,
      "grad_norm": 4.07191801071167,
      "learning_rate": 1.2282828282828284e-05,
      "loss": 0.2322,
      "step": 7568
    },
    {
      "epoch": 2.0137,
      "grad_norm": 6.054876804351807,
      "learning_rate": 1.2277777777777778e-05,
      "loss": 0.3871,
      "step": 7569
    },
    {
      "epoch": 2.0138,
      "grad_norm": 5.260079383850098,
      "learning_rate": 1.2272727272727273e-05,
      "loss": 0.3212,
      "step": 7570
    },
    {
      "epoch": 2.0139,
      "grad_norm": 6.362955093383789,
      "learning_rate": 1.2267676767676769e-05,
      "loss": 0.6587,
      "step": 7571
    },
    {
      "epoch": 2.014,
      "grad_norm": 6.419497489929199,
      "learning_rate": 1.2262626262626263e-05,
      "loss": 0.6614,
      "step": 7572
    },
    {
      "epoch": 2.0141,
      "grad_norm": 5.56779146194458,
      "learning_rate": 1.2257575757575757e-05,
      "loss": 0.9174,
      "step": 7573
    },
    {
      "epoch": 2.0142,
      "grad_norm": 6.433098793029785,
      "learning_rate": 1.2252525252525253e-05,
      "loss": 0.604,
      "step": 7574
    },
    {
      "epoch": 2.0143,
      "grad_norm": 7.635933876037598,
      "learning_rate": 1.2247474747474748e-05,
      "loss": 0.5898,
      "step": 7575
    },
    {
      "epoch": 2.0144,
      "grad_norm": 8.599373817443848,
      "learning_rate": 1.2242424242424242e-05,
      "loss": 0.4186,
      "step": 7576
    },
    {
      "epoch": 2.0145,
      "grad_norm": 5.704301357269287,
      "learning_rate": 1.2237373737373738e-05,
      "loss": 1.2187,
      "step": 7577
    },
    {
      "epoch": 2.0146,
      "grad_norm": 7.480515480041504,
      "learning_rate": 1.2232323232323232e-05,
      "loss": 0.5491,
      "step": 7578
    },
    {
      "epoch": 2.0147,
      "grad_norm": 9.428521156311035,
      "learning_rate": 1.2227272727272728e-05,
      "loss": 0.7779,
      "step": 7579
    },
    {
      "epoch": 2.0148,
      "grad_norm": 5.712530136108398,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 0.5363,
      "step": 7580
    },
    {
      "epoch": 2.0149,
      "grad_norm": 5.863683700561523,
      "learning_rate": 1.2217171717171717e-05,
      "loss": 0.2062,
      "step": 7581
    },
    {
      "epoch": 2.015,
      "grad_norm": 8.178627967834473,
      "learning_rate": 1.2212121212121213e-05,
      "loss": 0.2572,
      "step": 7582
    },
    {
      "epoch": 2.0151,
      "grad_norm": 8.554887771606445,
      "learning_rate": 1.2207070707070707e-05,
      "loss": 0.6571,
      "step": 7583
    },
    {
      "epoch": 2.0152,
      "grad_norm": 26.343303680419922,
      "learning_rate": 1.2202020202020201e-05,
      "loss": 0.3839,
      "step": 7584
    },
    {
      "epoch": 2.0153,
      "grad_norm": 7.7416090965271,
      "learning_rate": 1.2196969696969697e-05,
      "loss": 0.4667,
      "step": 7585
    },
    {
      "epoch": 2.0154,
      "grad_norm": 5.817751884460449,
      "learning_rate": 1.2191919191919192e-05,
      "loss": 0.4917,
      "step": 7586
    },
    {
      "epoch": 2.0155,
      "grad_norm": 9.030854225158691,
      "learning_rate": 1.2186868686868688e-05,
      "loss": 0.8719,
      "step": 7587
    },
    {
      "epoch": 2.0156,
      "grad_norm": 8.89795970916748,
      "learning_rate": 1.2181818181818182e-05,
      "loss": 0.457,
      "step": 7588
    },
    {
      "epoch": 2.0157,
      "grad_norm": 6.779231071472168,
      "learning_rate": 1.2176767676767676e-05,
      "loss": 1.406,
      "step": 7589
    },
    {
      "epoch": 2.0158,
      "grad_norm": 12.161600112915039,
      "learning_rate": 1.2171717171717172e-05,
      "loss": 0.5768,
      "step": 7590
    },
    {
      "epoch": 2.0159,
      "grad_norm": 7.5444254875183105,
      "learning_rate": 1.2166666666666668e-05,
      "loss": 0.4814,
      "step": 7591
    },
    {
      "epoch": 2.016,
      "grad_norm": 16.66742515563965,
      "learning_rate": 1.2161616161616162e-05,
      "loss": 0.7642,
      "step": 7592
    },
    {
      "epoch": 2.0161,
      "grad_norm": 6.385401725769043,
      "learning_rate": 1.2156565656565658e-05,
      "loss": 0.3484,
      "step": 7593
    },
    {
      "epoch": 2.0162,
      "grad_norm": 4.291913986206055,
      "learning_rate": 1.2151515151515153e-05,
      "loss": 0.1912,
      "step": 7594
    },
    {
      "epoch": 2.0163,
      "grad_norm": 7.526479244232178,
      "learning_rate": 1.2146464646464647e-05,
      "loss": 0.6292,
      "step": 7595
    },
    {
      "epoch": 2.0164,
      "grad_norm": 3.7319703102111816,
      "learning_rate": 1.2141414141414143e-05,
      "loss": 0.3903,
      "step": 7596
    },
    {
      "epoch": 2.0165,
      "grad_norm": 9.494080543518066,
      "learning_rate": 1.2136363636363637e-05,
      "loss": 0.4459,
      "step": 7597
    },
    {
      "epoch": 2.0166,
      "grad_norm": 7.624109745025635,
      "learning_rate": 1.2131313131313132e-05,
      "loss": 0.2852,
      "step": 7598
    },
    {
      "epoch": 2.0167,
      "grad_norm": 5.661098003387451,
      "learning_rate": 1.2126262626262628e-05,
      "loss": 0.2874,
      "step": 7599
    },
    {
      "epoch": 2.0168,
      "grad_norm": 8.424521446228027,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.8643,
      "step": 7600
    },
    {
      "epoch": 2.0169,
      "grad_norm": 5.347314834594727,
      "learning_rate": 1.2116161616161616e-05,
      "loss": 0.3051,
      "step": 7601
    },
    {
      "epoch": 2.017,
      "grad_norm": 6.26395845413208,
      "learning_rate": 1.2111111111111112e-05,
      "loss": 0.4634,
      "step": 7602
    },
    {
      "epoch": 2.0171,
      "grad_norm": 9.677337646484375,
      "learning_rate": 1.2106060606060607e-05,
      "loss": 0.7102,
      "step": 7603
    },
    {
      "epoch": 2.0172,
      "grad_norm": 9.14966869354248,
      "learning_rate": 1.2101010101010103e-05,
      "loss": 1.0493,
      "step": 7604
    },
    {
      "epoch": 2.0173,
      "grad_norm": 4.887302398681641,
      "learning_rate": 1.2095959595959597e-05,
      "loss": 0.3952,
      "step": 7605
    },
    {
      "epoch": 2.0174,
      "grad_norm": 4.629301071166992,
      "learning_rate": 1.2090909090909091e-05,
      "loss": 0.6627,
      "step": 7606
    },
    {
      "epoch": 2.0175,
      "grad_norm": 4.6270432472229,
      "learning_rate": 1.2085858585858587e-05,
      "loss": 0.3923,
      "step": 7607
    },
    {
      "epoch": 2.0176,
      "grad_norm": 7.1817498207092285,
      "learning_rate": 1.2080808080808081e-05,
      "loss": 0.6412,
      "step": 7608
    },
    {
      "epoch": 2.0177,
      "grad_norm": 3.7551450729370117,
      "learning_rate": 1.2075757575757576e-05,
      "loss": 0.21,
      "step": 7609
    },
    {
      "epoch": 2.0178,
      "grad_norm": 5.514759540557861,
      "learning_rate": 1.2070707070707072e-05,
      "loss": 0.2659,
      "step": 7610
    },
    {
      "epoch": 2.0179,
      "grad_norm": 6.047269344329834,
      "learning_rate": 1.2065656565656566e-05,
      "loss": 1.1,
      "step": 7611
    },
    {
      "epoch": 2.018,
      "grad_norm": 5.347559452056885,
      "learning_rate": 1.2060606060606062e-05,
      "loss": 0.4582,
      "step": 7612
    },
    {
      "epoch": 2.0181,
      "grad_norm": 8.967748641967773,
      "learning_rate": 1.2055555555555556e-05,
      "loss": 0.9328,
      "step": 7613
    },
    {
      "epoch": 2.0182,
      "grad_norm": 11.436724662780762,
      "learning_rate": 1.205050505050505e-05,
      "loss": 0.6793,
      "step": 7614
    },
    {
      "epoch": 2.0183,
      "grad_norm": 5.264969348907471,
      "learning_rate": 1.2045454545454547e-05,
      "loss": 0.3067,
      "step": 7615
    },
    {
      "epoch": 2.0184,
      "grad_norm": 17.36578369140625,
      "learning_rate": 1.2040404040404041e-05,
      "loss": 0.4001,
      "step": 7616
    },
    {
      "epoch": 2.0185,
      "grad_norm": 25.61361312866211,
      "learning_rate": 1.2035353535353535e-05,
      "loss": 0.8769,
      "step": 7617
    },
    {
      "epoch": 2.0186,
      "grad_norm": 13.040432929992676,
      "learning_rate": 1.2030303030303031e-05,
      "loss": 0.9506,
      "step": 7618
    },
    {
      "epoch": 2.0187,
      "grad_norm": 6.463250160217285,
      "learning_rate": 1.2025252525252525e-05,
      "loss": 0.6011,
      "step": 7619
    },
    {
      "epoch": 2.0188,
      "grad_norm": 3.869845151901245,
      "learning_rate": 1.202020202020202e-05,
      "loss": 0.1951,
      "step": 7620
    },
    {
      "epoch": 2.0189,
      "grad_norm": 7.886239051818848,
      "learning_rate": 1.2015151515151516e-05,
      "loss": 0.4984,
      "step": 7621
    },
    {
      "epoch": 2.019,
      "grad_norm": 7.328726291656494,
      "learning_rate": 1.201010101010101e-05,
      "loss": 0.7146,
      "step": 7622
    },
    {
      "epoch": 2.0191,
      "grad_norm": 6.469283580780029,
      "learning_rate": 1.2005050505050506e-05,
      "loss": 0.4083,
      "step": 7623
    },
    {
      "epoch": 2.0192,
      "grad_norm": 7.062155723571777,
      "learning_rate": 1.2e-05,
      "loss": 0.7786,
      "step": 7624
    },
    {
      "epoch": 2.0193,
      "grad_norm": 7.276419162750244,
      "learning_rate": 1.1994949494949495e-05,
      "loss": 0.3233,
      "step": 7625
    },
    {
      "epoch": 2.0194,
      "grad_norm": 5.267397403717041,
      "learning_rate": 1.198989898989899e-05,
      "loss": 0.3051,
      "step": 7626
    },
    {
      "epoch": 2.0195,
      "grad_norm": 4.464624881744385,
      "learning_rate": 1.1984848484848485e-05,
      "loss": 0.3191,
      "step": 7627
    },
    {
      "epoch": 2.0196,
      "grad_norm": 8.40725326538086,
      "learning_rate": 1.197979797979798e-05,
      "loss": 0.5987,
      "step": 7628
    },
    {
      "epoch": 2.0197,
      "grad_norm": 23.059499740600586,
      "learning_rate": 1.1974747474747475e-05,
      "loss": 0.7674,
      "step": 7629
    },
    {
      "epoch": 2.0198,
      "grad_norm": 3.428713798522949,
      "learning_rate": 1.196969696969697e-05,
      "loss": 0.0935,
      "step": 7630
    },
    {
      "epoch": 2.0199,
      "grad_norm": 16.54574203491211,
      "learning_rate": 1.1964646464646464e-05,
      "loss": 0.5325,
      "step": 7631
    },
    {
      "epoch": 2.02,
      "grad_norm": 6.640668869018555,
      "learning_rate": 1.1959595959595961e-05,
      "loss": 0.1637,
      "step": 7632
    },
    {
      "epoch": 2.0201,
      "grad_norm": 5.921253681182861,
      "learning_rate": 1.1954545454545456e-05,
      "loss": 0.2865,
      "step": 7633
    },
    {
      "epoch": 2.0202,
      "grad_norm": 7.042607307434082,
      "learning_rate": 1.194949494949495e-05,
      "loss": 0.2824,
      "step": 7634
    },
    {
      "epoch": 2.0203,
      "grad_norm": 11.211319923400879,
      "learning_rate": 1.1944444444444446e-05,
      "loss": 1.0101,
      "step": 7635
    },
    {
      "epoch": 2.0204,
      "grad_norm": 10.66618537902832,
      "learning_rate": 1.193939393939394e-05,
      "loss": 0.5209,
      "step": 7636
    },
    {
      "epoch": 2.0205,
      "grad_norm": 28.056480407714844,
      "learning_rate": 1.1934343434343435e-05,
      "loss": 0.5101,
      "step": 7637
    },
    {
      "epoch": 2.0206,
      "grad_norm": 8.141128540039062,
      "learning_rate": 1.192929292929293e-05,
      "loss": 0.6334,
      "step": 7638
    },
    {
      "epoch": 2.0207,
      "grad_norm": 5.6327009201049805,
      "learning_rate": 1.1924242424242425e-05,
      "loss": 0.1704,
      "step": 7639
    },
    {
      "epoch": 2.0208,
      "grad_norm": 5.895510196685791,
      "learning_rate": 1.1919191919191921e-05,
      "loss": 0.5567,
      "step": 7640
    },
    {
      "epoch": 2.0209,
      "grad_norm": 6.311949729919434,
      "learning_rate": 1.1914141414141415e-05,
      "loss": 0.4464,
      "step": 7641
    },
    {
      "epoch": 2.021,
      "grad_norm": 5.80855131149292,
      "learning_rate": 1.190909090909091e-05,
      "loss": 0.1584,
      "step": 7642
    },
    {
      "epoch": 2.0211,
      "grad_norm": 6.673916816711426,
      "learning_rate": 1.1904040404040405e-05,
      "loss": 0.436,
      "step": 7643
    },
    {
      "epoch": 2.0212,
      "grad_norm": 7.162682056427002,
      "learning_rate": 1.18989898989899e-05,
      "loss": 0.2686,
      "step": 7644
    },
    {
      "epoch": 2.0213,
      "grad_norm": 7.283148765563965,
      "learning_rate": 1.1893939393939394e-05,
      "loss": 0.274,
      "step": 7645
    },
    {
      "epoch": 2.0214,
      "grad_norm": 6.108485698699951,
      "learning_rate": 1.188888888888889e-05,
      "loss": 0.4983,
      "step": 7646
    },
    {
      "epoch": 2.0215,
      "grad_norm": 5.118884086608887,
      "learning_rate": 1.1883838383838384e-05,
      "loss": 0.2797,
      "step": 7647
    },
    {
      "epoch": 2.0216,
      "grad_norm": 6.932680606842041,
      "learning_rate": 1.187878787878788e-05,
      "loss": 0.589,
      "step": 7648
    },
    {
      "epoch": 2.0217,
      "grad_norm": 3.919480562210083,
      "learning_rate": 1.1873737373737375e-05,
      "loss": 0.3946,
      "step": 7649
    },
    {
      "epoch": 2.0218,
      "grad_norm": 6.817378044128418,
      "learning_rate": 1.1868686868686869e-05,
      "loss": 0.5428,
      "step": 7650
    },
    {
      "epoch": 2.0219,
      "grad_norm": 6.432758808135986,
      "learning_rate": 1.1863636363636365e-05,
      "loss": 0.4996,
      "step": 7651
    },
    {
      "epoch": 2.022,
      "grad_norm": 5.424415588378906,
      "learning_rate": 1.185858585858586e-05,
      "loss": 0.2417,
      "step": 7652
    },
    {
      "epoch": 2.0221,
      "grad_norm": 8.016623497009277,
      "learning_rate": 1.1853535353535354e-05,
      "loss": 1.8412,
      "step": 7653
    },
    {
      "epoch": 2.0222,
      "grad_norm": 7.829955577850342,
      "learning_rate": 1.184848484848485e-05,
      "loss": 0.3401,
      "step": 7654
    },
    {
      "epoch": 2.0223,
      "grad_norm": 3.9329471588134766,
      "learning_rate": 1.1843434343434344e-05,
      "loss": 0.1895,
      "step": 7655
    },
    {
      "epoch": 2.0224,
      "grad_norm": 6.703199863433838,
      "learning_rate": 1.1838383838383838e-05,
      "loss": 0.7711,
      "step": 7656
    },
    {
      "epoch": 2.0225,
      "grad_norm": 8.219666481018066,
      "learning_rate": 1.1833333333333334e-05,
      "loss": 0.6008,
      "step": 7657
    },
    {
      "epoch": 2.0226,
      "grad_norm": 14.06610107421875,
      "learning_rate": 1.1828282828282828e-05,
      "loss": 1.034,
      "step": 7658
    },
    {
      "epoch": 2.0227,
      "grad_norm": 6.037161350250244,
      "learning_rate": 1.1823232323232324e-05,
      "loss": 0.6113,
      "step": 7659
    },
    {
      "epoch": 2.0228,
      "grad_norm": 8.589093208312988,
      "learning_rate": 1.1818181818181819e-05,
      "loss": 0.513,
      "step": 7660
    },
    {
      "epoch": 2.0229,
      "grad_norm": 10.957663536071777,
      "learning_rate": 1.1813131313131313e-05,
      "loss": 0.9234,
      "step": 7661
    },
    {
      "epoch": 2.023,
      "grad_norm": 4.592494010925293,
      "learning_rate": 1.1808080808080809e-05,
      "loss": 0.571,
      "step": 7662
    },
    {
      "epoch": 2.0231,
      "grad_norm": 4.128093719482422,
      "learning_rate": 1.1803030303030303e-05,
      "loss": 0.2401,
      "step": 7663
    },
    {
      "epoch": 2.0232,
      "grad_norm": 7.452649116516113,
      "learning_rate": 1.1797979797979798e-05,
      "loss": 0.2865,
      "step": 7664
    },
    {
      "epoch": 2.0233,
      "grad_norm": 5.538349628448486,
      "learning_rate": 1.1792929292929294e-05,
      "loss": 0.1805,
      "step": 7665
    },
    {
      "epoch": 2.0234,
      "grad_norm": 7.183323860168457,
      "learning_rate": 1.1787878787878788e-05,
      "loss": 0.4745,
      "step": 7666
    },
    {
      "epoch": 2.0235,
      "grad_norm": 5.398412704467773,
      "learning_rate": 1.1782828282828282e-05,
      "loss": 0.2753,
      "step": 7667
    },
    {
      "epoch": 2.0236,
      "grad_norm": 6.364796161651611,
      "learning_rate": 1.1777777777777778e-05,
      "loss": 0.3656,
      "step": 7668
    },
    {
      "epoch": 2.0237,
      "grad_norm": 4.901348114013672,
      "learning_rate": 1.1772727272727272e-05,
      "loss": 0.2299,
      "step": 7669
    },
    {
      "epoch": 2.0238,
      "grad_norm": 7.517785549163818,
      "learning_rate": 1.1767676767676768e-05,
      "loss": 0.7735,
      "step": 7670
    },
    {
      "epoch": 2.0239,
      "grad_norm": 5.054324626922607,
      "learning_rate": 1.1762626262626263e-05,
      "loss": 0.7503,
      "step": 7671
    },
    {
      "epoch": 2.024,
      "grad_norm": 5.99261474609375,
      "learning_rate": 1.1757575757575757e-05,
      "loss": 0.3924,
      "step": 7672
    },
    {
      "epoch": 2.0241,
      "grad_norm": 17.79062843322754,
      "learning_rate": 1.1752525252525253e-05,
      "loss": 0.6605,
      "step": 7673
    },
    {
      "epoch": 2.0242,
      "grad_norm": 7.122779846191406,
      "learning_rate": 1.1747474747474747e-05,
      "loss": 0.8915,
      "step": 7674
    },
    {
      "epoch": 2.0243,
      "grad_norm": 7.516395568847656,
      "learning_rate": 1.1742424242424243e-05,
      "loss": 0.6562,
      "step": 7675
    },
    {
      "epoch": 2.0244,
      "grad_norm": 7.660529136657715,
      "learning_rate": 1.173737373737374e-05,
      "loss": 0.2425,
      "step": 7676
    },
    {
      "epoch": 2.0245,
      "grad_norm": 7.689420700073242,
      "learning_rate": 1.1732323232323234e-05,
      "loss": 0.7892,
      "step": 7677
    },
    {
      "epoch": 2.0246,
      "grad_norm": 6.814364910125732,
      "learning_rate": 1.1727272727272728e-05,
      "loss": 0.9022,
      "step": 7678
    },
    {
      "epoch": 2.0247,
      "grad_norm": 8.376973152160645,
      "learning_rate": 1.1722222222222224e-05,
      "loss": 0.3862,
      "step": 7679
    },
    {
      "epoch": 2.0248,
      "grad_norm": 4.986111640930176,
      "learning_rate": 1.1717171717171718e-05,
      "loss": 0.252,
      "step": 7680
    },
    {
      "epoch": 2.0249,
      "grad_norm": 6.350673198699951,
      "learning_rate": 1.1712121212121212e-05,
      "loss": 0.5014,
      "step": 7681
    },
    {
      "epoch": 2.025,
      "grad_norm": 7.840259075164795,
      "learning_rate": 1.1707070707070708e-05,
      "loss": 0.8794,
      "step": 7682
    },
    {
      "epoch": 2.0251,
      "grad_norm": 7.7516984939575195,
      "learning_rate": 1.1702020202020203e-05,
      "loss": 0.7192,
      "step": 7683
    },
    {
      "epoch": 2.0252,
      "grad_norm": 16.179847717285156,
      "learning_rate": 1.1696969696969699e-05,
      "loss": 0.4618,
      "step": 7684
    },
    {
      "epoch": 2.0253,
      "grad_norm": 6.102339744567871,
      "learning_rate": 1.1691919191919193e-05,
      "loss": 0.6979,
      "step": 7685
    },
    {
      "epoch": 2.0254,
      "grad_norm": 7.681690216064453,
      "learning_rate": 1.1686868686868687e-05,
      "loss": 0.3164,
      "step": 7686
    },
    {
      "epoch": 2.0255,
      "grad_norm": 13.718053817749023,
      "learning_rate": 1.1681818181818183e-05,
      "loss": 0.7008,
      "step": 7687
    },
    {
      "epoch": 2.0256,
      "grad_norm": 6.937062740325928,
      "learning_rate": 1.1676767676767678e-05,
      "loss": 0.6133,
      "step": 7688
    },
    {
      "epoch": 2.0257,
      "grad_norm": 40.071685791015625,
      "learning_rate": 1.1671717171717172e-05,
      "loss": 0.305,
      "step": 7689
    },
    {
      "epoch": 2.0258,
      "grad_norm": 5.879273414611816,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7751,
      "step": 7690
    },
    {
      "epoch": 2.0259,
      "grad_norm": 4.282869338989258,
      "learning_rate": 1.1661616161616162e-05,
      "loss": 0.1453,
      "step": 7691
    },
    {
      "epoch": 2.026,
      "grad_norm": 7.287415027618408,
      "learning_rate": 1.1656565656565656e-05,
      "loss": 0.1484,
      "step": 7692
    },
    {
      "epoch": 2.0261,
      "grad_norm": 12.254324913024902,
      "learning_rate": 1.1651515151515152e-05,
      "loss": 0.511,
      "step": 7693
    },
    {
      "epoch": 2.0262000000000002,
      "grad_norm": 13.545607566833496,
      "learning_rate": 1.1646464646464647e-05,
      "loss": 0.3884,
      "step": 7694
    },
    {
      "epoch": 2.0263,
      "grad_norm": 6.174824237823486,
      "learning_rate": 1.1641414141414143e-05,
      "loss": 0.6104,
      "step": 7695
    },
    {
      "epoch": 2.0264,
      "grad_norm": 10.100305557250977,
      "learning_rate": 1.1636363636363637e-05,
      "loss": 0.3418,
      "step": 7696
    },
    {
      "epoch": 2.0265,
      "grad_norm": 4.135922431945801,
      "learning_rate": 1.1631313131313131e-05,
      "loss": 0.8143,
      "step": 7697
    },
    {
      "epoch": 2.0266,
      "grad_norm": 14.331792831420898,
      "learning_rate": 1.1626262626262627e-05,
      "loss": 0.1368,
      "step": 7698
    },
    {
      "epoch": 2.0267,
      "grad_norm": 8.359907150268555,
      "learning_rate": 1.1621212121212122e-05,
      "loss": 0.5896,
      "step": 7699
    },
    {
      "epoch": 2.0268,
      "grad_norm": 7.061450958251953,
      "learning_rate": 1.1616161616161616e-05,
      "loss": 0.2963,
      "step": 7700
    },
    {
      "epoch": 2.0269,
      "grad_norm": 8.046835899353027,
      "learning_rate": 1.1611111111111112e-05,
      "loss": 1.067,
      "step": 7701
    },
    {
      "epoch": 2.027,
      "grad_norm": 6.253084182739258,
      "learning_rate": 1.1606060606060606e-05,
      "loss": 0.6853,
      "step": 7702
    },
    {
      "epoch": 2.0271,
      "grad_norm": 4.70768404006958,
      "learning_rate": 1.16010101010101e-05,
      "loss": 0.1341,
      "step": 7703
    },
    {
      "epoch": 2.0272,
      "grad_norm": 5.513238906860352,
      "learning_rate": 1.1595959595959597e-05,
      "loss": 0.317,
      "step": 7704
    },
    {
      "epoch": 2.0273,
      "grad_norm": 5.508908271789551,
      "learning_rate": 1.159090909090909e-05,
      "loss": 0.1442,
      "step": 7705
    },
    {
      "epoch": 2.0274,
      "grad_norm": 5.622928619384766,
      "learning_rate": 1.1585858585858587e-05,
      "loss": 0.7024,
      "step": 7706
    },
    {
      "epoch": 2.0275,
      "grad_norm": 7.220510482788086,
      "learning_rate": 1.1580808080808081e-05,
      "loss": 1.0002,
      "step": 7707
    },
    {
      "epoch": 2.0276,
      "grad_norm": 9.027481079101562,
      "learning_rate": 1.1575757575757575e-05,
      "loss": 0.2156,
      "step": 7708
    },
    {
      "epoch": 2.0277,
      "grad_norm": 6.102471351623535,
      "learning_rate": 1.1570707070707071e-05,
      "loss": 0.8401,
      "step": 7709
    },
    {
      "epoch": 2.0278,
      "grad_norm": 4.369470596313477,
      "learning_rate": 1.1565656565656566e-05,
      "loss": 0.5663,
      "step": 7710
    },
    {
      "epoch": 2.0279,
      "grad_norm": 9.172537803649902,
      "learning_rate": 1.156060606060606e-05,
      "loss": 1.0224,
      "step": 7711
    },
    {
      "epoch": 2.028,
      "grad_norm": 5.82987117767334,
      "learning_rate": 1.1555555555555556e-05,
      "loss": 0.4384,
      "step": 7712
    },
    {
      "epoch": 2.0281,
      "grad_norm": 5.468423843383789,
      "learning_rate": 1.155050505050505e-05,
      "loss": 0.4188,
      "step": 7713
    },
    {
      "epoch": 2.0282,
      "grad_norm": 6.212398052215576,
      "learning_rate": 1.1545454545454545e-05,
      "loss": 0.4163,
      "step": 7714
    },
    {
      "epoch": 2.0283,
      "grad_norm": 4.629080772399902,
      "learning_rate": 1.154040404040404e-05,
      "loss": 0.5718,
      "step": 7715
    },
    {
      "epoch": 2.0284,
      "grad_norm": 7.516354084014893,
      "learning_rate": 1.1535353535353535e-05,
      "loss": 0.5992,
      "step": 7716
    },
    {
      "epoch": 2.0285,
      "grad_norm": 6.087996006011963,
      "learning_rate": 1.153030303030303e-05,
      "loss": 0.5759,
      "step": 7717
    },
    {
      "epoch": 2.0286,
      "grad_norm": 19.69278335571289,
      "learning_rate": 1.1525252525252527e-05,
      "loss": 0.7981,
      "step": 7718
    },
    {
      "epoch": 2.0287,
      "grad_norm": 6.1889472007751465,
      "learning_rate": 1.1520202020202021e-05,
      "loss": 0.4605,
      "step": 7719
    },
    {
      "epoch": 2.0288,
      "grad_norm": 6.997664451599121,
      "learning_rate": 1.1515151515151517e-05,
      "loss": 0.7532,
      "step": 7720
    },
    {
      "epoch": 2.0289,
      "grad_norm": 4.660118579864502,
      "learning_rate": 1.1510101010101011e-05,
      "loss": 0.6534,
      "step": 7721
    },
    {
      "epoch": 2.029,
      "grad_norm": 6.397185802459717,
      "learning_rate": 1.1505050505050506e-05,
      "loss": 0.3344,
      "step": 7722
    },
    {
      "epoch": 2.0291,
      "grad_norm": 27.173311233520508,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 1.4507,
      "step": 7723
    },
    {
      "epoch": 2.0292,
      "grad_norm": 6.275936603546143,
      "learning_rate": 1.1494949494949496e-05,
      "loss": 0.477,
      "step": 7724
    },
    {
      "epoch": 2.0293,
      "grad_norm": 7.152045726776123,
      "learning_rate": 1.148989898989899e-05,
      "loss": 1.3851,
      "step": 7725
    },
    {
      "epoch": 2.0294,
      "grad_norm": 4.9154253005981445,
      "learning_rate": 1.1484848484848486e-05,
      "loss": 0.9559,
      "step": 7726
    },
    {
      "epoch": 2.0295,
      "grad_norm": 7.162940979003906,
      "learning_rate": 1.147979797979798e-05,
      "loss": 0.3396,
      "step": 7727
    },
    {
      "epoch": 2.0296,
      "grad_norm": 27.805086135864258,
      "learning_rate": 1.1474747474747475e-05,
      "loss": 0.5173,
      "step": 7728
    },
    {
      "epoch": 2.0297,
      "grad_norm": 5.734896183013916,
      "learning_rate": 1.1469696969696971e-05,
      "loss": 0.2721,
      "step": 7729
    },
    {
      "epoch": 2.0298,
      "grad_norm": 3.6618263721466064,
      "learning_rate": 1.1464646464646465e-05,
      "loss": 0.2666,
      "step": 7730
    },
    {
      "epoch": 2.0299,
      "grad_norm": 102.58828735351562,
      "learning_rate": 1.1459595959595961e-05,
      "loss": 1.3182,
      "step": 7731
    },
    {
      "epoch": 2.03,
      "grad_norm": 5.5092997550964355,
      "learning_rate": 1.1454545454545455e-05,
      "loss": 0.2582,
      "step": 7732
    },
    {
      "epoch": 2.0301,
      "grad_norm": 7.280786037445068,
      "learning_rate": 1.144949494949495e-05,
      "loss": 0.4742,
      "step": 7733
    },
    {
      "epoch": 2.0302,
      "grad_norm": 9.048171997070312,
      "learning_rate": 1.1444444444444446e-05,
      "loss": 0.7907,
      "step": 7734
    },
    {
      "epoch": 2.0303,
      "grad_norm": 28.765331268310547,
      "learning_rate": 1.143939393939394e-05,
      "loss": 1.0781,
      "step": 7735
    },
    {
      "epoch": 2.0304,
      "grad_norm": 8.176796913146973,
      "learning_rate": 1.1434343434343434e-05,
      "loss": 1.0774,
      "step": 7736
    },
    {
      "epoch": 2.0305,
      "grad_norm": 6.640288352966309,
      "learning_rate": 1.142929292929293e-05,
      "loss": 0.3765,
      "step": 7737
    },
    {
      "epoch": 2.0306,
      "grad_norm": 28.925159454345703,
      "learning_rate": 1.1424242424242425e-05,
      "loss": 0.6607,
      "step": 7738
    },
    {
      "epoch": 2.0307,
      "grad_norm": 9.677739143371582,
      "learning_rate": 1.1419191919191919e-05,
      "loss": 0.4385,
      "step": 7739
    },
    {
      "epoch": 2.0308,
      "grad_norm": 6.510425567626953,
      "learning_rate": 1.1414141414141415e-05,
      "loss": 0.8784,
      "step": 7740
    },
    {
      "epoch": 2.0309,
      "grad_norm": 40.918060302734375,
      "learning_rate": 1.140909090909091e-05,
      "loss": 0.315,
      "step": 7741
    },
    {
      "epoch": 2.031,
      "grad_norm": 9.460686683654785,
      "learning_rate": 1.1404040404040405e-05,
      "loss": 0.6437,
      "step": 7742
    },
    {
      "epoch": 2.0311,
      "grad_norm": 6.984875202178955,
      "learning_rate": 1.13989898989899e-05,
      "loss": 0.7549,
      "step": 7743
    },
    {
      "epoch": 2.0312,
      "grad_norm": 6.316888809204102,
      "learning_rate": 1.1393939393939394e-05,
      "loss": 0.8292,
      "step": 7744
    },
    {
      "epoch": 2.0313,
      "grad_norm": 6.76798152923584,
      "learning_rate": 1.138888888888889e-05,
      "loss": 0.5949,
      "step": 7745
    },
    {
      "epoch": 2.0314,
      "grad_norm": 3.1531553268432617,
      "learning_rate": 1.1383838383838384e-05,
      "loss": 0.3917,
      "step": 7746
    },
    {
      "epoch": 2.0315,
      "grad_norm": 8.33465576171875,
      "learning_rate": 1.1378787878787878e-05,
      "loss": 0.8754,
      "step": 7747
    },
    {
      "epoch": 2.0316,
      "grad_norm": 2.9960265159606934,
      "learning_rate": 1.1373737373737374e-05,
      "loss": 0.1628,
      "step": 7748
    },
    {
      "epoch": 2.0317,
      "grad_norm": 7.384845733642578,
      "learning_rate": 1.1368686868686869e-05,
      "loss": 0.5988,
      "step": 7749
    },
    {
      "epoch": 2.0318,
      "grad_norm": 6.016829013824463,
      "learning_rate": 1.1363636363636365e-05,
      "loss": 0.353,
      "step": 7750
    },
    {
      "epoch": 2.0319,
      "grad_norm": 8.803135871887207,
      "learning_rate": 1.1358585858585859e-05,
      "loss": 0.341,
      "step": 7751
    },
    {
      "epoch": 2.032,
      "grad_norm": 6.508093357086182,
      "learning_rate": 1.1353535353535353e-05,
      "loss": 0.2154,
      "step": 7752
    },
    {
      "epoch": 2.0321,
      "grad_norm": 5.797252655029297,
      "learning_rate": 1.134848484848485e-05,
      "loss": 0.783,
      "step": 7753
    },
    {
      "epoch": 2.0322,
      "grad_norm": 5.495855808258057,
      "learning_rate": 1.1343434343434344e-05,
      "loss": 0.4773,
      "step": 7754
    },
    {
      "epoch": 2.0323,
      "grad_norm": 8.185564041137695,
      "learning_rate": 1.1338383838383838e-05,
      "loss": 0.6767,
      "step": 7755
    },
    {
      "epoch": 2.0324,
      "grad_norm": 2.238506317138672,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 0.0971,
      "step": 7756
    },
    {
      "epoch": 2.0325,
      "grad_norm": 4.313523292541504,
      "learning_rate": 1.1328282828282828e-05,
      "loss": 0.4938,
      "step": 7757
    },
    {
      "epoch": 2.0326,
      "grad_norm": 10.71692943572998,
      "learning_rate": 1.1323232323232322e-05,
      "loss": 0.393,
      "step": 7758
    },
    {
      "epoch": 2.0327,
      "grad_norm": 7.774270534515381,
      "learning_rate": 1.1318181818181818e-05,
      "loss": 0.4457,
      "step": 7759
    },
    {
      "epoch": 2.0328,
      "grad_norm": 8.79232120513916,
      "learning_rate": 1.1313131313131314e-05,
      "loss": 0.1755,
      "step": 7760
    },
    {
      "epoch": 2.0329,
      "grad_norm": 7.876147270202637,
      "learning_rate": 1.1308080808080809e-05,
      "loss": 0.8126,
      "step": 7761
    },
    {
      "epoch": 2.033,
      "grad_norm": 12.9800386428833,
      "learning_rate": 1.1303030303030305e-05,
      "loss": 0.3785,
      "step": 7762
    },
    {
      "epoch": 2.0331,
      "grad_norm": 8.632781982421875,
      "learning_rate": 1.1297979797979799e-05,
      "loss": 0.3293,
      "step": 7763
    },
    {
      "epoch": 2.0332,
      "grad_norm": 4.6328349113464355,
      "learning_rate": 1.1292929292929293e-05,
      "loss": 0.272,
      "step": 7764
    },
    {
      "epoch": 2.0333,
      "grad_norm": 6.8980712890625,
      "learning_rate": 1.128787878787879e-05,
      "loss": 0.8108,
      "step": 7765
    },
    {
      "epoch": 2.0334,
      "grad_norm": 10.715271949768066,
      "learning_rate": 1.1282828282828284e-05,
      "loss": 0.7637,
      "step": 7766
    },
    {
      "epoch": 2.0335,
      "grad_norm": 11.939146041870117,
      "learning_rate": 1.127777777777778e-05,
      "loss": 0.6982,
      "step": 7767
    },
    {
      "epoch": 2.0336,
      "grad_norm": 31.45446014404297,
      "learning_rate": 1.1272727272727274e-05,
      "loss": 0.3299,
      "step": 7768
    },
    {
      "epoch": 2.0337,
      "grad_norm": 4.030155658721924,
      "learning_rate": 1.1267676767676768e-05,
      "loss": 0.1258,
      "step": 7769
    },
    {
      "epoch": 2.0338,
      "grad_norm": 5.793026924133301,
      "learning_rate": 1.1262626262626264e-05,
      "loss": 0.3458,
      "step": 7770
    },
    {
      "epoch": 2.0339,
      "grad_norm": 8.813907623291016,
      "learning_rate": 1.1257575757575758e-05,
      "loss": 0.6104,
      "step": 7771
    },
    {
      "epoch": 2.034,
      "grad_norm": 4.926126003265381,
      "learning_rate": 1.1252525252525253e-05,
      "loss": 0.3955,
      "step": 7772
    },
    {
      "epoch": 2.0341,
      "grad_norm": 4.672784328460693,
      "learning_rate": 1.1247474747474749e-05,
      "loss": 0.3043,
      "step": 7773
    },
    {
      "epoch": 2.0342,
      "grad_norm": 3.410083293914795,
      "learning_rate": 1.1242424242424243e-05,
      "loss": 0.0876,
      "step": 7774
    },
    {
      "epoch": 2.0343,
      "grad_norm": 6.670752048492432,
      "learning_rate": 1.1237373737373737e-05,
      "loss": 0.6372,
      "step": 7775
    },
    {
      "epoch": 2.0344,
      "grad_norm": 12.83764934539795,
      "learning_rate": 1.1232323232323233e-05,
      "loss": 0.6526,
      "step": 7776
    },
    {
      "epoch": 2.0345,
      "grad_norm": 8.06689167022705,
      "learning_rate": 1.1227272727272728e-05,
      "loss": 0.8346,
      "step": 7777
    },
    {
      "epoch": 2.0346,
      "grad_norm": 6.512429714202881,
      "learning_rate": 1.1222222222222224e-05,
      "loss": 1.0729,
      "step": 7778
    },
    {
      "epoch": 2.0347,
      "grad_norm": 7.233400821685791,
      "learning_rate": 1.1217171717171718e-05,
      "loss": 0.4122,
      "step": 7779
    },
    {
      "epoch": 2.0348,
      "grad_norm": 7.020813941955566,
      "learning_rate": 1.1212121212121212e-05,
      "loss": 0.8358,
      "step": 7780
    },
    {
      "epoch": 2.0349,
      "grad_norm": 6.813151836395264,
      "learning_rate": 1.1207070707070708e-05,
      "loss": 0.8933,
      "step": 7781
    },
    {
      "epoch": 2.035,
      "grad_norm": 6.145717620849609,
      "learning_rate": 1.1202020202020202e-05,
      "loss": 0.357,
      "step": 7782
    },
    {
      "epoch": 2.0351,
      "grad_norm": 7.799468040466309,
      "learning_rate": 1.1196969696969697e-05,
      "loss": 0.481,
      "step": 7783
    },
    {
      "epoch": 2.0352,
      "grad_norm": 11.74467945098877,
      "learning_rate": 1.1191919191919193e-05,
      "loss": 0.7798,
      "step": 7784
    },
    {
      "epoch": 2.0353,
      "grad_norm": 4.375615119934082,
      "learning_rate": 1.1186868686868687e-05,
      "loss": 0.3265,
      "step": 7785
    },
    {
      "epoch": 2.0354,
      "grad_norm": 5.132450103759766,
      "learning_rate": 1.1181818181818183e-05,
      "loss": 0.4862,
      "step": 7786
    },
    {
      "epoch": 2.0355,
      "grad_norm": 6.310489654541016,
      "learning_rate": 1.1176767676767677e-05,
      "loss": 0.4476,
      "step": 7787
    },
    {
      "epoch": 2.0356,
      "grad_norm": 10.133865356445312,
      "learning_rate": 1.1171717171717172e-05,
      "loss": 0.5679,
      "step": 7788
    },
    {
      "epoch": 2.0357,
      "grad_norm": 4.869781970977783,
      "learning_rate": 1.1166666666666668e-05,
      "loss": 0.183,
      "step": 7789
    },
    {
      "epoch": 2.0358,
      "grad_norm": 5.595122814178467,
      "learning_rate": 1.1161616161616162e-05,
      "loss": 0.7227,
      "step": 7790
    },
    {
      "epoch": 2.0359,
      "grad_norm": 4.218642711639404,
      "learning_rate": 1.1156565656565656e-05,
      "loss": 0.4321,
      "step": 7791
    },
    {
      "epoch": 2.036,
      "grad_norm": 4.801213264465332,
      "learning_rate": 1.1151515151515152e-05,
      "loss": 0.3653,
      "step": 7792
    },
    {
      "epoch": 2.0361,
      "grad_norm": 8.040764808654785,
      "learning_rate": 1.1146464646464646e-05,
      "loss": 0.4764,
      "step": 7793
    },
    {
      "epoch": 2.0362,
      "grad_norm": 6.866008758544922,
      "learning_rate": 1.114141414141414e-05,
      "loss": 1.009,
      "step": 7794
    },
    {
      "epoch": 2.0362999999999998,
      "grad_norm": 6.0579657554626465,
      "learning_rate": 1.1136363636363637e-05,
      "loss": 0.2639,
      "step": 7795
    },
    {
      "epoch": 2.0364,
      "grad_norm": 6.790030479431152,
      "learning_rate": 1.1131313131313131e-05,
      "loss": 0.5072,
      "step": 7796
    },
    {
      "epoch": 2.0365,
      "grad_norm": 6.755161762237549,
      "learning_rate": 1.1126262626262627e-05,
      "loss": 0.5367,
      "step": 7797
    },
    {
      "epoch": 2.0366,
      "grad_norm": 7.031276226043701,
      "learning_rate": 1.1121212121212121e-05,
      "loss": 0.4935,
      "step": 7798
    },
    {
      "epoch": 2.0367,
      "grad_norm": 7.493284702301025,
      "learning_rate": 1.1116161616161616e-05,
      "loss": 0.425,
      "step": 7799
    },
    {
      "epoch": 2.0368,
      "grad_norm": 8.14472770690918,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.5618,
      "step": 7800
    },
    {
      "epoch": 2.0369,
      "grad_norm": 5.09262752532959,
      "learning_rate": 1.1106060606060606e-05,
      "loss": 0.212,
      "step": 7801
    },
    {
      "epoch": 2.037,
      "grad_norm": 5.585345268249512,
      "learning_rate": 1.11010101010101e-05,
      "loss": 0.3068,
      "step": 7802
    },
    {
      "epoch": 2.0371,
      "grad_norm": 4.618856906890869,
      "learning_rate": 1.1095959595959598e-05,
      "loss": 0.6643,
      "step": 7803
    },
    {
      "epoch": 2.0372,
      "grad_norm": 29.166860580444336,
      "learning_rate": 1.1090909090909092e-05,
      "loss": 0.4881,
      "step": 7804
    },
    {
      "epoch": 2.0373,
      "grad_norm": 7.299139022827148,
      "learning_rate": 1.1085858585858587e-05,
      "loss": 0.6122,
      "step": 7805
    },
    {
      "epoch": 2.0374,
      "grad_norm": 9.996185302734375,
      "learning_rate": 1.1080808080808083e-05,
      "loss": 0.9351,
      "step": 7806
    },
    {
      "epoch": 2.0375,
      "grad_norm": 6.597074031829834,
      "learning_rate": 1.1075757575757577e-05,
      "loss": 0.6889,
      "step": 7807
    },
    {
      "epoch": 2.0376,
      "grad_norm": 7.322464942932129,
      "learning_rate": 1.1070707070707071e-05,
      "loss": 1.1914,
      "step": 7808
    },
    {
      "epoch": 2.0377,
      "grad_norm": 8.281582832336426,
      "learning_rate": 1.1065656565656567e-05,
      "loss": 0.7994,
      "step": 7809
    },
    {
      "epoch": 2.0378,
      "grad_norm": 6.611993789672852,
      "learning_rate": 1.1060606060606061e-05,
      "loss": 0.6745,
      "step": 7810
    },
    {
      "epoch": 2.0379,
      "grad_norm": 8.538578033447266,
      "learning_rate": 1.1055555555555556e-05,
      "loss": 0.5474,
      "step": 7811
    },
    {
      "epoch": 2.038,
      "grad_norm": 11.722183227539062,
      "learning_rate": 1.1050505050505052e-05,
      "loss": 0.2412,
      "step": 7812
    },
    {
      "epoch": 2.0381,
      "grad_norm": 4.474798679351807,
      "learning_rate": 1.1045454545454546e-05,
      "loss": 0.196,
      "step": 7813
    },
    {
      "epoch": 2.0382,
      "grad_norm": 6.103868007659912,
      "learning_rate": 1.1040404040404042e-05,
      "loss": 0.8077,
      "step": 7814
    },
    {
      "epoch": 2.0383,
      "grad_norm": 4.421363830566406,
      "learning_rate": 1.1035353535353536e-05,
      "loss": 0.328,
      "step": 7815
    },
    {
      "epoch": 2.0384,
      "grad_norm": 3.278323173522949,
      "learning_rate": 1.103030303030303e-05,
      "loss": 0.1242,
      "step": 7816
    },
    {
      "epoch": 2.0385,
      "grad_norm": 7.492076396942139,
      "learning_rate": 1.1025252525252527e-05,
      "loss": 0.3264,
      "step": 7817
    },
    {
      "epoch": 2.0386,
      "grad_norm": 5.510987281799316,
      "learning_rate": 1.102020202020202e-05,
      "loss": 0.7897,
      "step": 7818
    },
    {
      "epoch": 2.0387,
      "grad_norm": 2.363109588623047,
      "learning_rate": 1.1015151515151515e-05,
      "loss": 0.2576,
      "step": 7819
    },
    {
      "epoch": 2.0388,
      "grad_norm": 9.142346382141113,
      "learning_rate": 1.1010101010101011e-05,
      "loss": 0.4688,
      "step": 7820
    },
    {
      "epoch": 2.0389,
      "grad_norm": 16.408695220947266,
      "learning_rate": 1.1005050505050505e-05,
      "loss": 0.3418,
      "step": 7821
    },
    {
      "epoch": 2.039,
      "grad_norm": 4.596287250518799,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.2473,
      "step": 7822
    },
    {
      "epoch": 2.0391,
      "grad_norm": 6.473719596862793,
      "learning_rate": 1.0994949494949496e-05,
      "loss": 0.2259,
      "step": 7823
    },
    {
      "epoch": 2.0392,
      "grad_norm": 5.357305526733398,
      "learning_rate": 1.098989898989899e-05,
      "loss": 0.6869,
      "step": 7824
    },
    {
      "epoch": 2.0393,
      "grad_norm": 6.036100387573242,
      "learning_rate": 1.0984848484848486e-05,
      "loss": 0.6883,
      "step": 7825
    },
    {
      "epoch": 2.0394,
      "grad_norm": 9.089659690856934,
      "learning_rate": 1.097979797979798e-05,
      "loss": 0.3304,
      "step": 7826
    },
    {
      "epoch": 2.0395,
      "grad_norm": 7.605237007141113,
      "learning_rate": 1.0974747474747475e-05,
      "loss": 0.953,
      "step": 7827
    },
    {
      "epoch": 2.0396,
      "grad_norm": 8.976008415222168,
      "learning_rate": 1.096969696969697e-05,
      "loss": 0.5746,
      "step": 7828
    },
    {
      "epoch": 2.0397,
      "grad_norm": 5.572580337524414,
      "learning_rate": 1.0964646464646465e-05,
      "loss": 0.9262,
      "step": 7829
    },
    {
      "epoch": 2.0398,
      "grad_norm": 4.248342514038086,
      "learning_rate": 1.095959595959596e-05,
      "loss": 0.2637,
      "step": 7830
    },
    {
      "epoch": 2.0399,
      "grad_norm": 7.725618839263916,
      "learning_rate": 1.0954545454545455e-05,
      "loss": 0.8479,
      "step": 7831
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.751453876495361,
      "learning_rate": 1.094949494949495e-05,
      "loss": 0.5887,
      "step": 7832
    },
    {
      "epoch": 2.0401,
      "grad_norm": 4.009005069732666,
      "learning_rate": 1.0944444444444445e-05,
      "loss": 0.3279,
      "step": 7833
    },
    {
      "epoch": 2.0402,
      "grad_norm": 7.127386093139648,
      "learning_rate": 1.093939393939394e-05,
      "loss": 0.9842,
      "step": 7834
    },
    {
      "epoch": 2.0403000000000002,
      "grad_norm": 4.3324737548828125,
      "learning_rate": 1.0934343434343434e-05,
      "loss": 0.1997,
      "step": 7835
    },
    {
      "epoch": 2.0404,
      "grad_norm": 11.034652709960938,
      "learning_rate": 1.092929292929293e-05,
      "loss": 0.8435,
      "step": 7836
    },
    {
      "epoch": 2.0405,
      "grad_norm": 4.850997447967529,
      "learning_rate": 1.0924242424242424e-05,
      "loss": 0.4739,
      "step": 7837
    },
    {
      "epoch": 2.0406,
      "grad_norm": 5.38137674331665,
      "learning_rate": 1.0919191919191919e-05,
      "loss": 0.8179,
      "step": 7838
    },
    {
      "epoch": 2.0407,
      "grad_norm": 9.482253074645996,
      "learning_rate": 1.0914141414141415e-05,
      "loss": 0.6707,
      "step": 7839
    },
    {
      "epoch": 2.0408,
      "grad_norm": 10.439462661743164,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 2.1101,
      "step": 7840
    },
    {
      "epoch": 2.0409,
      "grad_norm": 20.1929931640625,
      "learning_rate": 1.0904040404040403e-05,
      "loss": 0.3504,
      "step": 7841
    },
    {
      "epoch": 2.041,
      "grad_norm": 33.770320892333984,
      "learning_rate": 1.08989898989899e-05,
      "loss": 0.9415,
      "step": 7842
    },
    {
      "epoch": 2.0411,
      "grad_norm": 12.646649360656738,
      "learning_rate": 1.0893939393939393e-05,
      "loss": 0.7583,
      "step": 7843
    },
    {
      "epoch": 2.0412,
      "grad_norm": 5.840404987335205,
      "learning_rate": 1.088888888888889e-05,
      "loss": 0.5404,
      "step": 7844
    },
    {
      "epoch": 2.0413,
      "grad_norm": 9.533632278442383,
      "learning_rate": 1.0883838383838384e-05,
      "loss": 0.5755,
      "step": 7845
    },
    {
      "epoch": 2.0414,
      "grad_norm": 18.61954116821289,
      "learning_rate": 1.087878787878788e-05,
      "loss": 1.2317,
      "step": 7846
    },
    {
      "epoch": 2.0415,
      "grad_norm": 13.921025276184082,
      "learning_rate": 1.0873737373737374e-05,
      "loss": 1.1761,
      "step": 7847
    },
    {
      "epoch": 2.0416,
      "grad_norm": 5.512228965759277,
      "learning_rate": 1.086868686868687e-05,
      "loss": 0.2107,
      "step": 7848
    },
    {
      "epoch": 2.0417,
      "grad_norm": 4.567892551422119,
      "learning_rate": 1.0863636363636364e-05,
      "loss": 0.2857,
      "step": 7849
    },
    {
      "epoch": 2.0418,
      "grad_norm": 4.371391296386719,
      "learning_rate": 1.085858585858586e-05,
      "loss": 0.7048,
      "step": 7850
    },
    {
      "epoch": 2.0419,
      "grad_norm": 6.899272918701172,
      "learning_rate": 1.0853535353535355e-05,
      "loss": 0.7217,
      "step": 7851
    },
    {
      "epoch": 2.042,
      "grad_norm": 8.392151832580566,
      "learning_rate": 1.0848484848484849e-05,
      "loss": 0.7299,
      "step": 7852
    },
    {
      "epoch": 2.0421,
      "grad_norm": 7.311099529266357,
      "learning_rate": 1.0843434343434345e-05,
      "loss": 0.4802,
      "step": 7853
    },
    {
      "epoch": 2.0422,
      "grad_norm": 3.3827850818634033,
      "learning_rate": 1.083838383838384e-05,
      "loss": 0.3491,
      "step": 7854
    },
    {
      "epoch": 2.0423,
      "grad_norm": 2.9980294704437256,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.252,
      "step": 7855
    },
    {
      "epoch": 2.0424,
      "grad_norm": 7.858309268951416,
      "learning_rate": 1.082828282828283e-05,
      "loss": 0.4301,
      "step": 7856
    },
    {
      "epoch": 2.0425,
      "grad_norm": 8.231795310974121,
      "learning_rate": 1.0823232323232324e-05,
      "loss": 1.1592,
      "step": 7857
    },
    {
      "epoch": 2.0426,
      "grad_norm": 6.006837844848633,
      "learning_rate": 1.081818181818182e-05,
      "loss": 0.7122,
      "step": 7858
    },
    {
      "epoch": 2.0427,
      "grad_norm": 7.088186264038086,
      "learning_rate": 1.0813131313131314e-05,
      "loss": 1.2134,
      "step": 7859
    },
    {
      "epoch": 2.0428,
      "grad_norm": 8.653401374816895,
      "learning_rate": 1.0808080808080808e-05,
      "loss": 0.4973,
      "step": 7860
    },
    {
      "epoch": 2.0429,
      "grad_norm": 4.808138370513916,
      "learning_rate": 1.0803030303030304e-05,
      "loss": 0.5199,
      "step": 7861
    },
    {
      "epoch": 2.043,
      "grad_norm": 9.986310005187988,
      "learning_rate": 1.0797979797979799e-05,
      "loss": 0.7355,
      "step": 7862
    },
    {
      "epoch": 2.0431,
      "grad_norm": 7.016462802886963,
      "learning_rate": 1.0792929292929293e-05,
      "loss": 0.6185,
      "step": 7863
    },
    {
      "epoch": 2.0432,
      "grad_norm": 8.168205261230469,
      "learning_rate": 1.0787878787878789e-05,
      "loss": 0.5945,
      "step": 7864
    },
    {
      "epoch": 2.0433,
      "grad_norm": 10.672067642211914,
      "learning_rate": 1.0782828282828283e-05,
      "loss": 0.2651,
      "step": 7865
    },
    {
      "epoch": 2.0434,
      "grad_norm": 5.471961498260498,
      "learning_rate": 1.0777777777777778e-05,
      "loss": 0.3068,
      "step": 7866
    },
    {
      "epoch": 2.0435,
      "grad_norm": 5.587368965148926,
      "learning_rate": 1.0772727272727274e-05,
      "loss": 0.347,
      "step": 7867
    },
    {
      "epoch": 2.0436,
      "grad_norm": 13.90919017791748,
      "learning_rate": 1.0767676767676768e-05,
      "loss": 0.2306,
      "step": 7868
    },
    {
      "epoch": 2.0437,
      "grad_norm": 8.172571182250977,
      "learning_rate": 1.0762626262626264e-05,
      "loss": 0.3436,
      "step": 7869
    },
    {
      "epoch": 2.0438,
      "grad_norm": 6.439176559448242,
      "learning_rate": 1.0757575757575758e-05,
      "loss": 0.2666,
      "step": 7870
    },
    {
      "epoch": 2.0439,
      "grad_norm": 6.196496963500977,
      "learning_rate": 1.0752525252525252e-05,
      "loss": 0.2532,
      "step": 7871
    },
    {
      "epoch": 2.044,
      "grad_norm": 7.563130855560303,
      "learning_rate": 1.0747474747474748e-05,
      "loss": 0.2669,
      "step": 7872
    },
    {
      "epoch": 2.0441,
      "grad_norm": 4.200045585632324,
      "learning_rate": 1.0742424242424243e-05,
      "loss": 1.1025,
      "step": 7873
    },
    {
      "epoch": 2.0442,
      "grad_norm": 6.976571559906006,
      "learning_rate": 1.0737373737373737e-05,
      "loss": 0.4541,
      "step": 7874
    },
    {
      "epoch": 2.0443,
      "grad_norm": 6.779148578643799,
      "learning_rate": 1.0732323232323233e-05,
      "loss": 0.4573,
      "step": 7875
    },
    {
      "epoch": 2.0444,
      "grad_norm": 7.632964611053467,
      "learning_rate": 1.0727272727272727e-05,
      "loss": 0.5951,
      "step": 7876
    },
    {
      "epoch": 2.0445,
      "grad_norm": 6.524518966674805,
      "learning_rate": 1.0722222222222222e-05,
      "loss": 0.3001,
      "step": 7877
    },
    {
      "epoch": 2.0446,
      "grad_norm": 8.337782859802246,
      "learning_rate": 1.0717171717171718e-05,
      "loss": 0.2709,
      "step": 7878
    },
    {
      "epoch": 2.0447,
      "grad_norm": 16.341873168945312,
      "learning_rate": 1.0712121212121212e-05,
      "loss": 0.2798,
      "step": 7879
    },
    {
      "epoch": 2.0448,
      "grad_norm": 7.53145694732666,
      "learning_rate": 1.0707070707070708e-05,
      "loss": 1.0081,
      "step": 7880
    },
    {
      "epoch": 2.0449,
      "grad_norm": 5.146393775939941,
      "learning_rate": 1.0702020202020202e-05,
      "loss": 0.2751,
      "step": 7881
    },
    {
      "epoch": 2.045,
      "grad_norm": 4.480648994445801,
      "learning_rate": 1.0696969696969696e-05,
      "loss": 0.5591,
      "step": 7882
    },
    {
      "epoch": 2.0451,
      "grad_norm": 14.778848648071289,
      "learning_rate": 1.0691919191919192e-05,
      "loss": 0.446,
      "step": 7883
    },
    {
      "epoch": 2.0452,
      "grad_norm": 7.842984199523926,
      "learning_rate": 1.0686868686868687e-05,
      "loss": 0.4415,
      "step": 7884
    },
    {
      "epoch": 2.0453,
      "grad_norm": 10.423826217651367,
      "learning_rate": 1.0681818181818181e-05,
      "loss": 0.168,
      "step": 7885
    },
    {
      "epoch": 2.0454,
      "grad_norm": 11.154095649719238,
      "learning_rate": 1.0676767676767677e-05,
      "loss": 0.6643,
      "step": 7886
    },
    {
      "epoch": 2.0455,
      "grad_norm": 5.979033470153809,
      "learning_rate": 1.0671717171717171e-05,
      "loss": 0.5931,
      "step": 7887
    },
    {
      "epoch": 2.0456,
      "grad_norm": 5.757279396057129,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.3818,
      "step": 7888
    },
    {
      "epoch": 2.0457,
      "grad_norm": 32.9141960144043,
      "learning_rate": 1.0661616161616163e-05,
      "loss": 0.9868,
      "step": 7889
    },
    {
      "epoch": 2.0458,
      "grad_norm": 5.801530361175537,
      "learning_rate": 1.0656565656565658e-05,
      "loss": 0.4614,
      "step": 7890
    },
    {
      "epoch": 2.0459,
      "grad_norm": 5.838603973388672,
      "learning_rate": 1.0651515151515152e-05,
      "loss": 0.2965,
      "step": 7891
    },
    {
      "epoch": 2.046,
      "grad_norm": 6.705530643463135,
      "learning_rate": 1.0646464646464648e-05,
      "loss": 0.4128,
      "step": 7892
    },
    {
      "epoch": 2.0461,
      "grad_norm": 8.610803604125977,
      "learning_rate": 1.0641414141414142e-05,
      "loss": 0.4621,
      "step": 7893
    },
    {
      "epoch": 2.0462,
      "grad_norm": 8.429837226867676,
      "learning_rate": 1.0636363636363638e-05,
      "loss": 0.7742,
      "step": 7894
    },
    {
      "epoch": 2.0463,
      "grad_norm": 5.971479892730713,
      "learning_rate": 1.0631313131313132e-05,
      "loss": 0.4567,
      "step": 7895
    },
    {
      "epoch": 2.0464,
      "grad_norm": 5.209444999694824,
      "learning_rate": 1.0626262626262627e-05,
      "loss": 0.4449,
      "step": 7896
    },
    {
      "epoch": 2.0465,
      "grad_norm": 5.739396095275879,
      "learning_rate": 1.0621212121212123e-05,
      "loss": 0.3287,
      "step": 7897
    },
    {
      "epoch": 2.0466,
      "grad_norm": 4.417747974395752,
      "learning_rate": 1.0616161616161617e-05,
      "loss": 0.3256,
      "step": 7898
    },
    {
      "epoch": 2.0467,
      "grad_norm": 5.90683650970459,
      "learning_rate": 1.0611111111111111e-05,
      "loss": 0.3394,
      "step": 7899
    },
    {
      "epoch": 2.0468,
      "grad_norm": 18.250944137573242,
      "learning_rate": 1.0606060606060607e-05,
      "loss": 0.2854,
      "step": 7900
    },
    {
      "epoch": 2.0469,
      "grad_norm": 4.958743095397949,
      "learning_rate": 1.0601010101010102e-05,
      "loss": 0.9119,
      "step": 7901
    },
    {
      "epoch": 2.047,
      "grad_norm": 4.915347099304199,
      "learning_rate": 1.0595959595959596e-05,
      "loss": 0.1682,
      "step": 7902
    },
    {
      "epoch": 2.0471,
      "grad_norm": 10.476213455200195,
      "learning_rate": 1.0590909090909092e-05,
      "loss": 0.7381,
      "step": 7903
    },
    {
      "epoch": 2.0472,
      "grad_norm": 6.557709693908691,
      "learning_rate": 1.0585858585858586e-05,
      "loss": 0.2841,
      "step": 7904
    },
    {
      "epoch": 2.0473,
      "grad_norm": 6.2283244132995605,
      "learning_rate": 1.0580808080808082e-05,
      "loss": 1.0566,
      "step": 7905
    },
    {
      "epoch": 2.0474,
      "grad_norm": 6.035665512084961,
      "learning_rate": 1.0575757575757577e-05,
      "loss": 0.2688,
      "step": 7906
    },
    {
      "epoch": 2.0475,
      "grad_norm": 5.426872253417969,
      "learning_rate": 1.057070707070707e-05,
      "loss": 0.4274,
      "step": 7907
    },
    {
      "epoch": 2.0476,
      "grad_norm": 13.848220825195312,
      "learning_rate": 1.0565656565656567e-05,
      "loss": 0.2792,
      "step": 7908
    },
    {
      "epoch": 2.0477,
      "grad_norm": 5.725285053253174,
      "learning_rate": 1.0560606060606061e-05,
      "loss": 0.4518,
      "step": 7909
    },
    {
      "epoch": 2.0478,
      "grad_norm": 6.901704788208008,
      "learning_rate": 1.0555555555555555e-05,
      "loss": 0.3355,
      "step": 7910
    },
    {
      "epoch": 2.0479,
      "grad_norm": 56.055747985839844,
      "learning_rate": 1.0550505050505051e-05,
      "loss": 0.2707,
      "step": 7911
    },
    {
      "epoch": 2.048,
      "grad_norm": 7.4574408531188965,
      "learning_rate": 1.0545454545454546e-05,
      "loss": 0.37,
      "step": 7912
    },
    {
      "epoch": 2.0481,
      "grad_norm": 7.2031049728393555,
      "learning_rate": 1.054040404040404e-05,
      "loss": 0.3759,
      "step": 7913
    },
    {
      "epoch": 2.0482,
      "grad_norm": 6.247785568237305,
      "learning_rate": 1.0535353535353536e-05,
      "loss": 0.3571,
      "step": 7914
    },
    {
      "epoch": 2.0483,
      "grad_norm": 6.0669474601745605,
      "learning_rate": 1.053030303030303e-05,
      "loss": 0.7814,
      "step": 7915
    },
    {
      "epoch": 2.0484,
      "grad_norm": 6.799445152282715,
      "learning_rate": 1.0525252525252526e-05,
      "loss": 0.3926,
      "step": 7916
    },
    {
      "epoch": 2.0485,
      "grad_norm": 3.5960168838500977,
      "learning_rate": 1.052020202020202e-05,
      "loss": 0.5582,
      "step": 7917
    },
    {
      "epoch": 2.0486,
      "grad_norm": 5.115959644317627,
      "learning_rate": 1.0515151515151515e-05,
      "loss": 0.3213,
      "step": 7918
    },
    {
      "epoch": 2.0487,
      "grad_norm": 7.876897811889648,
      "learning_rate": 1.051010101010101e-05,
      "loss": 0.367,
      "step": 7919
    },
    {
      "epoch": 2.0488,
      "grad_norm": 8.18124771118164,
      "learning_rate": 1.0505050505050505e-05,
      "loss": 0.3588,
      "step": 7920
    },
    {
      "epoch": 2.0489,
      "grad_norm": 11.905903816223145,
      "learning_rate": 1.05e-05,
      "loss": 0.5739,
      "step": 7921
    },
    {
      "epoch": 2.049,
      "grad_norm": 7.279657363891602,
      "learning_rate": 1.0494949494949495e-05,
      "loss": 0.2969,
      "step": 7922
    },
    {
      "epoch": 2.0491,
      "grad_norm": 8.160224914550781,
      "learning_rate": 1.048989898989899e-05,
      "loss": 0.3642,
      "step": 7923
    },
    {
      "epoch": 2.0492,
      "grad_norm": 10.343058586120605,
      "learning_rate": 1.0484848484848486e-05,
      "loss": 0.733,
      "step": 7924
    },
    {
      "epoch": 2.0493,
      "grad_norm": 15.47484302520752,
      "learning_rate": 1.047979797979798e-05,
      "loss": 1.2172,
      "step": 7925
    },
    {
      "epoch": 2.0494,
      "grad_norm": 7.28474760055542,
      "learning_rate": 1.0474747474747474e-05,
      "loss": 0.529,
      "step": 7926
    },
    {
      "epoch": 2.0495,
      "grad_norm": 5.721981525421143,
      "learning_rate": 1.046969696969697e-05,
      "loss": 1.2257,
      "step": 7927
    },
    {
      "epoch": 2.0496,
      "grad_norm": 19.968820571899414,
      "learning_rate": 1.0464646464646465e-05,
      "loss": 0.5144,
      "step": 7928
    },
    {
      "epoch": 2.0497,
      "grad_norm": 13.492219924926758,
      "learning_rate": 1.0459595959595959e-05,
      "loss": 1.0046,
      "step": 7929
    },
    {
      "epoch": 2.0498,
      "grad_norm": 6.170838832855225,
      "learning_rate": 1.0454545454545455e-05,
      "loss": 0.871,
      "step": 7930
    },
    {
      "epoch": 2.0499,
      "grad_norm": 6.4061479568481445,
      "learning_rate": 1.0449494949494951e-05,
      "loss": 0.7646,
      "step": 7931
    },
    {
      "epoch": 2.05,
      "grad_norm": 7.17263650894165,
      "learning_rate": 1.0444444444444445e-05,
      "loss": 0.9493,
      "step": 7932
    },
    {
      "epoch": 2.0501,
      "grad_norm": 22.049455642700195,
      "learning_rate": 1.0439393939393941e-05,
      "loss": 0.4344,
      "step": 7933
    },
    {
      "epoch": 2.0502,
      "grad_norm": 8.230277061462402,
      "learning_rate": 1.0434343434343435e-05,
      "loss": 0.4907,
      "step": 7934
    },
    {
      "epoch": 2.0503,
      "grad_norm": 5.898164749145508,
      "learning_rate": 1.042929292929293e-05,
      "loss": 0.534,
      "step": 7935
    },
    {
      "epoch": 2.0504,
      "grad_norm": 8.010835647583008,
      "learning_rate": 1.0424242424242426e-05,
      "loss": 0.4577,
      "step": 7936
    },
    {
      "epoch": 2.0505,
      "grad_norm": 7.355976104736328,
      "learning_rate": 1.041919191919192e-05,
      "loss": 0.3326,
      "step": 7937
    },
    {
      "epoch": 2.0506,
      "grad_norm": 14.834680557250977,
      "learning_rate": 1.0414141414141414e-05,
      "loss": 0.3829,
      "step": 7938
    },
    {
      "epoch": 2.0507,
      "grad_norm": 4.648716926574707,
      "learning_rate": 1.040909090909091e-05,
      "loss": 0.4506,
      "step": 7939
    },
    {
      "epoch": 2.0508,
      "grad_norm": 5.243838787078857,
      "learning_rate": 1.0404040404040405e-05,
      "loss": 0.1687,
      "step": 7940
    },
    {
      "epoch": 2.0509,
      "grad_norm": 6.573238372802734,
      "learning_rate": 1.03989898989899e-05,
      "loss": 0.2103,
      "step": 7941
    },
    {
      "epoch": 2.051,
      "grad_norm": 37.282440185546875,
      "learning_rate": 1.0393939393939395e-05,
      "loss": 0.2171,
      "step": 7942
    },
    {
      "epoch": 2.0511,
      "grad_norm": 6.564283847808838,
      "learning_rate": 1.038888888888889e-05,
      "loss": 0.3941,
      "step": 7943
    },
    {
      "epoch": 2.0512,
      "grad_norm": 8.525093078613281,
      "learning_rate": 1.0383838383838385e-05,
      "loss": 1.19,
      "step": 7944
    },
    {
      "epoch": 2.0513,
      "grad_norm": 7.426860809326172,
      "learning_rate": 1.037878787878788e-05,
      "loss": 0.3955,
      "step": 7945
    },
    {
      "epoch": 2.0514,
      "grad_norm": 8.880045890808105,
      "learning_rate": 1.0373737373737374e-05,
      "loss": 0.5899,
      "step": 7946
    },
    {
      "epoch": 2.0515,
      "grad_norm": 11.409974098205566,
      "learning_rate": 1.036868686868687e-05,
      "loss": 0.5669,
      "step": 7947
    },
    {
      "epoch": 2.0516,
      "grad_norm": 9.674158096313477,
      "learning_rate": 1.0363636363636364e-05,
      "loss": 0.4166,
      "step": 7948
    },
    {
      "epoch": 2.0517,
      "grad_norm": 6.259641170501709,
      "learning_rate": 1.0358585858585858e-05,
      "loss": 0.3172,
      "step": 7949
    },
    {
      "epoch": 2.0518,
      "grad_norm": 10.618340492248535,
      "learning_rate": 1.0353535353535354e-05,
      "loss": 0.5599,
      "step": 7950
    },
    {
      "epoch": 2.0519,
      "grad_norm": 6.666693210601807,
      "learning_rate": 1.0348484848484849e-05,
      "loss": 0.3291,
      "step": 7951
    },
    {
      "epoch": 2.052,
      "grad_norm": 10.092100143432617,
      "learning_rate": 1.0343434343434345e-05,
      "loss": 0.7837,
      "step": 7952
    },
    {
      "epoch": 2.0521,
      "grad_norm": 5.362566947937012,
      "learning_rate": 1.0338383838383839e-05,
      "loss": 0.3451,
      "step": 7953
    },
    {
      "epoch": 2.0522,
      "grad_norm": 6.721336841583252,
      "learning_rate": 1.0333333333333333e-05,
      "loss": 0.3635,
      "step": 7954
    },
    {
      "epoch": 2.0523,
      "grad_norm": 6.869516849517822,
      "learning_rate": 1.032828282828283e-05,
      "loss": 0.426,
      "step": 7955
    },
    {
      "epoch": 2.0524,
      "grad_norm": 18.34858512878418,
      "learning_rate": 1.0323232323232324e-05,
      "loss": 1.9467,
      "step": 7956
    },
    {
      "epoch": 2.0525,
      "grad_norm": 7.715972900390625,
      "learning_rate": 1.0318181818181818e-05,
      "loss": 0.6491,
      "step": 7957
    },
    {
      "epoch": 2.0526,
      "grad_norm": 10.3684663772583,
      "learning_rate": 1.0313131313131314e-05,
      "loss": 1.1798,
      "step": 7958
    },
    {
      "epoch": 2.0527,
      "grad_norm": 6.9799299240112305,
      "learning_rate": 1.0308080808080808e-05,
      "loss": 0.7087,
      "step": 7959
    },
    {
      "epoch": 2.0528,
      "grad_norm": 10.093742370605469,
      "learning_rate": 1.0303030303030304e-05,
      "loss": 0.6736,
      "step": 7960
    },
    {
      "epoch": 2.0529,
      "grad_norm": 10.233283996582031,
      "learning_rate": 1.0297979797979798e-05,
      "loss": 0.3131,
      "step": 7961
    },
    {
      "epoch": 2.053,
      "grad_norm": 5.592309951782227,
      "learning_rate": 1.0292929292929293e-05,
      "loss": 0.8304,
      "step": 7962
    },
    {
      "epoch": 2.0531,
      "grad_norm": 7.574190139770508,
      "learning_rate": 1.0287878787878789e-05,
      "loss": 0.5795,
      "step": 7963
    },
    {
      "epoch": 2.0532,
      "grad_norm": 16.203472137451172,
      "learning_rate": 1.0282828282828283e-05,
      "loss": 0.8988,
      "step": 7964
    },
    {
      "epoch": 2.0533,
      "grad_norm": 4.3921427726745605,
      "learning_rate": 1.0277777777777777e-05,
      "loss": 0.1018,
      "step": 7965
    },
    {
      "epoch": 2.0534,
      "grad_norm": 16.082998275756836,
      "learning_rate": 1.0272727272727273e-05,
      "loss": 0.5467,
      "step": 7966
    },
    {
      "epoch": 2.0535,
      "grad_norm": 7.148820400238037,
      "learning_rate": 1.0267676767676768e-05,
      "loss": 0.469,
      "step": 7967
    },
    {
      "epoch": 2.0536,
      "grad_norm": 6.905325412750244,
      "learning_rate": 1.0262626262626262e-05,
      "loss": 0.2648,
      "step": 7968
    },
    {
      "epoch": 2.0537,
      "grad_norm": 13.806635856628418,
      "learning_rate": 1.0257575757575758e-05,
      "loss": 1.2824,
      "step": 7969
    },
    {
      "epoch": 2.0538,
      "grad_norm": 7.717479228973389,
      "learning_rate": 1.0252525252525252e-05,
      "loss": 0.4956,
      "step": 7970
    },
    {
      "epoch": 2.0539,
      "grad_norm": 6.689619541168213,
      "learning_rate": 1.0247474747474748e-05,
      "loss": 0.434,
      "step": 7971
    },
    {
      "epoch": 2.054,
      "grad_norm": 7.673614501953125,
      "learning_rate": 1.0242424242424242e-05,
      "loss": 0.2672,
      "step": 7972
    },
    {
      "epoch": 2.0541,
      "grad_norm": 9.326793670654297,
      "learning_rate": 1.0237373737373737e-05,
      "loss": 0.9657,
      "step": 7973
    },
    {
      "epoch": 2.0542,
      "grad_norm": 5.695932388305664,
      "learning_rate": 1.0232323232323233e-05,
      "loss": 1.7528,
      "step": 7974
    },
    {
      "epoch": 2.0543,
      "grad_norm": 7.226579189300537,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.4346,
      "step": 7975
    },
    {
      "epoch": 2.0544,
      "grad_norm": 7.1542534828186035,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 0.706,
      "step": 7976
    },
    {
      "epoch": 2.0545,
      "grad_norm": 12.729436874389648,
      "learning_rate": 1.0217171717171719e-05,
      "loss": 0.674,
      "step": 7977
    },
    {
      "epoch": 2.0546,
      "grad_norm": 11.029817581176758,
      "learning_rate": 1.0212121212121213e-05,
      "loss": 0.5072,
      "step": 7978
    },
    {
      "epoch": 2.0547,
      "grad_norm": 8.805672645568848,
      "learning_rate": 1.0207070707070708e-05,
      "loss": 0.449,
      "step": 7979
    },
    {
      "epoch": 2.0548,
      "grad_norm": 12.730386734008789,
      "learning_rate": 1.0202020202020204e-05,
      "loss": 0.4186,
      "step": 7980
    },
    {
      "epoch": 2.0549,
      "grad_norm": 4.5673651695251465,
      "learning_rate": 1.0196969696969698e-05,
      "loss": 0.239,
      "step": 7981
    },
    {
      "epoch": 2.055,
      "grad_norm": 6.855194091796875,
      "learning_rate": 1.0191919191919192e-05,
      "loss": 0.225,
      "step": 7982
    },
    {
      "epoch": 2.0551,
      "grad_norm": 7.174311637878418,
      "learning_rate": 1.0186868686868688e-05,
      "loss": 0.3411,
      "step": 7983
    },
    {
      "epoch": 2.0552,
      "grad_norm": 6.263615131378174,
      "learning_rate": 1.0181818181818182e-05,
      "loss": 0.316,
      "step": 7984
    },
    {
      "epoch": 2.0553,
      "grad_norm": 5.784483909606934,
      "learning_rate": 1.0176767676767677e-05,
      "loss": 0.309,
      "step": 7985
    },
    {
      "epoch": 2.0554,
      "grad_norm": 6.776238918304443,
      "learning_rate": 1.0171717171717173e-05,
      "loss": 0.2839,
      "step": 7986
    },
    {
      "epoch": 2.0555,
      "grad_norm": 7.9300150871276855,
      "learning_rate": 1.0166666666666667e-05,
      "loss": 0.4297,
      "step": 7987
    },
    {
      "epoch": 2.0556,
      "grad_norm": 6.629661560058594,
      "learning_rate": 1.0161616161616163e-05,
      "loss": 0.7935,
      "step": 7988
    },
    {
      "epoch": 2.0557,
      "grad_norm": 5.415837287902832,
      "learning_rate": 1.0156565656565657e-05,
      "loss": 0.3703,
      "step": 7989
    },
    {
      "epoch": 2.0558,
      "grad_norm": 7.148609161376953,
      "learning_rate": 1.0151515151515152e-05,
      "loss": 0.4437,
      "step": 7990
    },
    {
      "epoch": 2.0559,
      "grad_norm": 3.4707770347595215,
      "learning_rate": 1.0146464646464648e-05,
      "loss": 0.2424,
      "step": 7991
    },
    {
      "epoch": 2.056,
      "grad_norm": 9.6856050491333,
      "learning_rate": 1.0141414141414142e-05,
      "loss": 0.7749,
      "step": 7992
    },
    {
      "epoch": 2.0561,
      "grad_norm": 5.784102439880371,
      "learning_rate": 1.0136363636363636e-05,
      "loss": 0.6372,
      "step": 7993
    },
    {
      "epoch": 2.0562,
      "grad_norm": 5.856524467468262,
      "learning_rate": 1.0131313131313132e-05,
      "loss": 0.6362,
      "step": 7994
    },
    {
      "epoch": 2.0563,
      "grad_norm": 11.007902145385742,
      "learning_rate": 1.0126262626262626e-05,
      "loss": 0.4894,
      "step": 7995
    },
    {
      "epoch": 2.0564,
      "grad_norm": 16.492219924926758,
      "learning_rate": 1.0121212121212122e-05,
      "loss": 0.2849,
      "step": 7996
    },
    {
      "epoch": 2.0565,
      "grad_norm": 3.484018325805664,
      "learning_rate": 1.0116161616161617e-05,
      "loss": 0.5974,
      "step": 7997
    },
    {
      "epoch": 2.0566,
      "grad_norm": 5.139404773712158,
      "learning_rate": 1.0111111111111111e-05,
      "loss": 0.6112,
      "step": 7998
    },
    {
      "epoch": 2.0567,
      "grad_norm": 9.72391128540039,
      "learning_rate": 1.0106060606060607e-05,
      "loss": 0.6339,
      "step": 7999
    },
    {
      "epoch": 2.0568,
      "grad_norm": 19.219982147216797,
      "learning_rate": 1.0101010101010101e-05,
      "loss": 0.5223,
      "step": 8000
    },
    {
      "epoch": 2.0569,
      "grad_norm": 15.601153373718262,
      "learning_rate": 1.0095959595959596e-05,
      "loss": 0.6815,
      "step": 8001
    },
    {
      "epoch": 2.057,
      "grad_norm": 8.488984107971191,
      "learning_rate": 1.0090909090909092e-05,
      "loss": 0.2841,
      "step": 8002
    },
    {
      "epoch": 2.0571,
      "grad_norm": 7.481717109680176,
      "learning_rate": 1.0085858585858586e-05,
      "loss": 0.6458,
      "step": 8003
    },
    {
      "epoch": 2.0572,
      "grad_norm": 13.247354507446289,
      "learning_rate": 1.008080808080808e-05,
      "loss": 0.2315,
      "step": 8004
    },
    {
      "epoch": 2.0573,
      "grad_norm": 7.810291290283203,
      "learning_rate": 1.0075757575757576e-05,
      "loss": 1.4177,
      "step": 8005
    },
    {
      "epoch": 2.0574,
      "grad_norm": 9.531362533569336,
      "learning_rate": 1.007070707070707e-05,
      "loss": 0.637,
      "step": 8006
    },
    {
      "epoch": 2.0575,
      "grad_norm": 7.033144950866699,
      "learning_rate": 1.0065656565656566e-05,
      "loss": 0.3246,
      "step": 8007
    },
    {
      "epoch": 2.0576,
      "grad_norm": 5.835668087005615,
      "learning_rate": 1.006060606060606e-05,
      "loss": 0.2296,
      "step": 8008
    },
    {
      "epoch": 2.0577,
      "grad_norm": 7.549180507659912,
      "learning_rate": 1.0055555555555555e-05,
      "loss": 0.7515,
      "step": 8009
    },
    {
      "epoch": 2.0578,
      "grad_norm": 4.8892621994018555,
      "learning_rate": 1.0050505050505051e-05,
      "loss": 0.2098,
      "step": 8010
    },
    {
      "epoch": 2.0579,
      "grad_norm": 7.263841152191162,
      "learning_rate": 1.0045454545454545e-05,
      "loss": 0.6022,
      "step": 8011
    },
    {
      "epoch": 2.058,
      "grad_norm": 6.881012439727783,
      "learning_rate": 1.004040404040404e-05,
      "loss": 0.455,
      "step": 8012
    },
    {
      "epoch": 2.0581,
      "grad_norm": 12.97314739227295,
      "learning_rate": 1.0035353535353536e-05,
      "loss": 0.5485,
      "step": 8013
    },
    {
      "epoch": 2.0582,
      "grad_norm": 19.396617889404297,
      "learning_rate": 1.003030303030303e-05,
      "loss": 0.2153,
      "step": 8014
    },
    {
      "epoch": 2.0583,
      "grad_norm": 7.4457550048828125,
      "learning_rate": 1.0025252525252524e-05,
      "loss": 1.0865,
      "step": 8015
    },
    {
      "epoch": 2.0584,
      "grad_norm": 4.1951422691345215,
      "learning_rate": 1.0020202020202022e-05,
      "loss": 0.2038,
      "step": 8016
    },
    {
      "epoch": 2.0585,
      "grad_norm": 8.389963150024414,
      "learning_rate": 1.0015151515151516e-05,
      "loss": 1.3566,
      "step": 8017
    },
    {
      "epoch": 2.0586,
      "grad_norm": 4.366209030151367,
      "learning_rate": 1.001010101010101e-05,
      "loss": 0.3632,
      "step": 8018
    },
    {
      "epoch": 2.0587,
      "grad_norm": 12.678149223327637,
      "learning_rate": 1.0005050505050507e-05,
      "loss": 0.2986,
      "step": 8019
    },
    {
      "epoch": 2.0588,
      "grad_norm": 5.423270225524902,
      "learning_rate": 1e-05,
      "loss": 0.2986,
      "step": 8020
    },
    {
      "epoch": 2.0589,
      "grad_norm": 5.560509204864502,
      "learning_rate": 9.994949494949495e-06,
      "loss": 0.4297,
      "step": 8021
    },
    {
      "epoch": 2.059,
      "grad_norm": 8.881644248962402,
      "learning_rate": 9.989898989898991e-06,
      "loss": 0.3803,
      "step": 8022
    },
    {
      "epoch": 2.0591,
      "grad_norm": 4.257493019104004,
      "learning_rate": 9.984848484848485e-06,
      "loss": 0.6806,
      "step": 8023
    },
    {
      "epoch": 2.0592,
      "grad_norm": 3.073432207107544,
      "learning_rate": 9.979797979797981e-06,
      "loss": 0.1467,
      "step": 8024
    },
    {
      "epoch": 2.0593,
      "grad_norm": 6.591687202453613,
      "learning_rate": 9.974747474747476e-06,
      "loss": 0.6699,
      "step": 8025
    },
    {
      "epoch": 2.0594,
      "grad_norm": 5.23692512512207,
      "learning_rate": 9.96969696969697e-06,
      "loss": 1.3038,
      "step": 8026
    },
    {
      "epoch": 2.0595,
      "grad_norm": 15.331793785095215,
      "learning_rate": 9.964646464646466e-06,
      "loss": 0.3324,
      "step": 8027
    },
    {
      "epoch": 2.0596,
      "grad_norm": 8.702374458312988,
      "learning_rate": 9.95959595959596e-06,
      "loss": 0.3767,
      "step": 8028
    },
    {
      "epoch": 2.0597,
      "grad_norm": 9.964350700378418,
      "learning_rate": 9.954545454545455e-06,
      "loss": 0.7638,
      "step": 8029
    },
    {
      "epoch": 2.0598,
      "grad_norm": 6.12257719039917,
      "learning_rate": 9.94949494949495e-06,
      "loss": 0.6488,
      "step": 8030
    },
    {
      "epoch": 2.0599,
      "grad_norm": 4.708653450012207,
      "learning_rate": 9.944444444444445e-06,
      "loss": 0.2692,
      "step": 8031
    },
    {
      "epoch": 2.06,
      "grad_norm": 9.265249252319336,
      "learning_rate": 9.93939393939394e-06,
      "loss": 0.1979,
      "step": 8032
    },
    {
      "epoch": 2.0601,
      "grad_norm": 11.127909660339355,
      "learning_rate": 9.934343434343435e-06,
      "loss": 0.4789,
      "step": 8033
    },
    {
      "epoch": 2.0602,
      "grad_norm": 5.71251916885376,
      "learning_rate": 9.92929292929293e-06,
      "loss": 0.3572,
      "step": 8034
    },
    {
      "epoch": 2.0603,
      "grad_norm": 11.984149932861328,
      "learning_rate": 9.924242424242425e-06,
      "loss": 0.4186,
      "step": 8035
    },
    {
      "epoch": 2.0604,
      "grad_norm": 5.560181140899658,
      "learning_rate": 9.91919191919192e-06,
      "loss": 0.2189,
      "step": 8036
    },
    {
      "epoch": 2.0605,
      "grad_norm": 7.749444007873535,
      "learning_rate": 9.914141414141414e-06,
      "loss": 0.8425,
      "step": 8037
    },
    {
      "epoch": 2.0606,
      "grad_norm": 4.575434684753418,
      "learning_rate": 9.90909090909091e-06,
      "loss": 0.2284,
      "step": 8038
    },
    {
      "epoch": 2.0607,
      "grad_norm": 8.468398094177246,
      "learning_rate": 9.904040404040404e-06,
      "loss": 0.9308,
      "step": 8039
    },
    {
      "epoch": 2.0608,
      "grad_norm": 5.083129405975342,
      "learning_rate": 9.898989898989899e-06,
      "loss": 0.2853,
      "step": 8040
    },
    {
      "epoch": 2.0609,
      "grad_norm": 4.752821445465088,
      "learning_rate": 9.893939393939395e-06,
      "loss": 0.8005,
      "step": 8041
    },
    {
      "epoch": 2.061,
      "grad_norm": 97.44420623779297,
      "learning_rate": 9.888888888888889e-06,
      "loss": 0.3712,
      "step": 8042
    },
    {
      "epoch": 2.0611,
      "grad_norm": 3.809279203414917,
      "learning_rate": 9.883838383838385e-06,
      "loss": 0.5625,
      "step": 8043
    },
    {
      "epoch": 2.0612,
      "grad_norm": 4.470684051513672,
      "learning_rate": 9.87878787878788e-06,
      "loss": 0.183,
      "step": 8044
    },
    {
      "epoch": 2.0613,
      "grad_norm": 20.637916564941406,
      "learning_rate": 9.873737373737373e-06,
      "loss": 0.1945,
      "step": 8045
    },
    {
      "epoch": 2.0614,
      "grad_norm": 5.608370780944824,
      "learning_rate": 9.86868686868687e-06,
      "loss": 0.2318,
      "step": 8046
    },
    {
      "epoch": 2.0615,
      "grad_norm": 4.538869857788086,
      "learning_rate": 9.863636363636364e-06,
      "loss": 0.2095,
      "step": 8047
    },
    {
      "epoch": 2.0616,
      "grad_norm": 6.8762993812561035,
      "learning_rate": 9.858585858585858e-06,
      "loss": 0.2075,
      "step": 8048
    },
    {
      "epoch": 2.0617,
      "grad_norm": 8.45614242553711,
      "learning_rate": 9.853535353535354e-06,
      "loss": 0.2533,
      "step": 8049
    },
    {
      "epoch": 2.0618,
      "grad_norm": 9.92778205871582,
      "learning_rate": 9.848484848484848e-06,
      "loss": 0.3188,
      "step": 8050
    },
    {
      "epoch": 2.0619,
      "grad_norm": 9.667013168334961,
      "learning_rate": 9.843434343434343e-06,
      "loss": 0.2213,
      "step": 8051
    },
    {
      "epoch": 2.062,
      "grad_norm": 9.623188972473145,
      "learning_rate": 9.838383838383839e-06,
      "loss": 0.6953,
      "step": 8052
    },
    {
      "epoch": 2.0621,
      "grad_norm": 30.42800521850586,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.3668,
      "step": 8053
    },
    {
      "epoch": 2.0622,
      "grad_norm": 3.493880271911621,
      "learning_rate": 9.828282828282829e-06,
      "loss": 0.576,
      "step": 8054
    },
    {
      "epoch": 2.0623,
      "grad_norm": 4.871952533721924,
      "learning_rate": 9.823232323232323e-06,
      "loss": 0.2118,
      "step": 8055
    },
    {
      "epoch": 2.0624,
      "grad_norm": 7.98559045791626,
      "learning_rate": 9.818181818181818e-06,
      "loss": 0.6578,
      "step": 8056
    },
    {
      "epoch": 2.0625,
      "grad_norm": 3.870793342590332,
      "learning_rate": 9.813131313131313e-06,
      "loss": 0.1999,
      "step": 8057
    },
    {
      "epoch": 2.0626,
      "grad_norm": 6.040370941162109,
      "learning_rate": 9.808080808080808e-06,
      "loss": 0.3884,
      "step": 8058
    },
    {
      "epoch": 2.0627,
      "grad_norm": 7.285652160644531,
      "learning_rate": 9.803030303030304e-06,
      "loss": 0.2213,
      "step": 8059
    },
    {
      "epoch": 2.0628,
      "grad_norm": 1.5303298234939575,
      "learning_rate": 9.7979797979798e-06,
      "loss": 0.0315,
      "step": 8060
    },
    {
      "epoch": 2.0629,
      "grad_norm": 3.885775566101074,
      "learning_rate": 9.792929292929294e-06,
      "loss": 0.0999,
      "step": 8061
    },
    {
      "epoch": 2.063,
      "grad_norm": 12.231526374816895,
      "learning_rate": 9.787878787878788e-06,
      "loss": 0.3764,
      "step": 8062
    },
    {
      "epoch": 2.0631,
      "grad_norm": 3.3583168983459473,
      "learning_rate": 9.782828282828284e-06,
      "loss": 0.4342,
      "step": 8063
    },
    {
      "epoch": 2.0632,
      "grad_norm": 1.9706108570098877,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.0912,
      "step": 8064
    },
    {
      "epoch": 2.0633,
      "grad_norm": 6.205560207366943,
      "learning_rate": 9.772727272727273e-06,
      "loss": 0.3201,
      "step": 8065
    },
    {
      "epoch": 2.0634,
      "grad_norm": 4.668756008148193,
      "learning_rate": 9.767676767676769e-06,
      "loss": 0.0893,
      "step": 8066
    },
    {
      "epoch": 2.0635,
      "grad_norm": 5.496912479400635,
      "learning_rate": 9.762626262626263e-06,
      "loss": 0.2165,
      "step": 8067
    },
    {
      "epoch": 2.0636,
      "grad_norm": 11.596955299377441,
      "learning_rate": 9.75757575757576e-06,
      "loss": 0.315,
      "step": 8068
    },
    {
      "epoch": 2.0637,
      "grad_norm": 3.425208330154419,
      "learning_rate": 9.752525252525254e-06,
      "loss": 0.0513,
      "step": 8069
    },
    {
      "epoch": 2.0638,
      "grad_norm": 1.5099778175354004,
      "learning_rate": 9.747474747474748e-06,
      "loss": 0.0155,
      "step": 8070
    },
    {
      "epoch": 2.0639,
      "grad_norm": 0.886323094367981,
      "learning_rate": 9.742424242424244e-06,
      "loss": 0.0202,
      "step": 8071
    },
    {
      "epoch": 2.064,
      "grad_norm": 1.9494019746780396,
      "learning_rate": 9.737373737373738e-06,
      "loss": 0.2037,
      "step": 8072
    },
    {
      "epoch": 2.0641,
      "grad_norm": 6.498388290405273,
      "learning_rate": 9.732323232323232e-06,
      "loss": 0.3553,
      "step": 8073
    },
    {
      "epoch": 2.0642,
      "grad_norm": 3.203371524810791,
      "learning_rate": 9.727272727272728e-06,
      "loss": 0.1158,
      "step": 8074
    },
    {
      "epoch": 2.0643,
      "grad_norm": 9.504006385803223,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.7166,
      "step": 8075
    },
    {
      "epoch": 2.0644,
      "grad_norm": 5.5203776359558105,
      "learning_rate": 9.717171717171717e-06,
      "loss": 0.347,
      "step": 8076
    },
    {
      "epoch": 2.0645,
      "grad_norm": 6.457955837249756,
      "learning_rate": 9.712121212121213e-06,
      "loss": 0.3862,
      "step": 8077
    },
    {
      "epoch": 2.0646,
      "grad_norm": 1.4865494966506958,
      "learning_rate": 9.707070707070707e-06,
      "loss": 0.0332,
      "step": 8078
    },
    {
      "epoch": 2.0647,
      "grad_norm": 3.1380722522735596,
      "learning_rate": 9.702020202020203e-06,
      "loss": 0.5394,
      "step": 8079
    },
    {
      "epoch": 2.0648,
      "grad_norm": 5.797513484954834,
      "learning_rate": 9.696969696969698e-06,
      "loss": 0.0582,
      "step": 8080
    },
    {
      "epoch": 2.0649,
      "grad_norm": 2.011821985244751,
      "learning_rate": 9.691919191919192e-06,
      "loss": 0.234,
      "step": 8081
    },
    {
      "epoch": 2.065,
      "grad_norm": 2.3099136352539062,
      "learning_rate": 9.686868686868688e-06,
      "loss": 0.0313,
      "step": 8082
    },
    {
      "epoch": 2.0651,
      "grad_norm": 8.593579292297363,
      "learning_rate": 9.681818181818182e-06,
      "loss": 0.2787,
      "step": 8083
    },
    {
      "epoch": 2.0652,
      "grad_norm": 13.272555351257324,
      "learning_rate": 9.676767676767676e-06,
      "loss": 1.2939,
      "step": 8084
    },
    {
      "epoch": 2.0653,
      "grad_norm": 5.821134090423584,
      "learning_rate": 9.671717171717172e-06,
      "loss": 1.5166,
      "step": 8085
    },
    {
      "epoch": 2.0654,
      "grad_norm": 49.554840087890625,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.4256,
      "step": 8086
    },
    {
      "epoch": 2.0655,
      "grad_norm": 5.042214870452881,
      "learning_rate": 9.661616161616161e-06,
      "loss": 0.5802,
      "step": 8087
    },
    {
      "epoch": 2.0656,
      "grad_norm": 9.206768989562988,
      "learning_rate": 9.656565656565657e-06,
      "loss": 0.212,
      "step": 8088
    },
    {
      "epoch": 2.0657,
      "grad_norm": 11.72902774810791,
      "learning_rate": 9.651515151515151e-06,
      "loss": 0.7342,
      "step": 8089
    },
    {
      "epoch": 2.0658,
      "grad_norm": 4.786969184875488,
      "learning_rate": 9.646464646464647e-06,
      "loss": 0.1065,
      "step": 8090
    },
    {
      "epoch": 2.0659,
      "grad_norm": 4.162802219390869,
      "learning_rate": 9.641414141414142e-06,
      "loss": 0.2852,
      "step": 8091
    },
    {
      "epoch": 2.066,
      "grad_norm": 2.078045606613159,
      "learning_rate": 9.636363636363636e-06,
      "loss": 0.0096,
      "step": 8092
    },
    {
      "epoch": 2.0661,
      "grad_norm": 3.1280057430267334,
      "learning_rate": 9.631313131313132e-06,
      "loss": 0.5011,
      "step": 8093
    },
    {
      "epoch": 2.0662,
      "grad_norm": 1.623963475227356,
      "learning_rate": 9.626262626262626e-06,
      "loss": 0.0373,
      "step": 8094
    },
    {
      "epoch": 2.0663,
      "grad_norm": 5.062607765197754,
      "learning_rate": 9.62121212121212e-06,
      "loss": 0.4231,
      "step": 8095
    },
    {
      "epoch": 2.0664,
      "grad_norm": 28.881948471069336,
      "learning_rate": 9.616161616161616e-06,
      "loss": 0.118,
      "step": 8096
    },
    {
      "epoch": 2.0665,
      "grad_norm": 6.804059982299805,
      "learning_rate": 9.61111111111111e-06,
      "loss": 0.3269,
      "step": 8097
    },
    {
      "epoch": 2.0666,
      "grad_norm": 11.760186195373535,
      "learning_rate": 9.606060606060607e-06,
      "loss": 0.7916,
      "step": 8098
    },
    {
      "epoch": 2.0667,
      "grad_norm": 3.2276768684387207,
      "learning_rate": 9.601010101010101e-06,
      "loss": 0.1145,
      "step": 8099
    },
    {
      "epoch": 2.0668,
      "grad_norm": 2.1920201778411865,
      "learning_rate": 9.595959595959595e-06,
      "loss": 0.2141,
      "step": 8100
    },
    {
      "epoch": 2.0669,
      "grad_norm": 7.964486122131348,
      "learning_rate": 9.590909090909091e-06,
      "loss": 0.6186,
      "step": 8101
    },
    {
      "epoch": 2.067,
      "grad_norm": 1.4578795433044434,
      "learning_rate": 9.585858585858587e-06,
      "loss": 0.0332,
      "step": 8102
    },
    {
      "epoch": 2.0671,
      "grad_norm": 2.9818332195281982,
      "learning_rate": 9.580808080808082e-06,
      "loss": 0.54,
      "step": 8103
    },
    {
      "epoch": 2.0672,
      "grad_norm": 6.342127323150635,
      "learning_rate": 9.575757575757578e-06,
      "loss": 0.0916,
      "step": 8104
    },
    {
      "epoch": 2.0673,
      "grad_norm": 15.268851280212402,
      "learning_rate": 9.570707070707072e-06,
      "loss": 0.3802,
      "step": 8105
    },
    {
      "epoch": 2.0674,
      "grad_norm": 7.064785480499268,
      "learning_rate": 9.565656565656566e-06,
      "loss": 0.148,
      "step": 8106
    },
    {
      "epoch": 2.0675,
      "grad_norm": 3.4544713497161865,
      "learning_rate": 9.560606060606062e-06,
      "loss": 0.2778,
      "step": 8107
    },
    {
      "epoch": 2.0676,
      "grad_norm": 8.323359489440918,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.3556,
      "step": 8108
    },
    {
      "epoch": 2.0677,
      "grad_norm": 15.053864479064941,
      "learning_rate": 9.55050505050505e-06,
      "loss": 0.1892,
      "step": 8109
    },
    {
      "epoch": 2.0678,
      "grad_norm": 6.750093936920166,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.197,
      "step": 8110
    },
    {
      "epoch": 2.0679,
      "grad_norm": 1.3469505310058594,
      "learning_rate": 9.540404040404041e-06,
      "loss": 0.0235,
      "step": 8111
    },
    {
      "epoch": 2.068,
      "grad_norm": 4.9094767570495605,
      "learning_rate": 9.535353535353535e-06,
      "loss": 0.1592,
      "step": 8112
    },
    {
      "epoch": 2.0681,
      "grad_norm": 3.170062780380249,
      "learning_rate": 9.530303030303031e-06,
      "loss": 0.031,
      "step": 8113
    },
    {
      "epoch": 2.0682,
      "grad_norm": 6.799862861633301,
      "learning_rate": 9.525252525252526e-06,
      "loss": 0.0459,
      "step": 8114
    },
    {
      "epoch": 2.0683,
      "grad_norm": 30.540193557739258,
      "learning_rate": 9.520202020202022e-06,
      "loss": 1.6127,
      "step": 8115
    },
    {
      "epoch": 2.0684,
      "grad_norm": 5.843868732452393,
      "learning_rate": 9.515151515151516e-06,
      "loss": 0.3398,
      "step": 8116
    },
    {
      "epoch": 2.0685000000000002,
      "grad_norm": 5.281851291656494,
      "learning_rate": 9.51010101010101e-06,
      "loss": 0.2491,
      "step": 8117
    },
    {
      "epoch": 2.0686,
      "grad_norm": 4.289869785308838,
      "learning_rate": 9.505050505050506e-06,
      "loss": 0.1016,
      "step": 8118
    },
    {
      "epoch": 2.0687,
      "grad_norm": 3.7567877769470215,
      "learning_rate": 9.5e-06,
      "loss": 0.1434,
      "step": 8119
    },
    {
      "epoch": 2.0688,
      "grad_norm": 1.0690585374832153,
      "learning_rate": 9.494949494949495e-06,
      "loss": 0.4738,
      "step": 8120
    },
    {
      "epoch": 2.0689,
      "grad_norm": 3.7131497859954834,
      "learning_rate": 9.48989898989899e-06,
      "loss": 1.0311,
      "step": 8121
    },
    {
      "epoch": 2.069,
      "grad_norm": 4.9453206062316895,
      "learning_rate": 9.484848484848485e-06,
      "loss": 0.2192,
      "step": 8122
    },
    {
      "epoch": 2.0691,
      "grad_norm": 6.127968788146973,
      "learning_rate": 9.47979797979798e-06,
      "loss": 0.1036,
      "step": 8123
    },
    {
      "epoch": 2.0692,
      "grad_norm": 2.486339569091797,
      "learning_rate": 9.474747474747475e-06,
      "loss": 0.0489,
      "step": 8124
    },
    {
      "epoch": 2.0693,
      "grad_norm": 1.2888368368148804,
      "learning_rate": 9.46969696969697e-06,
      "loss": 0.0241,
      "step": 8125
    },
    {
      "epoch": 2.0694,
      "grad_norm": 6.775229454040527,
      "learning_rate": 9.464646464646466e-06,
      "loss": 0.2539,
      "step": 8126
    },
    {
      "epoch": 2.0695,
      "grad_norm": 21.584922790527344,
      "learning_rate": 9.45959595959596e-06,
      "loss": 0.5033,
      "step": 8127
    },
    {
      "epoch": 2.0696,
      "grad_norm": 7.916574478149414,
      "learning_rate": 9.454545454545454e-06,
      "loss": 0.2401,
      "step": 8128
    },
    {
      "epoch": 2.0697,
      "grad_norm": 3.128427267074585,
      "learning_rate": 9.44949494949495e-06,
      "loss": 0.0437,
      "step": 8129
    },
    {
      "epoch": 2.0698,
      "grad_norm": 1.2253137826919556,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.0194,
      "step": 8130
    },
    {
      "epoch": 2.0699,
      "grad_norm": 2.4226436614990234,
      "learning_rate": 9.439393939393939e-06,
      "loss": 0.0564,
      "step": 8131
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.735198974609375,
      "learning_rate": 9.434343434343435e-06,
      "loss": 0.5185,
      "step": 8132
    },
    {
      "epoch": 2.0701,
      "grad_norm": 6.986057758331299,
      "learning_rate": 9.429292929292929e-06,
      "loss": 0.5712,
      "step": 8133
    },
    {
      "epoch": 2.0702,
      "grad_norm": 2.6293156147003174,
      "learning_rate": 9.424242424242425e-06,
      "loss": 0.1115,
      "step": 8134
    },
    {
      "epoch": 2.0703,
      "grad_norm": 7.001469135284424,
      "learning_rate": 9.41919191919192e-06,
      "loss": 0.5292,
      "step": 8135
    },
    {
      "epoch": 2.0704,
      "grad_norm": 8.227751731872559,
      "learning_rate": 9.414141414141414e-06,
      "loss": 1.2362,
      "step": 8136
    },
    {
      "epoch": 2.0705,
      "grad_norm": 3.1946794986724854,
      "learning_rate": 9.40909090909091e-06,
      "loss": 0.2135,
      "step": 8137
    },
    {
      "epoch": 2.0705999999999998,
      "grad_norm": 7.622948169708252,
      "learning_rate": 9.404040404040404e-06,
      "loss": 0.682,
      "step": 8138
    },
    {
      "epoch": 2.0707,
      "grad_norm": 8.358927726745605,
      "learning_rate": 9.398989898989898e-06,
      "loss": 0.1144,
      "step": 8139
    },
    {
      "epoch": 2.0708,
      "grad_norm": 6.423135280609131,
      "learning_rate": 9.393939393939394e-06,
      "loss": 0.16,
      "step": 8140
    },
    {
      "epoch": 2.0709,
      "grad_norm": 1.7942233085632324,
      "learning_rate": 9.388888888888889e-06,
      "loss": 0.0395,
      "step": 8141
    },
    {
      "epoch": 2.071,
      "grad_norm": 18.245058059692383,
      "learning_rate": 9.383838383838383e-06,
      "loss": 0.0596,
      "step": 8142
    },
    {
      "epoch": 2.0711,
      "grad_norm": 6.889775276184082,
      "learning_rate": 9.378787878787879e-06,
      "loss": 0.5988,
      "step": 8143
    },
    {
      "epoch": 2.0712,
      "grad_norm": 7.860355377197266,
      "learning_rate": 9.373737373737375e-06,
      "loss": 0.5211,
      "step": 8144
    },
    {
      "epoch": 2.0713,
      "grad_norm": 5.36297607421875,
      "learning_rate": 9.36868686868687e-06,
      "loss": 0.6357,
      "step": 8145
    },
    {
      "epoch": 2.0714,
      "grad_norm": 4.257750988006592,
      "learning_rate": 9.363636363636365e-06,
      "loss": 0.7203,
      "step": 8146
    },
    {
      "epoch": 2.0715,
      "grad_norm": 4.427670955657959,
      "learning_rate": 9.35858585858586e-06,
      "loss": 1.2395,
      "step": 8147
    },
    {
      "epoch": 2.0716,
      "grad_norm": 3.078942060470581,
      "learning_rate": 9.353535353535354e-06,
      "loss": 0.0376,
      "step": 8148
    },
    {
      "epoch": 2.0717,
      "grad_norm": 0.6068412065505981,
      "learning_rate": 9.34848484848485e-06,
      "loss": 0.0083,
      "step": 8149
    },
    {
      "epoch": 2.0718,
      "grad_norm": 3.7366089820861816,
      "learning_rate": 9.343434343434344e-06,
      "loss": 0.5145,
      "step": 8150
    },
    {
      "epoch": 2.0719,
      "grad_norm": 5.876011848449707,
      "learning_rate": 9.33838383838384e-06,
      "loss": 0.3045,
      "step": 8151
    },
    {
      "epoch": 2.072,
      "grad_norm": 6.8809380531311035,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.8414,
      "step": 8152
    },
    {
      "epoch": 2.0721,
      "grad_norm": 4.2047576904296875,
      "learning_rate": 9.328282828282829e-06,
      "loss": 0.1564,
      "step": 8153
    },
    {
      "epoch": 2.0722,
      "grad_norm": 5.253091335296631,
      "learning_rate": 9.323232323232325e-06,
      "loss": 0.2776,
      "step": 8154
    },
    {
      "epoch": 2.0723,
      "grad_norm": 1.9352107048034668,
      "learning_rate": 9.318181818181819e-06,
      "loss": 0.0483,
      "step": 8155
    },
    {
      "epoch": 2.0724,
      "grad_norm": 0.8674207925796509,
      "learning_rate": 9.313131313131313e-06,
      "loss": 0.4445,
      "step": 8156
    },
    {
      "epoch": 2.0725,
      "grad_norm": 1.5654754638671875,
      "learning_rate": 9.30808080808081e-06,
      "loss": 0.021,
      "step": 8157
    },
    {
      "epoch": 2.0726,
      "grad_norm": 5.020177364349365,
      "learning_rate": 9.303030303030303e-06,
      "loss": 0.697,
      "step": 8158
    },
    {
      "epoch": 2.0727,
      "grad_norm": 6.122411251068115,
      "learning_rate": 9.297979797979798e-06,
      "loss": 0.1592,
      "step": 8159
    },
    {
      "epoch": 2.0728,
      "grad_norm": 4.235081672668457,
      "learning_rate": 9.292929292929294e-06,
      "loss": 0.1879,
      "step": 8160
    },
    {
      "epoch": 2.0729,
      "grad_norm": 6.109891891479492,
      "learning_rate": 9.287878787878788e-06,
      "loss": 0.1708,
      "step": 8161
    },
    {
      "epoch": 2.073,
      "grad_norm": 4.756462097167969,
      "learning_rate": 9.282828282828284e-06,
      "loss": 0.1339,
      "step": 8162
    },
    {
      "epoch": 2.0731,
      "grad_norm": 2.1436243057250977,
      "learning_rate": 9.277777777777778e-06,
      "loss": 0.6302,
      "step": 8163
    },
    {
      "epoch": 2.0732,
      "grad_norm": 2.592557668685913,
      "learning_rate": 9.272727272727273e-06,
      "loss": 0.0195,
      "step": 8164
    },
    {
      "epoch": 2.0733,
      "grad_norm": 8.078291893005371,
      "learning_rate": 9.267676767676769e-06,
      "loss": 0.1758,
      "step": 8165
    },
    {
      "epoch": 2.0734,
      "grad_norm": 5.598640441894531,
      "learning_rate": 9.262626262626263e-06,
      "loss": 0.1247,
      "step": 8166
    },
    {
      "epoch": 2.0735,
      "grad_norm": 4.631314277648926,
      "learning_rate": 9.257575757575757e-06,
      "loss": 0.3705,
      "step": 8167
    },
    {
      "epoch": 2.0736,
      "grad_norm": 0.7004482746124268,
      "learning_rate": 9.252525252525253e-06,
      "loss": 0.4462,
      "step": 8168
    },
    {
      "epoch": 2.0737,
      "grad_norm": 5.134483814239502,
      "learning_rate": 9.247474747474748e-06,
      "loss": 0.2842,
      "step": 8169
    },
    {
      "epoch": 2.0738,
      "grad_norm": 2.8210997581481934,
      "learning_rate": 9.242424242424244e-06,
      "loss": 0.4747,
      "step": 8170
    },
    {
      "epoch": 2.0739,
      "grad_norm": 1.1315772533416748,
      "learning_rate": 9.237373737373738e-06,
      "loss": 0.0131,
      "step": 8171
    },
    {
      "epoch": 2.074,
      "grad_norm": 1.3096588850021362,
      "learning_rate": 9.232323232323232e-06,
      "loss": 0.0169,
      "step": 8172
    },
    {
      "epoch": 2.0741,
      "grad_norm": 14.284083366394043,
      "learning_rate": 9.227272727272728e-06,
      "loss": 0.7316,
      "step": 8173
    },
    {
      "epoch": 2.0742,
      "grad_norm": 3.4988043308258057,
      "learning_rate": 9.222222222222222e-06,
      "loss": 0.2431,
      "step": 8174
    },
    {
      "epoch": 2.0743,
      "grad_norm": 2.684617042541504,
      "learning_rate": 9.217171717171717e-06,
      "loss": 0.2611,
      "step": 8175
    },
    {
      "epoch": 2.0744,
      "grad_norm": 11.099268913269043,
      "learning_rate": 9.212121212121213e-06,
      "loss": 0.7826,
      "step": 8176
    },
    {
      "epoch": 2.0745,
      "grad_norm": 5.961709499359131,
      "learning_rate": 9.207070707070707e-06,
      "loss": 0.1356,
      "step": 8177
    },
    {
      "epoch": 2.0746,
      "grad_norm": 7.294568061828613,
      "learning_rate": 9.202020202020201e-06,
      "loss": 0.3174,
      "step": 8178
    },
    {
      "epoch": 2.0747,
      "grad_norm": 2.1836607456207275,
      "learning_rate": 9.196969696969697e-06,
      "loss": 0.06,
      "step": 8179
    },
    {
      "epoch": 2.0748,
      "grad_norm": 4.666771411895752,
      "learning_rate": 9.191919191919192e-06,
      "loss": 0.6784,
      "step": 8180
    },
    {
      "epoch": 2.0749,
      "grad_norm": 1.0530509948730469,
      "learning_rate": 9.186868686868688e-06,
      "loss": 0.0169,
      "step": 8181
    },
    {
      "epoch": 2.075,
      "grad_norm": 24.14056396484375,
      "learning_rate": 9.181818181818182e-06,
      "loss": 0.3021,
      "step": 8182
    },
    {
      "epoch": 2.0751,
      "grad_norm": 3.4257500171661377,
      "learning_rate": 9.176767676767676e-06,
      "loss": 0.2477,
      "step": 8183
    },
    {
      "epoch": 2.0752,
      "grad_norm": 1.1370246410369873,
      "learning_rate": 9.171717171717172e-06,
      "loss": 0.025,
      "step": 8184
    },
    {
      "epoch": 2.0753,
      "grad_norm": 3.645282745361328,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.1036,
      "step": 8185
    },
    {
      "epoch": 2.0754,
      "grad_norm": 4.590909957885742,
      "learning_rate": 9.16161616161616e-06,
      "loss": 0.1235,
      "step": 8186
    },
    {
      "epoch": 2.0755,
      "grad_norm": 1.7227556705474854,
      "learning_rate": 9.156565656565658e-06,
      "loss": 0.0587,
      "step": 8187
    },
    {
      "epoch": 2.0756,
      "grad_norm": 6.613701820373535,
      "learning_rate": 9.151515151515153e-06,
      "loss": 0.3547,
      "step": 8188
    },
    {
      "epoch": 2.0757,
      "grad_norm": 4.372057914733887,
      "learning_rate": 9.146464646464647e-06,
      "loss": 0.1026,
      "step": 8189
    },
    {
      "epoch": 2.0758,
      "grad_norm": 6.214123249053955,
      "learning_rate": 9.141414141414143e-06,
      "loss": 0.4559,
      "step": 8190
    },
    {
      "epoch": 2.0759,
      "grad_norm": 7.173543930053711,
      "learning_rate": 9.136363636363637e-06,
      "loss": 0.59,
      "step": 8191
    },
    {
      "epoch": 2.076,
      "grad_norm": 14.329381942749023,
      "learning_rate": 9.131313131313132e-06,
      "loss": 0.8421,
      "step": 8192
    },
    {
      "epoch": 2.0761,
      "grad_norm": 3.3848328590393066,
      "learning_rate": 9.126262626262628e-06,
      "loss": 0.6844,
      "step": 8193
    },
    {
      "epoch": 2.0762,
      "grad_norm": 4.1351704597473145,
      "learning_rate": 9.121212121212122e-06,
      "loss": 0.1135,
      "step": 8194
    },
    {
      "epoch": 2.0763,
      "grad_norm": 3.166936159133911,
      "learning_rate": 9.116161616161616e-06,
      "loss": 0.068,
      "step": 8195
    },
    {
      "epoch": 2.0764,
      "grad_norm": 6.440316677093506,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.3601,
      "step": 8196
    },
    {
      "epoch": 2.0765,
      "grad_norm": 4.703660488128662,
      "learning_rate": 9.106060606060606e-06,
      "loss": 0.7722,
      "step": 8197
    },
    {
      "epoch": 2.0766,
      "grad_norm": 5.674811363220215,
      "learning_rate": 9.101010101010102e-06,
      "loss": 0.1696,
      "step": 8198
    },
    {
      "epoch": 2.0767,
      "grad_norm": 3.204075336456299,
      "learning_rate": 9.095959595959597e-06,
      "loss": 0.0544,
      "step": 8199
    },
    {
      "epoch": 2.0768,
      "grad_norm": 2.937525749206543,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.2321,
      "step": 8200
    },
    {
      "epoch": 2.0769,
      "grad_norm": 2.528459310531616,
      "learning_rate": 9.085858585858587e-06,
      "loss": 0.0673,
      "step": 8201
    },
    {
      "epoch": 2.077,
      "grad_norm": 5.371732234954834,
      "learning_rate": 9.080808080808081e-06,
      "loss": 0.0531,
      "step": 8202
    },
    {
      "epoch": 2.0771,
      "grad_norm": 7.427900314331055,
      "learning_rate": 9.075757575757576e-06,
      "loss": 0.3647,
      "step": 8203
    },
    {
      "epoch": 2.0772,
      "grad_norm": 4.181116580963135,
      "learning_rate": 9.070707070707072e-06,
      "loss": 0.0469,
      "step": 8204
    },
    {
      "epoch": 2.0773,
      "grad_norm": 8.294105529785156,
      "learning_rate": 9.065656565656566e-06,
      "loss": 0.4387,
      "step": 8205
    },
    {
      "epoch": 2.0774,
      "grad_norm": 5.367554664611816,
      "learning_rate": 9.060606060606062e-06,
      "loss": 0.1186,
      "step": 8206
    },
    {
      "epoch": 2.0775,
      "grad_norm": 4.621823310852051,
      "learning_rate": 9.055555555555556e-06,
      "loss": 0.127,
      "step": 8207
    },
    {
      "epoch": 2.0776,
      "grad_norm": 8.095108032226562,
      "learning_rate": 9.05050505050505e-06,
      "loss": 0.2712,
      "step": 8208
    },
    {
      "epoch": 2.0777,
      "grad_norm": 3.571167230606079,
      "learning_rate": 9.045454545454546e-06,
      "loss": 0.5228,
      "step": 8209
    },
    {
      "epoch": 2.0778,
      "grad_norm": 7.386204242706299,
      "learning_rate": 9.04040404040404e-06,
      "loss": 0.6752,
      "step": 8210
    },
    {
      "epoch": 2.0779,
      "grad_norm": 4.697990417480469,
      "learning_rate": 9.035353535353535e-06,
      "loss": 0.2687,
      "step": 8211
    },
    {
      "epoch": 2.078,
      "grad_norm": 4.837382793426514,
      "learning_rate": 9.030303030303031e-06,
      "loss": 0.1098,
      "step": 8212
    },
    {
      "epoch": 2.0781,
      "grad_norm": 6.678194522857666,
      "learning_rate": 9.025252525252525e-06,
      "loss": 0.7834,
      "step": 8213
    },
    {
      "epoch": 2.0782,
      "grad_norm": 6.1951799392700195,
      "learning_rate": 9.02020202020202e-06,
      "loss": 0.7002,
      "step": 8214
    },
    {
      "epoch": 2.0783,
      "grad_norm": 5.6435723304748535,
      "learning_rate": 9.015151515151516e-06,
      "loss": 0.322,
      "step": 8215
    },
    {
      "epoch": 2.0784,
      "grad_norm": 4.424534320831299,
      "learning_rate": 9.01010101010101e-06,
      "loss": 0.2198,
      "step": 8216
    },
    {
      "epoch": 2.0785,
      "grad_norm": 1.6751519441604614,
      "learning_rate": 9.005050505050506e-06,
      "loss": 0.0396,
      "step": 8217
    },
    {
      "epoch": 2.0786,
      "grad_norm": 7.176828861236572,
      "learning_rate": 9e-06,
      "loss": 0.1007,
      "step": 8218
    },
    {
      "epoch": 2.0787,
      "grad_norm": 2.089249610900879,
      "learning_rate": 8.994949494949495e-06,
      "loss": 0.0365,
      "step": 8219
    },
    {
      "epoch": 2.0788,
      "grad_norm": 3.4951188564300537,
      "learning_rate": 8.98989898989899e-06,
      "loss": 0.0949,
      "step": 8220
    },
    {
      "epoch": 2.0789,
      "grad_norm": 6.077789306640625,
      "learning_rate": 8.984848484848485e-06,
      "loss": 0.5276,
      "step": 8221
    },
    {
      "epoch": 2.079,
      "grad_norm": 4.579398155212402,
      "learning_rate": 8.979797979797979e-06,
      "loss": 0.183,
      "step": 8222
    },
    {
      "epoch": 2.0791,
      "grad_norm": 3.0340399742126465,
      "learning_rate": 8.974747474747475e-06,
      "loss": 0.1118,
      "step": 8223
    },
    {
      "epoch": 2.0792,
      "grad_norm": 2.2746970653533936,
      "learning_rate": 8.96969696969697e-06,
      "loss": 0.0448,
      "step": 8224
    },
    {
      "epoch": 2.0793,
      "grad_norm": 10.905118942260742,
      "learning_rate": 8.964646464646464e-06,
      "loss": 0.2669,
      "step": 8225
    },
    {
      "epoch": 2.0794,
      "grad_norm": 3.1494383811950684,
      "learning_rate": 8.95959595959596e-06,
      "loss": 0.1082,
      "step": 8226
    },
    {
      "epoch": 2.0795,
      "grad_norm": 5.44968318939209,
      "learning_rate": 8.954545454545454e-06,
      "loss": 0.3549,
      "step": 8227
    },
    {
      "epoch": 2.0796,
      "grad_norm": 1.0155854225158691,
      "learning_rate": 8.94949494949495e-06,
      "loss": 0.0357,
      "step": 8228
    },
    {
      "epoch": 2.0797,
      "grad_norm": 5.0908522605896,
      "learning_rate": 8.944444444444444e-06,
      "loss": 0.1632,
      "step": 8229
    },
    {
      "epoch": 2.0798,
      "grad_norm": 4.287431240081787,
      "learning_rate": 8.93939393939394e-06,
      "loss": 0.1163,
      "step": 8230
    },
    {
      "epoch": 2.0799,
      "grad_norm": 18.887779235839844,
      "learning_rate": 8.934343434343436e-06,
      "loss": 0.3742,
      "step": 8231
    },
    {
      "epoch": 2.08,
      "grad_norm": 21.765085220336914,
      "learning_rate": 8.92929292929293e-06,
      "loss": 0.1053,
      "step": 8232
    },
    {
      "epoch": 2.0801,
      "grad_norm": 4.191557884216309,
      "learning_rate": 8.924242424242425e-06,
      "loss": 0.5503,
      "step": 8233
    },
    {
      "epoch": 2.0802,
      "grad_norm": 7.551177024841309,
      "learning_rate": 8.91919191919192e-06,
      "loss": 1.2322,
      "step": 8234
    },
    {
      "epoch": 2.0803,
      "grad_norm": 3.6781864166259766,
      "learning_rate": 8.914141414141415e-06,
      "loss": 0.2747,
      "step": 8235
    },
    {
      "epoch": 2.0804,
      "grad_norm": 10.071224212646484,
      "learning_rate": 8.90909090909091e-06,
      "loss": 0.2697,
      "step": 8236
    },
    {
      "epoch": 2.0805,
      "grad_norm": 6.878726005554199,
      "learning_rate": 8.904040404040405e-06,
      "loss": 0.2504,
      "step": 8237
    },
    {
      "epoch": 2.0806,
      "grad_norm": 1.6683140993118286,
      "learning_rate": 8.8989898989899e-06,
      "loss": 0.0582,
      "step": 8238
    },
    {
      "epoch": 2.0807,
      "grad_norm": 2.015449047088623,
      "learning_rate": 8.893939393939394e-06,
      "loss": 0.0524,
      "step": 8239
    },
    {
      "epoch": 2.0808,
      "grad_norm": 5.683555603027344,
      "learning_rate": 8.88888888888889e-06,
      "loss": 0.7678,
      "step": 8240
    },
    {
      "epoch": 2.0809,
      "grad_norm": 1.2113044261932373,
      "learning_rate": 8.883838383838384e-06,
      "loss": 0.0271,
      "step": 8241
    },
    {
      "epoch": 2.081,
      "grad_norm": 22.5190486907959,
      "learning_rate": 8.87878787878788e-06,
      "loss": 0.0496,
      "step": 8242
    },
    {
      "epoch": 2.0811,
      "grad_norm": 7.444010257720947,
      "learning_rate": 8.873737373737375e-06,
      "loss": 0.1204,
      "step": 8243
    },
    {
      "epoch": 2.0812,
      "grad_norm": 4.484131813049316,
      "learning_rate": 8.868686868686869e-06,
      "loss": 0.6261,
      "step": 8244
    },
    {
      "epoch": 2.0813,
      "grad_norm": 7.39255952835083,
      "learning_rate": 8.863636363636365e-06,
      "loss": 0.3269,
      "step": 8245
    },
    {
      "epoch": 2.0814,
      "grad_norm": 4.550231456756592,
      "learning_rate": 8.85858585858586e-06,
      "loss": 0.3292,
      "step": 8246
    },
    {
      "epoch": 2.0815,
      "grad_norm": 9.636701583862305,
      "learning_rate": 8.853535353535353e-06,
      "loss": 0.2997,
      "step": 8247
    },
    {
      "epoch": 2.0816,
      "grad_norm": 2.5582776069641113,
      "learning_rate": 8.84848484848485e-06,
      "loss": 0.0995,
      "step": 8248
    },
    {
      "epoch": 2.0817,
      "grad_norm": 9.976761817932129,
      "learning_rate": 8.843434343434344e-06,
      "loss": 0.8022,
      "step": 8249
    },
    {
      "epoch": 2.0818,
      "grad_norm": 4.688576698303223,
      "learning_rate": 8.838383838383838e-06,
      "loss": 0.1235,
      "step": 8250
    },
    {
      "epoch": 2.0819,
      "grad_norm": 1.1338003873825073,
      "learning_rate": 8.833333333333334e-06,
      "loss": 0.0277,
      "step": 8251
    },
    {
      "epoch": 2.082,
      "grad_norm": 3.308408498764038,
      "learning_rate": 8.828282828282828e-06,
      "loss": 0.018,
      "step": 8252
    },
    {
      "epoch": 2.0821,
      "grad_norm": 36.051517486572266,
      "learning_rate": 8.823232323232324e-06,
      "loss": 0.2427,
      "step": 8253
    },
    {
      "epoch": 2.0822,
      "grad_norm": 2.8814566135406494,
      "learning_rate": 8.818181818181819e-06,
      "loss": 0.0354,
      "step": 8254
    },
    {
      "epoch": 2.0823,
      "grad_norm": 4.365131855010986,
      "learning_rate": 8.813131313131313e-06,
      "loss": 0.5758,
      "step": 8255
    },
    {
      "epoch": 2.0824,
      "grad_norm": 4.490920066833496,
      "learning_rate": 8.808080808080809e-06,
      "loss": 1.0474,
      "step": 8256
    },
    {
      "epoch": 2.0825,
      "grad_norm": 3.9323678016662598,
      "learning_rate": 8.803030303030303e-06,
      "loss": 0.5668,
      "step": 8257
    },
    {
      "epoch": 2.0826000000000002,
      "grad_norm": 2.2375998497009277,
      "learning_rate": 8.797979797979797e-06,
      "loss": 0.0517,
      "step": 8258
    },
    {
      "epoch": 2.0827,
      "grad_norm": 4.703373432159424,
      "learning_rate": 8.792929292929293e-06,
      "loss": 0.2803,
      "step": 8259
    },
    {
      "epoch": 2.0828,
      "grad_norm": 4.181105613708496,
      "learning_rate": 8.787878787878788e-06,
      "loss": 0.1546,
      "step": 8260
    },
    {
      "epoch": 2.0829,
      "grad_norm": 4.82415246963501,
      "learning_rate": 8.782828282828282e-06,
      "loss": 0.5254,
      "step": 8261
    },
    {
      "epoch": 2.083,
      "grad_norm": 0.7149443030357361,
      "learning_rate": 8.777777777777778e-06,
      "loss": 0.0113,
      "step": 8262
    },
    {
      "epoch": 2.0831,
      "grad_norm": 7.970099449157715,
      "learning_rate": 8.772727272727272e-06,
      "loss": 0.2104,
      "step": 8263
    },
    {
      "epoch": 2.0832,
      "grad_norm": 4.798157215118408,
      "learning_rate": 8.767676767676768e-06,
      "loss": 0.1307,
      "step": 8264
    },
    {
      "epoch": 2.0833,
      "grad_norm": 2.2388224601745605,
      "learning_rate": 8.762626262626263e-06,
      "loss": 0.2234,
      "step": 8265
    },
    {
      "epoch": 2.0834,
      "grad_norm": 7.84674596786499,
      "learning_rate": 8.757575757575757e-06,
      "loss": 0.4251,
      "step": 8266
    },
    {
      "epoch": 2.0835,
      "grad_norm": 2.8419189453125,
      "learning_rate": 8.752525252525253e-06,
      "loss": 0.0489,
      "step": 8267
    },
    {
      "epoch": 2.0836,
      "grad_norm": 3.7134673595428467,
      "learning_rate": 8.747474747474747e-06,
      "loss": 0.1709,
      "step": 8268
    },
    {
      "epoch": 2.0837,
      "grad_norm": 3.3112409114837646,
      "learning_rate": 8.742424242424242e-06,
      "loss": 0.083,
      "step": 8269
    },
    {
      "epoch": 2.0838,
      "grad_norm": 5.712350845336914,
      "learning_rate": 8.737373737373738e-06,
      "loss": 0.1025,
      "step": 8270
    },
    {
      "epoch": 2.0839,
      "grad_norm": 2.800114154815674,
      "learning_rate": 8.732323232323232e-06,
      "loss": 0.2429,
      "step": 8271
    },
    {
      "epoch": 2.084,
      "grad_norm": 4.121086597442627,
      "learning_rate": 8.727272727272728e-06,
      "loss": 0.0226,
      "step": 8272
    },
    {
      "epoch": 2.0841,
      "grad_norm": 6.248196601867676,
      "learning_rate": 8.722222222222224e-06,
      "loss": 0.3053,
      "step": 8273
    },
    {
      "epoch": 2.0842,
      "grad_norm": 3.4625813961029053,
      "learning_rate": 8.717171717171718e-06,
      "loss": 0.507,
      "step": 8274
    },
    {
      "epoch": 2.0843,
      "grad_norm": 3.3886935710906982,
      "learning_rate": 8.712121212121212e-06,
      "loss": 0.3002,
      "step": 8275
    },
    {
      "epoch": 2.0844,
      "grad_norm": 3.3240203857421875,
      "learning_rate": 8.707070707070708e-06,
      "loss": 0.072,
      "step": 8276
    },
    {
      "epoch": 2.0845,
      "grad_norm": 5.23786735534668,
      "learning_rate": 8.702020202020203e-06,
      "loss": 0.1225,
      "step": 8277
    },
    {
      "epoch": 2.0846,
      "grad_norm": 5.5836310386657715,
      "learning_rate": 8.696969696969699e-06,
      "loss": 0.2366,
      "step": 8278
    },
    {
      "epoch": 2.0846999999999998,
      "grad_norm": 4.814572811126709,
      "learning_rate": 8.691919191919193e-06,
      "loss": 0.0917,
      "step": 8279
    },
    {
      "epoch": 2.0848,
      "grad_norm": 3.596740245819092,
      "learning_rate": 8.686868686868687e-06,
      "loss": 0.1198,
      "step": 8280
    },
    {
      "epoch": 2.0849,
      "grad_norm": 1.5018556118011475,
      "learning_rate": 8.681818181818183e-06,
      "loss": 0.4977,
      "step": 8281
    },
    {
      "epoch": 2.085,
      "grad_norm": 9.509903907775879,
      "learning_rate": 8.676767676767678e-06,
      "loss": 0.2523,
      "step": 8282
    },
    {
      "epoch": 2.0851,
      "grad_norm": 4.305244445800781,
      "learning_rate": 8.671717171717172e-06,
      "loss": 0.2706,
      "step": 8283
    },
    {
      "epoch": 2.0852,
      "grad_norm": 4.682595729827881,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.2583,
      "step": 8284
    },
    {
      "epoch": 2.0853,
      "grad_norm": 4.363198280334473,
      "learning_rate": 8.661616161616162e-06,
      "loss": 0.5458,
      "step": 8285
    },
    {
      "epoch": 2.0854,
      "grad_norm": 23.028135299682617,
      "learning_rate": 8.656565656565656e-06,
      "loss": 0.1924,
      "step": 8286
    },
    {
      "epoch": 2.0855,
      "grad_norm": 8.925971031188965,
      "learning_rate": 8.651515151515152e-06,
      "loss": 0.7868,
      "step": 8287
    },
    {
      "epoch": 2.0856,
      "grad_norm": 4.385901927947998,
      "learning_rate": 8.646464646464647e-06,
      "loss": 0.2189,
      "step": 8288
    },
    {
      "epoch": 2.0857,
      "grad_norm": 5.59233283996582,
      "learning_rate": 8.641414141414143e-06,
      "loss": 0.5789,
      "step": 8289
    },
    {
      "epoch": 2.0858,
      "grad_norm": 2.5960052013397217,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.2489,
      "step": 8290
    },
    {
      "epoch": 2.0859,
      "grad_norm": 7.346348285675049,
      "learning_rate": 8.631313131313131e-06,
      "loss": 0.3018,
      "step": 8291
    },
    {
      "epoch": 2.086,
      "grad_norm": 1.8792016506195068,
      "learning_rate": 8.626262626262627e-06,
      "loss": 0.0512,
      "step": 8292
    },
    {
      "epoch": 2.0861,
      "grad_norm": 4.854771614074707,
      "learning_rate": 8.621212121212122e-06,
      "loss": 0.0808,
      "step": 8293
    },
    {
      "epoch": 2.0862,
      "grad_norm": 3.118868350982666,
      "learning_rate": 8.616161616161616e-06,
      "loss": 0.0124,
      "step": 8294
    },
    {
      "epoch": 2.0863,
      "grad_norm": 1.6146717071533203,
      "learning_rate": 8.611111111111112e-06,
      "loss": 0.0298,
      "step": 8295
    },
    {
      "epoch": 2.0864,
      "grad_norm": 3.028447151184082,
      "learning_rate": 8.606060606060606e-06,
      "loss": 0.0881,
      "step": 8296
    },
    {
      "epoch": 2.0865,
      "grad_norm": 4.493955135345459,
      "learning_rate": 8.6010101010101e-06,
      "loss": 0.5373,
      "step": 8297
    },
    {
      "epoch": 2.0866,
      "grad_norm": 0.40880918502807617,
      "learning_rate": 8.595959595959596e-06,
      "loss": 0.0033,
      "step": 8298
    },
    {
      "epoch": 2.0867,
      "grad_norm": 5.493949890136719,
      "learning_rate": 8.59090909090909e-06,
      "loss": 0.1107,
      "step": 8299
    },
    {
      "epoch": 2.0868,
      "grad_norm": 2.9775655269622803,
      "learning_rate": 8.585858585858587e-06,
      "loss": 0.134,
      "step": 8300
    },
    {
      "epoch": 2.0869,
      "grad_norm": 3.4409713745117188,
      "learning_rate": 8.580808080808081e-06,
      "loss": 0.1152,
      "step": 8301
    },
    {
      "epoch": 2.087,
      "grad_norm": 3.7290351390838623,
      "learning_rate": 8.575757575757575e-06,
      "loss": 0.0709,
      "step": 8302
    },
    {
      "epoch": 2.0871,
      "grad_norm": 5.748896598815918,
      "learning_rate": 8.570707070707071e-06,
      "loss": 0.6409,
      "step": 8303
    },
    {
      "epoch": 2.0872,
      "grad_norm": 0.698280930519104,
      "learning_rate": 8.565656565656566e-06,
      "loss": 0.0086,
      "step": 8304
    },
    {
      "epoch": 2.0873,
      "grad_norm": 2.807546377182007,
      "learning_rate": 8.56060606060606e-06,
      "loss": 0.2237,
      "step": 8305
    },
    {
      "epoch": 2.0874,
      "grad_norm": 8.23444938659668,
      "learning_rate": 8.555555555555556e-06,
      "loss": 0.6461,
      "step": 8306
    },
    {
      "epoch": 2.0875,
      "grad_norm": 5.42748498916626,
      "learning_rate": 8.55050505050505e-06,
      "loss": 0.6023,
      "step": 8307
    },
    {
      "epoch": 2.0876,
      "grad_norm": 3.5657877922058105,
      "learning_rate": 8.545454545454546e-06,
      "loss": 0.1147,
      "step": 8308
    },
    {
      "epoch": 2.0877,
      "grad_norm": 3.6526403427124023,
      "learning_rate": 8.54040404040404e-06,
      "loss": 0.4609,
      "step": 8309
    },
    {
      "epoch": 2.0878,
      "grad_norm": 6.112612247467041,
      "learning_rate": 8.535353535353535e-06,
      "loss": 0.1821,
      "step": 8310
    },
    {
      "epoch": 2.0879,
      "grad_norm": 5.341404438018799,
      "learning_rate": 8.53030303030303e-06,
      "loss": 0.2173,
      "step": 8311
    },
    {
      "epoch": 2.088,
      "grad_norm": 4.660440444946289,
      "learning_rate": 8.525252525252525e-06,
      "loss": 0.3024,
      "step": 8312
    },
    {
      "epoch": 2.0881,
      "grad_norm": 2.0678212642669678,
      "learning_rate": 8.52020202020202e-06,
      "loss": 0.0975,
      "step": 8313
    },
    {
      "epoch": 2.0882,
      "grad_norm": 2.0739800930023193,
      "learning_rate": 8.515151515151515e-06,
      "loss": 0.495,
      "step": 8314
    },
    {
      "epoch": 2.0883,
      "grad_norm": 2.0350937843322754,
      "learning_rate": 8.510101010101011e-06,
      "loss": 0.0673,
      "step": 8315
    },
    {
      "epoch": 2.0884,
      "grad_norm": 20.573823928833008,
      "learning_rate": 8.505050505050506e-06,
      "loss": 0.1953,
      "step": 8316
    },
    {
      "epoch": 2.0885,
      "grad_norm": 5.495995998382568,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.7368,
      "step": 8317
    },
    {
      "epoch": 2.0886,
      "grad_norm": 1.2642676830291748,
      "learning_rate": 8.494949494949496e-06,
      "loss": 0.0209,
      "step": 8318
    },
    {
      "epoch": 2.0887000000000002,
      "grad_norm": 5.503946304321289,
      "learning_rate": 8.48989898989899e-06,
      "loss": 0.1222,
      "step": 8319
    },
    {
      "epoch": 2.0888,
      "grad_norm": 3.770504951477051,
      "learning_rate": 8.484848484848486e-06,
      "loss": 0.0923,
      "step": 8320
    },
    {
      "epoch": 2.0889,
      "grad_norm": 3.4579615592956543,
      "learning_rate": 8.47979797979798e-06,
      "loss": 0.1552,
      "step": 8321
    },
    {
      "epoch": 2.089,
      "grad_norm": 1.2260949611663818,
      "learning_rate": 8.474747474747475e-06,
      "loss": 0.0201,
      "step": 8322
    },
    {
      "epoch": 2.0891,
      "grad_norm": 0.5097748637199402,
      "learning_rate": 8.46969696969697e-06,
      "loss": 0.0044,
      "step": 8323
    },
    {
      "epoch": 2.0892,
      "grad_norm": 3.1843953132629395,
      "learning_rate": 8.464646464646465e-06,
      "loss": 0.061,
      "step": 8324
    },
    {
      "epoch": 2.0893,
      "grad_norm": 5.076607704162598,
      "learning_rate": 8.459595959595961e-06,
      "loss": 0.0632,
      "step": 8325
    },
    {
      "epoch": 2.0894,
      "grad_norm": 8.410199165344238,
      "learning_rate": 8.454545454545455e-06,
      "loss": 0.5212,
      "step": 8326
    },
    {
      "epoch": 2.0895,
      "grad_norm": 5.320818901062012,
      "learning_rate": 8.44949494949495e-06,
      "loss": 0.5754,
      "step": 8327
    },
    {
      "epoch": 2.0896,
      "grad_norm": 2.412588357925415,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.0607,
      "step": 8328
    },
    {
      "epoch": 2.0897,
      "grad_norm": 5.221775531768799,
      "learning_rate": 8.43939393939394e-06,
      "loss": 0.0774,
      "step": 8329
    },
    {
      "epoch": 2.0898,
      "grad_norm": 2.9507856369018555,
      "learning_rate": 8.434343434343434e-06,
      "loss": 0.2956,
      "step": 8330
    },
    {
      "epoch": 2.0899,
      "grad_norm": 4.322076797485352,
      "learning_rate": 8.42929292929293e-06,
      "loss": 0.4756,
      "step": 8331
    },
    {
      "epoch": 2.09,
      "grad_norm": 5.465985298156738,
      "learning_rate": 8.424242424242425e-06,
      "loss": 0.6756,
      "step": 8332
    },
    {
      "epoch": 2.0901,
      "grad_norm": 9.294780731201172,
      "learning_rate": 8.41919191919192e-06,
      "loss": 0.3172,
      "step": 8333
    },
    {
      "epoch": 2.0902,
      "grad_norm": 4.682340621948242,
      "learning_rate": 8.414141414141415e-06,
      "loss": 0.096,
      "step": 8334
    },
    {
      "epoch": 2.0903,
      "grad_norm": 8.42701244354248,
      "learning_rate": 8.409090909090909e-06,
      "loss": 0.0638,
      "step": 8335
    },
    {
      "epoch": 2.0904,
      "grad_norm": 1.9525656700134277,
      "learning_rate": 8.404040404040405e-06,
      "loss": 0.1998,
      "step": 8336
    },
    {
      "epoch": 2.0905,
      "grad_norm": 4.1694488525390625,
      "learning_rate": 8.3989898989899e-06,
      "loss": 0.6118,
      "step": 8337
    },
    {
      "epoch": 2.0906,
      "grad_norm": 1.8733659982681274,
      "learning_rate": 8.393939393939394e-06,
      "loss": 0.0634,
      "step": 8338
    },
    {
      "epoch": 2.0907,
      "grad_norm": 1.347395896911621,
      "learning_rate": 8.38888888888889e-06,
      "loss": 0.0289,
      "step": 8339
    },
    {
      "epoch": 2.0908,
      "grad_norm": 4.0821757316589355,
      "learning_rate": 8.383838383838384e-06,
      "loss": 0.079,
      "step": 8340
    },
    {
      "epoch": 2.0909,
      "grad_norm": 3.404991626739502,
      "learning_rate": 8.378787878787878e-06,
      "loss": 0.5161,
      "step": 8341
    },
    {
      "epoch": 2.091,
      "grad_norm": 3.537285089492798,
      "learning_rate": 8.373737373737374e-06,
      "loss": 0.5565,
      "step": 8342
    },
    {
      "epoch": 2.0911,
      "grad_norm": 3.287252902984619,
      "learning_rate": 8.368686868686869e-06,
      "loss": 0.0881,
      "step": 8343
    },
    {
      "epoch": 2.0912,
      "grad_norm": 5.175621032714844,
      "learning_rate": 8.363636363636365e-06,
      "loss": 0.5727,
      "step": 8344
    },
    {
      "epoch": 2.0913,
      "grad_norm": 3.7094523906707764,
      "learning_rate": 8.358585858585859e-06,
      "loss": 0.1247,
      "step": 8345
    },
    {
      "epoch": 2.0914,
      "grad_norm": 3.221021890640259,
      "learning_rate": 8.353535353535353e-06,
      "loss": 0.078,
      "step": 8346
    },
    {
      "epoch": 2.0915,
      "grad_norm": 8.157715797424316,
      "learning_rate": 8.348484848484849e-06,
      "loss": 0.6603,
      "step": 8347
    },
    {
      "epoch": 2.0916,
      "grad_norm": 4.075290203094482,
      "learning_rate": 8.343434343434343e-06,
      "loss": 0.1213,
      "step": 8348
    },
    {
      "epoch": 2.0917,
      "grad_norm": 3.894411087036133,
      "learning_rate": 8.338383838383838e-06,
      "loss": 0.1005,
      "step": 8349
    },
    {
      "epoch": 2.0918,
      "grad_norm": 2.569836139678955,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0277,
      "step": 8350
    },
    {
      "epoch": 2.0919,
      "grad_norm": 5.174343585968018,
      "learning_rate": 8.328282828282828e-06,
      "loss": 0.7124,
      "step": 8351
    },
    {
      "epoch": 2.092,
      "grad_norm": 3.7215816974639893,
      "learning_rate": 8.323232323232322e-06,
      "loss": 0.2966,
      "step": 8352
    },
    {
      "epoch": 2.0921,
      "grad_norm": 3.143409490585327,
      "learning_rate": 8.318181818181818e-06,
      "loss": 0.0721,
      "step": 8353
    },
    {
      "epoch": 2.0922,
      "grad_norm": 1.7170485258102417,
      "learning_rate": 8.313131313131313e-06,
      "loss": 0.0307,
      "step": 8354
    },
    {
      "epoch": 2.0923,
      "grad_norm": 0.5668884515762329,
      "learning_rate": 8.308080808080809e-06,
      "loss": 0.0194,
      "step": 8355
    },
    {
      "epoch": 2.0924,
      "grad_norm": 15.06493854522705,
      "learning_rate": 8.303030303030303e-06,
      "loss": 0.6177,
      "step": 8356
    },
    {
      "epoch": 2.0925,
      "grad_norm": 1.9032890796661377,
      "learning_rate": 8.297979797979797e-06,
      "loss": 0.0326,
      "step": 8357
    },
    {
      "epoch": 2.0926,
      "grad_norm": 3.155622720718384,
      "learning_rate": 8.292929292929293e-06,
      "loss": 0.2949,
      "step": 8358
    },
    {
      "epoch": 2.0927,
      "grad_norm": 6.2281670570373535,
      "learning_rate": 8.28787878787879e-06,
      "loss": 0.3163,
      "step": 8359
    },
    {
      "epoch": 2.0928,
      "grad_norm": 3.0065979957580566,
      "learning_rate": 8.282828282828283e-06,
      "loss": 0.5261,
      "step": 8360
    },
    {
      "epoch": 2.0929,
      "grad_norm": 20.02219009399414,
      "learning_rate": 8.27777777777778e-06,
      "loss": 0.2733,
      "step": 8361
    },
    {
      "epoch": 2.093,
      "grad_norm": 2.653541088104248,
      "learning_rate": 8.272727272727274e-06,
      "loss": 0.112,
      "step": 8362
    },
    {
      "epoch": 2.0931,
      "grad_norm": 2.169816493988037,
      "learning_rate": 8.267676767676768e-06,
      "loss": 0.0432,
      "step": 8363
    },
    {
      "epoch": 2.0932,
      "grad_norm": 9.218538284301758,
      "learning_rate": 8.262626262626264e-06,
      "loss": 0.0929,
      "step": 8364
    },
    {
      "epoch": 2.0933,
      "grad_norm": 1.5800268650054932,
      "learning_rate": 8.257575757575758e-06,
      "loss": 0.4544,
      "step": 8365
    },
    {
      "epoch": 2.0934,
      "grad_norm": 2.7908427715301514,
      "learning_rate": 8.252525252525253e-06,
      "loss": 0.054,
      "step": 8366
    },
    {
      "epoch": 2.0935,
      "grad_norm": 2.247694253921509,
      "learning_rate": 8.247474747474749e-06,
      "loss": 0.2026,
      "step": 8367
    },
    {
      "epoch": 2.0936,
      "grad_norm": 15.520794868469238,
      "learning_rate": 8.242424242424243e-06,
      "loss": 0.3416,
      "step": 8368
    },
    {
      "epoch": 2.0937,
      "grad_norm": 5.32484769821167,
      "learning_rate": 8.237373737373739e-06,
      "loss": 0.7502,
      "step": 8369
    },
    {
      "epoch": 2.0938,
      "grad_norm": 2.5402956008911133,
      "learning_rate": 8.232323232323233e-06,
      "loss": 0.0405,
      "step": 8370
    },
    {
      "epoch": 2.0939,
      "grad_norm": 2.5889647006988525,
      "learning_rate": 8.227272727272728e-06,
      "loss": 0.6601,
      "step": 8371
    },
    {
      "epoch": 2.094,
      "grad_norm": 2.677809953689575,
      "learning_rate": 8.222222222222223e-06,
      "loss": 0.0908,
      "step": 8372
    },
    {
      "epoch": 2.0941,
      "grad_norm": 5.263143062591553,
      "learning_rate": 8.217171717171718e-06,
      "loss": 0.3304,
      "step": 8373
    },
    {
      "epoch": 2.0942,
      "grad_norm": 4.386558532714844,
      "learning_rate": 8.212121212121212e-06,
      "loss": 0.5346,
      "step": 8374
    },
    {
      "epoch": 2.0943,
      "grad_norm": 2.936912775039673,
      "learning_rate": 8.207070707070708e-06,
      "loss": 0.0829,
      "step": 8375
    },
    {
      "epoch": 2.0944,
      "grad_norm": 6.386848449707031,
      "learning_rate": 8.202020202020202e-06,
      "loss": 0.1137,
      "step": 8376
    },
    {
      "epoch": 2.0945,
      "grad_norm": 0.9253285527229309,
      "learning_rate": 8.196969696969697e-06,
      "loss": 0.0144,
      "step": 8377
    },
    {
      "epoch": 2.0946,
      "grad_norm": 2.5039632320404053,
      "learning_rate": 8.191919191919193e-06,
      "loss": 0.2257,
      "step": 8378
    },
    {
      "epoch": 2.0947,
      "grad_norm": 2.032074213027954,
      "learning_rate": 8.186868686868687e-06,
      "loss": 0.0479,
      "step": 8379
    },
    {
      "epoch": 2.0948,
      "grad_norm": 10.415968894958496,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.5514,
      "step": 8380
    },
    {
      "epoch": 2.0949,
      "grad_norm": 3.1523754596710205,
      "learning_rate": 8.176767676767677e-06,
      "loss": 0.5826,
      "step": 8381
    },
    {
      "epoch": 2.095,
      "grad_norm": 2.6133944988250732,
      "learning_rate": 8.171717171717172e-06,
      "loss": 0.5636,
      "step": 8382
    },
    {
      "epoch": 2.0951,
      "grad_norm": 2.884615659713745,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.2087,
      "step": 8383
    },
    {
      "epoch": 2.0952,
      "grad_norm": 8.837599754333496,
      "learning_rate": 8.161616161616162e-06,
      "loss": 0.3266,
      "step": 8384
    },
    {
      "epoch": 2.0953,
      "grad_norm": 1.9356902837753296,
      "learning_rate": 8.156565656565656e-06,
      "loss": 0.0423,
      "step": 8385
    },
    {
      "epoch": 2.0954,
      "grad_norm": 2.3210346698760986,
      "learning_rate": 8.151515151515152e-06,
      "loss": 0.0656,
      "step": 8386
    },
    {
      "epoch": 2.0955,
      "grad_norm": 3.4417622089385986,
      "learning_rate": 8.146464646464646e-06,
      "loss": 0.1572,
      "step": 8387
    },
    {
      "epoch": 2.0956,
      "grad_norm": 7.327733993530273,
      "learning_rate": 8.14141414141414e-06,
      "loss": 0.1441,
      "step": 8388
    },
    {
      "epoch": 2.0957,
      "grad_norm": 3.9609556198120117,
      "learning_rate": 8.136363636363637e-06,
      "loss": 0.2309,
      "step": 8389
    },
    {
      "epoch": 2.0958,
      "grad_norm": 9.860526084899902,
      "learning_rate": 8.131313131313131e-06,
      "loss": 0.4872,
      "step": 8390
    },
    {
      "epoch": 2.0959,
      "grad_norm": 2.3627560138702393,
      "learning_rate": 8.126262626262627e-06,
      "loss": 0.0438,
      "step": 8391
    },
    {
      "epoch": 2.096,
      "grad_norm": 4.227623462677002,
      "learning_rate": 8.121212121212121e-06,
      "loss": 0.261,
      "step": 8392
    },
    {
      "epoch": 2.0961,
      "grad_norm": 5.877654552459717,
      "learning_rate": 8.116161616161616e-06,
      "loss": 0.276,
      "step": 8393
    },
    {
      "epoch": 2.0962,
      "grad_norm": 3.194059371948242,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.0993,
      "step": 8394
    },
    {
      "epoch": 2.0963,
      "grad_norm": 0.8386120200157166,
      "learning_rate": 8.106060606060606e-06,
      "loss": 0.0225,
      "step": 8395
    },
    {
      "epoch": 2.0964,
      "grad_norm": 2.1727402210235596,
      "learning_rate": 8.1010101010101e-06,
      "loss": 0.0126,
      "step": 8396
    },
    {
      "epoch": 2.0965,
      "grad_norm": 4.077530860900879,
      "learning_rate": 8.095959595959596e-06,
      "loss": 0.1098,
      "step": 8397
    },
    {
      "epoch": 2.0966,
      "grad_norm": 2.730673313140869,
      "learning_rate": 8.09090909090909e-06,
      "loss": 0.2246,
      "step": 8398
    },
    {
      "epoch": 2.0967,
      "grad_norm": 3.3750991821289062,
      "learning_rate": 8.085858585858585e-06,
      "loss": 0.0801,
      "step": 8399
    },
    {
      "epoch": 2.0968,
      "grad_norm": 4.751196384429932,
      "learning_rate": 8.080808080808082e-06,
      "loss": 0.0693,
      "step": 8400
    },
    {
      "epoch": 2.0969,
      "grad_norm": 2.643112897872925,
      "learning_rate": 8.075757575757577e-06,
      "loss": 0.1574,
      "step": 8401
    },
    {
      "epoch": 2.097,
      "grad_norm": 7.28637170791626,
      "learning_rate": 8.070707070707071e-06,
      "loss": 0.3103,
      "step": 8402
    },
    {
      "epoch": 2.0971,
      "grad_norm": 4.343286037445068,
      "learning_rate": 8.065656565656567e-06,
      "loss": 0.0395,
      "step": 8403
    },
    {
      "epoch": 2.0972,
      "grad_norm": 1.589514136314392,
      "learning_rate": 8.060606060606061e-06,
      "loss": 0.0244,
      "step": 8404
    },
    {
      "epoch": 2.0973,
      "grad_norm": 5.335031986236572,
      "learning_rate": 8.055555555555557e-06,
      "loss": 0.1245,
      "step": 8405
    },
    {
      "epoch": 2.0974,
      "grad_norm": 1.2724292278289795,
      "learning_rate": 8.050505050505052e-06,
      "loss": 0.007,
      "step": 8406
    },
    {
      "epoch": 2.0975,
      "grad_norm": 3.591002941131592,
      "learning_rate": 8.045454545454546e-06,
      "loss": 0.9615,
      "step": 8407
    },
    {
      "epoch": 2.0976,
      "grad_norm": 13.474946975708008,
      "learning_rate": 8.040404040404042e-06,
      "loss": 0.1417,
      "step": 8408
    },
    {
      "epoch": 2.0977,
      "grad_norm": 7.459499835968018,
      "learning_rate": 8.035353535353536e-06,
      "loss": 0.4123,
      "step": 8409
    },
    {
      "epoch": 2.0978,
      "grad_norm": 10.339279174804688,
      "learning_rate": 8.03030303030303e-06,
      "loss": 0.6717,
      "step": 8410
    },
    {
      "epoch": 2.0979,
      "grad_norm": 1.4853198528289795,
      "learning_rate": 8.025252525252526e-06,
      "loss": 0.0273,
      "step": 8411
    },
    {
      "epoch": 2.098,
      "grad_norm": 13.451398849487305,
      "learning_rate": 8.02020202020202e-06,
      "loss": 0.3497,
      "step": 8412
    },
    {
      "epoch": 2.0981,
      "grad_norm": 4.193254470825195,
      "learning_rate": 8.015151515151515e-06,
      "loss": 0.983,
      "step": 8413
    },
    {
      "epoch": 2.0982,
      "grad_norm": 3.404106855392456,
      "learning_rate": 8.010101010101011e-06,
      "loss": 0.0723,
      "step": 8414
    },
    {
      "epoch": 2.0983,
      "grad_norm": 1.9284934997558594,
      "learning_rate": 8.005050505050505e-06,
      "loss": 0.0309,
      "step": 8415
    },
    {
      "epoch": 2.0984,
      "grad_norm": 4.939024925231934,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2366,
      "step": 8416
    },
    {
      "epoch": 2.0985,
      "grad_norm": 3.771944522857666,
      "learning_rate": 7.994949494949496e-06,
      "loss": 0.1048,
      "step": 8417
    },
    {
      "epoch": 2.0986,
      "grad_norm": 9.659847259521484,
      "learning_rate": 7.98989898989899e-06,
      "loss": 0.6583,
      "step": 8418
    },
    {
      "epoch": 2.0987,
      "grad_norm": 5.558146953582764,
      "learning_rate": 7.984848484848486e-06,
      "loss": 0.066,
      "step": 8419
    },
    {
      "epoch": 2.0987999999999998,
      "grad_norm": 4.864151954650879,
      "learning_rate": 7.97979797979798e-06,
      "loss": 0.2104,
      "step": 8420
    },
    {
      "epoch": 2.0989,
      "grad_norm": 4.555793285369873,
      "learning_rate": 7.974747474747475e-06,
      "loss": 0.5006,
      "step": 8421
    },
    {
      "epoch": 2.099,
      "grad_norm": 1.8390192985534668,
      "learning_rate": 7.96969696969697e-06,
      "loss": 0.1939,
      "step": 8422
    },
    {
      "epoch": 2.0991,
      "grad_norm": 0.41212019324302673,
      "learning_rate": 7.964646464646465e-06,
      "loss": 0.0105,
      "step": 8423
    },
    {
      "epoch": 2.0992,
      "grad_norm": 3.119842290878296,
      "learning_rate": 7.959595959595959e-06,
      "loss": 0.066,
      "step": 8424
    },
    {
      "epoch": 2.0993,
      "grad_norm": 2.769488573074341,
      "learning_rate": 7.954545454545455e-06,
      "loss": 0.1795,
      "step": 8425
    },
    {
      "epoch": 2.0994,
      "grad_norm": 20.836458206176758,
      "learning_rate": 7.94949494949495e-06,
      "loss": 0.2971,
      "step": 8426
    },
    {
      "epoch": 2.0995,
      "grad_norm": 0.4990002512931824,
      "learning_rate": 7.944444444444445e-06,
      "loss": 0.0042,
      "step": 8427
    },
    {
      "epoch": 2.0996,
      "grad_norm": 4.027958869934082,
      "learning_rate": 7.93939393939394e-06,
      "loss": 0.0903,
      "step": 8428
    },
    {
      "epoch": 2.0997,
      "grad_norm": 7.158139228820801,
      "learning_rate": 7.934343434343434e-06,
      "loss": 0.1814,
      "step": 8429
    },
    {
      "epoch": 2.0998,
      "grad_norm": 3.3891806602478027,
      "learning_rate": 7.92929292929293e-06,
      "loss": 0.0354,
      "step": 8430
    },
    {
      "epoch": 2.0999,
      "grad_norm": 6.224186420440674,
      "learning_rate": 7.924242424242424e-06,
      "loss": 0.18,
      "step": 8431
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.7184135913848877,
      "learning_rate": 7.919191919191919e-06,
      "loss": 0.0193,
      "step": 8432
    },
    {
      "epoch": 2.1001,
      "grad_norm": 0.711763858795166,
      "learning_rate": 7.914141414141415e-06,
      "loss": 0.0059,
      "step": 8433
    },
    {
      "epoch": 2.1002,
      "grad_norm": 1.4353039264678955,
      "learning_rate": 7.909090909090909e-06,
      "loss": 0.0608,
      "step": 8434
    },
    {
      "epoch": 2.1003,
      "grad_norm": 6.840133190155029,
      "learning_rate": 7.904040404040403e-06,
      "loss": 0.3092,
      "step": 8435
    },
    {
      "epoch": 2.1004,
      "grad_norm": 4.281399250030518,
      "learning_rate": 7.898989898989899e-06,
      "loss": 0.5261,
      "step": 8436
    },
    {
      "epoch": 2.1005,
      "grad_norm": 6.873133182525635,
      "learning_rate": 7.893939393939393e-06,
      "loss": 1.0512,
      "step": 8437
    },
    {
      "epoch": 2.1006,
      "grad_norm": 10.902819633483887,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.5765,
      "step": 8438
    },
    {
      "epoch": 2.1007,
      "grad_norm": 3.0187253952026367,
      "learning_rate": 7.883838383838384e-06,
      "loss": 0.0355,
      "step": 8439
    },
    {
      "epoch": 2.1008,
      "grad_norm": 3.1948490142822266,
      "learning_rate": 7.878787878787878e-06,
      "loss": 0.0428,
      "step": 8440
    },
    {
      "epoch": 2.1009,
      "grad_norm": 5.869186878204346,
      "learning_rate": 7.873737373737374e-06,
      "loss": 0.1104,
      "step": 8441
    },
    {
      "epoch": 2.101,
      "grad_norm": 1.4026079177856445,
      "learning_rate": 7.868686868686868e-06,
      "loss": 0.0227,
      "step": 8442
    },
    {
      "epoch": 2.1011,
      "grad_norm": 5.640840530395508,
      "learning_rate": 7.863636363636364e-06,
      "loss": 0.2806,
      "step": 8443
    },
    {
      "epoch": 2.1012,
      "grad_norm": 1.364786148071289,
      "learning_rate": 7.85858585858586e-06,
      "loss": 0.015,
      "step": 8444
    },
    {
      "epoch": 2.1013,
      "grad_norm": 6.401142120361328,
      "learning_rate": 7.853535353535355e-06,
      "loss": 0.1319,
      "step": 8445
    },
    {
      "epoch": 2.1014,
      "grad_norm": 8.70822525024414,
      "learning_rate": 7.848484848484849e-06,
      "loss": 0.2769,
      "step": 8446
    },
    {
      "epoch": 2.1015,
      "grad_norm": 2.7708282470703125,
      "learning_rate": 7.843434343434345e-06,
      "loss": 0.4909,
      "step": 8447
    },
    {
      "epoch": 2.1016,
      "grad_norm": 4.654189586639404,
      "learning_rate": 7.838383838383839e-06,
      "loss": 0.2532,
      "step": 8448
    },
    {
      "epoch": 2.1017,
      "grad_norm": 1.2688452005386353,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.4739,
      "step": 8449
    },
    {
      "epoch": 2.1018,
      "grad_norm": 19.124229431152344,
      "learning_rate": 7.82828282828283e-06,
      "loss": 0.1025,
      "step": 8450
    },
    {
      "epoch": 2.1019,
      "grad_norm": 13.8939790725708,
      "learning_rate": 7.823232323232324e-06,
      "loss": 0.1288,
      "step": 8451
    },
    {
      "epoch": 2.102,
      "grad_norm": 2.661203384399414,
      "learning_rate": 7.81818181818182e-06,
      "loss": 0.2509,
      "step": 8452
    },
    {
      "epoch": 2.1021,
      "grad_norm": 10.06440258026123,
      "learning_rate": 7.813131313131314e-06,
      "loss": 0.1162,
      "step": 8453
    },
    {
      "epoch": 2.1022,
      "grad_norm": 3.4321482181549072,
      "learning_rate": 7.808080808080808e-06,
      "loss": 0.1238,
      "step": 8454
    },
    {
      "epoch": 2.1023,
      "grad_norm": 2.0316367149353027,
      "learning_rate": 7.803030303030304e-06,
      "loss": 0.6499,
      "step": 8455
    },
    {
      "epoch": 2.1024,
      "grad_norm": 6.929229259490967,
      "learning_rate": 7.797979797979799e-06,
      "loss": 0.7725,
      "step": 8456
    },
    {
      "epoch": 2.1025,
      "grad_norm": 4.2721052169799805,
      "learning_rate": 7.792929292929293e-06,
      "loss": 0.1464,
      "step": 8457
    },
    {
      "epoch": 2.1026,
      "grad_norm": 0.7906271815299988,
      "learning_rate": 7.787878787878789e-06,
      "loss": 0.0071,
      "step": 8458
    },
    {
      "epoch": 2.1027,
      "grad_norm": 6.631953239440918,
      "learning_rate": 7.782828282828283e-06,
      "loss": 0.2168,
      "step": 8459
    },
    {
      "epoch": 2.1028000000000002,
      "grad_norm": 5.825253009796143,
      "learning_rate": 7.777777777777777e-06,
      "loss": 0.2915,
      "step": 8460
    },
    {
      "epoch": 2.1029,
      "grad_norm": 8.687987327575684,
      "learning_rate": 7.772727272727273e-06,
      "loss": 0.5898,
      "step": 8461
    },
    {
      "epoch": 2.103,
      "grad_norm": 1.9530099630355835,
      "learning_rate": 7.767676767676768e-06,
      "loss": 0.0611,
      "step": 8462
    },
    {
      "epoch": 2.1031,
      "grad_norm": 3.8343052864074707,
      "learning_rate": 7.762626262626264e-06,
      "loss": 0.0767,
      "step": 8463
    },
    {
      "epoch": 2.1032,
      "grad_norm": 4.867516994476318,
      "learning_rate": 7.757575757575758e-06,
      "loss": 0.0949,
      "step": 8464
    },
    {
      "epoch": 2.1033,
      "grad_norm": 6.363272666931152,
      "learning_rate": 7.752525252525252e-06,
      "loss": 0.6041,
      "step": 8465
    },
    {
      "epoch": 2.1034,
      "grad_norm": 10.93178653717041,
      "learning_rate": 7.747474747474748e-06,
      "loss": 0.2463,
      "step": 8466
    },
    {
      "epoch": 2.1035,
      "grad_norm": 4.293187618255615,
      "learning_rate": 7.742424242424243e-06,
      "loss": 0.9593,
      "step": 8467
    },
    {
      "epoch": 2.1036,
      "grad_norm": 1.3920074701309204,
      "learning_rate": 7.737373737373737e-06,
      "loss": 0.035,
      "step": 8468
    },
    {
      "epoch": 2.1037,
      "grad_norm": 5.448213577270508,
      "learning_rate": 7.732323232323233e-06,
      "loss": 0.2997,
      "step": 8469
    },
    {
      "epoch": 2.1038,
      "grad_norm": 7.424942493438721,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.5928,
      "step": 8470
    },
    {
      "epoch": 2.1039,
      "grad_norm": 3.466822862625122,
      "learning_rate": 7.722222222222223e-06,
      "loss": 0.1063,
      "step": 8471
    },
    {
      "epoch": 2.104,
      "grad_norm": 9.088530540466309,
      "learning_rate": 7.717171717171717e-06,
      "loss": 0.2867,
      "step": 8472
    },
    {
      "epoch": 2.1041,
      "grad_norm": 1.9852373600006104,
      "learning_rate": 7.712121212121212e-06,
      "loss": 0.4712,
      "step": 8473
    },
    {
      "epoch": 2.1042,
      "grad_norm": 5.39760684967041,
      "learning_rate": 7.707070707070708e-06,
      "loss": 0.2307,
      "step": 8474
    },
    {
      "epoch": 2.1043,
      "grad_norm": 1.1646716594696045,
      "learning_rate": 7.702020202020202e-06,
      "loss": 0.017,
      "step": 8475
    },
    {
      "epoch": 2.1044,
      "grad_norm": 4.175097465515137,
      "learning_rate": 7.696969696969696e-06,
      "loss": 0.0579,
      "step": 8476
    },
    {
      "epoch": 2.1045,
      "grad_norm": 1.6457709074020386,
      "learning_rate": 7.691919191919192e-06,
      "loss": 0.0328,
      "step": 8477
    },
    {
      "epoch": 2.1046,
      "grad_norm": 4.242518901824951,
      "learning_rate": 7.686868686868687e-06,
      "loss": 0.1544,
      "step": 8478
    },
    {
      "epoch": 2.1047,
      "grad_norm": 0.8973390460014343,
      "learning_rate": 7.681818181818181e-06,
      "loss": 0.0138,
      "step": 8479
    },
    {
      "epoch": 2.1048,
      "grad_norm": 4.322976589202881,
      "learning_rate": 7.676767676767677e-06,
      "loss": 0.1112,
      "step": 8480
    },
    {
      "epoch": 2.1048999999999998,
      "grad_norm": 11.974750518798828,
      "learning_rate": 7.671717171717171e-06,
      "loss": 0.9919,
      "step": 8481
    },
    {
      "epoch": 2.105,
      "grad_norm": 1.089332938194275,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.0079,
      "step": 8482
    },
    {
      "epoch": 2.1051,
      "grad_norm": 3.752842664718628,
      "learning_rate": 7.661616161616162e-06,
      "loss": 0.1014,
      "step": 8483
    },
    {
      "epoch": 2.1052,
      "grad_norm": 6.320112228393555,
      "learning_rate": 7.656565656565656e-06,
      "loss": 0.1783,
      "step": 8484
    },
    {
      "epoch": 2.1053,
      "grad_norm": 2.5094361305236816,
      "learning_rate": 7.651515151515152e-06,
      "loss": 0.1844,
      "step": 8485
    },
    {
      "epoch": 2.1054,
      "grad_norm": 2.8210716247558594,
      "learning_rate": 7.646464646464648e-06,
      "loss": 0.0472,
      "step": 8486
    },
    {
      "epoch": 2.1055,
      "grad_norm": 3.0685315132141113,
      "learning_rate": 7.641414141414142e-06,
      "loss": 0.0411,
      "step": 8487
    },
    {
      "epoch": 2.1056,
      "grad_norm": 2.4093074798583984,
      "learning_rate": 7.636363636363638e-06,
      "loss": 0.0812,
      "step": 8488
    },
    {
      "epoch": 2.1057,
      "grad_norm": 1.1659431457519531,
      "learning_rate": 7.631313131313132e-06,
      "loss": 0.4841,
      "step": 8489
    },
    {
      "epoch": 2.1058,
      "grad_norm": 5.002840995788574,
      "learning_rate": 7.6262626262626275e-06,
      "loss": 0.0547,
      "step": 8490
    },
    {
      "epoch": 2.1059,
      "grad_norm": 8.0558443069458,
      "learning_rate": 7.621212121212122e-06,
      "loss": 0.8203,
      "step": 8491
    },
    {
      "epoch": 2.106,
      "grad_norm": 1.988905906677246,
      "learning_rate": 7.616161616161617e-06,
      "loss": 0.4848,
      "step": 8492
    },
    {
      "epoch": 2.1061,
      "grad_norm": 0.34531256556510925,
      "learning_rate": 7.611111111111112e-06,
      "loss": 0.003,
      "step": 8493
    },
    {
      "epoch": 2.1062,
      "grad_norm": 2.2090890407562256,
      "learning_rate": 7.606060606060606e-06,
      "loss": 0.0581,
      "step": 8494
    },
    {
      "epoch": 2.1063,
      "grad_norm": 5.579421520233154,
      "learning_rate": 7.6010101010101016e-06,
      "loss": 0.198,
      "step": 8495
    },
    {
      "epoch": 2.1064,
      "grad_norm": 0.9743647575378418,
      "learning_rate": 7.595959595959597e-06,
      "loss": 0.0165,
      "step": 8496
    },
    {
      "epoch": 2.1065,
      "grad_norm": 2.060535192489624,
      "learning_rate": 7.590909090909092e-06,
      "loss": 0.0518,
      "step": 8497
    },
    {
      "epoch": 2.1066,
      "grad_norm": 7.384819984436035,
      "learning_rate": 7.585858585858586e-06,
      "loss": 0.4632,
      "step": 8498
    },
    {
      "epoch": 2.1067,
      "grad_norm": 3.179088592529297,
      "learning_rate": 7.580808080808081e-06,
      "loss": 0.5068,
      "step": 8499
    },
    {
      "epoch": 2.1068,
      "grad_norm": 6.807212829589844,
      "learning_rate": 7.5757575757575764e-06,
      "loss": 0.3942,
      "step": 8500
    }
  ],
  "logging_steps": 1,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}