{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9609375,
  "eval_steps": 500,
  "global_step": 340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01171875,
      "grad_norm": 34.936946868896484,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 4.5195,
      "step": 1
    },
    {
      "epoch": 0.0234375,
      "grad_norm": 35.17010498046875,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 4.5214,
      "step": 2
    },
    {
      "epoch": 0.03515625,
      "grad_norm": 36.44556427001953,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 4.5497,
      "step": 3
    },
    {
      "epoch": 0.046875,
      "grad_norm": 35.30234146118164,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 4.5195,
      "step": 4
    },
    {
      "epoch": 0.05859375,
      "grad_norm": 35.90467834472656,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 4.5572,
      "step": 5
    },
    {
      "epoch": 0.0703125,
      "grad_norm": 35.25331497192383,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 4.522,
      "step": 6
    },
    {
      "epoch": 0.08203125,
      "grad_norm": 35.77690124511719,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 4.532,
      "step": 7
    },
    {
      "epoch": 0.09375,
      "grad_norm": 35.59796905517578,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.5413,
      "step": 8
    },
    {
      "epoch": 0.10546875,
      "grad_norm": 34.550079345703125,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 4.444,
      "step": 9
    },
    {
      "epoch": 0.1171875,
      "grad_norm": 35.464324951171875,
      "learning_rate": 5.000000000000001e-07,
      "loss": 4.3859,
      "step": 10
    },
    {
      "epoch": 0.12890625,
      "grad_norm": 34.73582458496094,
      "learning_rate": 5.5e-07,
      "loss": 4.4206,
      "step": 11
    },
    {
      "epoch": 0.140625,
      "grad_norm": 34.19432067871094,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.3844,
      "step": 12
    },
    {
      "epoch": 0.15234375,
      "grad_norm": 33.5693473815918,
      "learning_rate": 6.5e-07,
      "loss": 4.3484,
      "step": 13
    },
    {
      "epoch": 0.1640625,
      "grad_norm": 33.903499603271484,
      "learning_rate": 7.000000000000001e-07,
      "loss": 4.3156,
      "step": 14
    },
    {
      "epoch": 0.17578125,
      "grad_norm": 31.61307716369629,
      "learning_rate": 7.5e-07,
      "loss": 4.1674,
      "step": 15
    },
    {
      "epoch": 0.1875,
      "grad_norm": 29.78888702392578,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.0061,
      "step": 16
    },
    {
      "epoch": 0.19921875,
      "grad_norm": 28.59884262084961,
      "learning_rate": 8.500000000000001e-07,
      "loss": 3.9706,
      "step": 17
    },
    {
      "epoch": 0.2109375,
      "grad_norm": 25.763689041137695,
      "learning_rate": 9.000000000000001e-07,
      "loss": 3.7961,
      "step": 18
    },
    {
      "epoch": 0.22265625,
      "grad_norm": 24.51258659362793,
      "learning_rate": 9.500000000000001e-07,
      "loss": 3.7835,
      "step": 19
    },
    {
      "epoch": 0.234375,
      "grad_norm": 22.484188079833984,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 3.5655,
      "step": 20
    },
    {
      "epoch": 0.24609375,
      "grad_norm": 20.98280143737793,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 3.483,
      "step": 21
    },
    {
      "epoch": 0.2578125,
      "grad_norm": 19.596670150756836,
      "learning_rate": 1.1e-06,
      "loss": 3.3479,
      "step": 22
    },
    {
      "epoch": 0.26953125,
      "grad_norm": 18.059001922607422,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 3.1264,
      "step": 23
    },
    {
      "epoch": 0.28125,
      "grad_norm": 17.812040328979492,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 3.1063,
      "step": 24
    },
    {
      "epoch": 0.29296875,
      "grad_norm": 18.554685592651367,
      "learning_rate": 1.25e-06,
      "loss": 2.9743,
      "step": 25
    },
    {
      "epoch": 0.3046875,
      "grad_norm": 19.290931701660156,
      "learning_rate": 1.3e-06,
      "loss": 2.7898,
      "step": 26
    },
    {
      "epoch": 0.31640625,
      "grad_norm": 20.526958465576172,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 2.6767,
      "step": 27
    },
    {
      "epoch": 0.328125,
      "grad_norm": 20.140962600708008,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.476,
      "step": 28
    },
    {
      "epoch": 0.33984375,
      "grad_norm": 18.80613136291504,
      "learning_rate": 1.45e-06,
      "loss": 2.4107,
      "step": 29
    },
    {
      "epoch": 0.3515625,
      "grad_norm": 16.77231788635254,
      "learning_rate": 1.5e-06,
      "loss": 2.2789,
      "step": 30
    },
    {
      "epoch": 0.36328125,
      "grad_norm": 15.546438217163086,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.118,
      "step": 31
    },
    {
      "epoch": 0.375,
      "grad_norm": 15.454617500305176,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.9487,
      "step": 32
    },
    {
      "epoch": 0.38671875,
      "grad_norm": 15.873698234558105,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 1.8438,
      "step": 33
    },
    {
      "epoch": 0.3984375,
      "grad_norm": 15.691689491271973,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.6935,
      "step": 34
    },
    {
      "epoch": 0.41015625,
      "grad_norm": 15.127069473266602,
      "learning_rate": 1.75e-06,
      "loss": 1.5188,
      "step": 35
    },
    {
      "epoch": 0.421875,
      "grad_norm": 14.441194534301758,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.345,
      "step": 36
    },
    {
      "epoch": 0.43359375,
      "grad_norm": 14.392739295959473,
      "learning_rate": 1.85e-06,
      "loss": 1.2395,
      "step": 37
    },
    {
      "epoch": 0.4453125,
      "grad_norm": 14.283534049987793,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.0954,
      "step": 38
    },
    {
      "epoch": 0.45703125,
      "grad_norm": 13.821670532226562,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.9594,
      "step": 39
    },
    {
      "epoch": 0.46875,
      "grad_norm": 13.498543739318848,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7972,
      "step": 40
    },
    {
      "epoch": 0.48046875,
      "grad_norm": 12.805685997009277,
      "learning_rate": 2.05e-06,
      "loss": 0.6578,
      "step": 41
    },
    {
      "epoch": 0.4921875,
      "grad_norm": 11.584136962890625,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.5148,
      "step": 42
    },
    {
      "epoch": 0.50390625,
      "grad_norm": 9.801186561584473,
      "learning_rate": 2.15e-06,
      "loss": 0.416,
      "step": 43
    },
    {
      "epoch": 0.515625,
      "grad_norm": 9.404278755187988,
      "learning_rate": 2.2e-06,
      "loss": 0.3419,
      "step": 44
    },
    {
      "epoch": 0.52734375,
      "grad_norm": 7.185790061950684,
      "learning_rate": 2.25e-06,
      "loss": 0.24,
      "step": 45
    },
    {
      "epoch": 0.5390625,
      "grad_norm": 4.915873050689697,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.1769,
      "step": 46
    },
    {
      "epoch": 0.55078125,
      "grad_norm": 2.8505806922912598,
      "learning_rate": 2.35e-06,
      "loss": 0.1435,
      "step": 47
    },
    {
      "epoch": 0.5625,
      "grad_norm": 1.7713755369186401,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.119,
      "step": 48
    },
    {
      "epoch": 0.57421875,
      "grad_norm": 1.5673104524612427,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.1064,
      "step": 49
    },
    {
      "epoch": 0.5859375,
      "grad_norm": 1.3322486877441406,
      "learning_rate": 2.5e-06,
      "loss": 0.0995,
      "step": 50
    },
    {
      "epoch": 0.59765625,
      "grad_norm": 1.2245675325393677,
      "learning_rate": 2.55e-06,
      "loss": 0.0889,
      "step": 51
    },
    {
      "epoch": 0.609375,
      "grad_norm": 0.8981238007545471,
      "learning_rate": 2.6e-06,
      "loss": 0.0847,
      "step": 52
    },
    {
      "epoch": 0.62109375,
      "grad_norm": 0.6747678518295288,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0731,
      "step": 53
    },
    {
      "epoch": 0.6328125,
      "grad_norm": 0.7690777778625488,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0812,
      "step": 54
    },
    {
      "epoch": 0.64453125,
      "grad_norm": 0.9659140706062317,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0827,
      "step": 55
    },
    {
      "epoch": 0.65625,
      "grad_norm": 0.7463306784629822,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0734,
      "step": 56
    },
    {
      "epoch": 0.66796875,
      "grad_norm": 0.6388681530952454,
      "learning_rate": 2.85e-06,
      "loss": 0.0741,
      "step": 57
    },
    {
      "epoch": 0.6796875,
      "grad_norm": 0.5176859498023987,
      "learning_rate": 2.9e-06,
      "loss": 0.0687,
      "step": 58
    },
    {
      "epoch": 0.69140625,
      "grad_norm": 0.5809365510940552,
      "learning_rate": 2.95e-06,
      "loss": 0.0759,
      "step": 59
    },
    {
      "epoch": 0.703125,
      "grad_norm": 0.4900883436203003,
      "learning_rate": 3e-06,
      "loss": 0.0637,
      "step": 60
    },
    {
      "epoch": 0.71484375,
      "grad_norm": 0.46506503224372864,
      "learning_rate": 3.05e-06,
      "loss": 0.066,
      "step": 61
    },
    {
      "epoch": 0.7265625,
      "grad_norm": 0.4315766394138336,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0619,
      "step": 62
    },
    {
      "epoch": 0.73828125,
      "grad_norm": 0.6561943888664246,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0718,
      "step": 63
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.44578853249549866,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0662,
      "step": 64
    },
    {
      "epoch": 0.76171875,
      "grad_norm": 0.352586567401886,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0612,
      "step": 65
    },
    {
      "epoch": 0.7734375,
      "grad_norm": 0.3779674172401428,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0696,
      "step": 66
    },
    {
      "epoch": 0.78515625,
      "grad_norm": 0.45150119066238403,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0665,
      "step": 67
    },
    {
      "epoch": 0.796875,
      "grad_norm": 0.32507383823394775,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0652,
      "step": 68
    },
    {
      "epoch": 0.80859375,
      "grad_norm": 0.3733697533607483,
      "learning_rate": 3.45e-06,
      "loss": 0.061,
      "step": 69
    },
    {
      "epoch": 0.8203125,
      "grad_norm": 0.3483797609806061,
      "learning_rate": 3.5e-06,
      "loss": 0.067,
      "step": 70
    },
    {
      "epoch": 0.83203125,
      "grad_norm": 0.2772752642631531,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0597,
      "step": 71
    },
    {
      "epoch": 0.84375,
      "grad_norm": 0.32659995555877686,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0657,
      "step": 72
    },
    {
      "epoch": 0.85546875,
      "grad_norm": 0.3197411000728607,
      "learning_rate": 3.65e-06,
      "loss": 0.0649,
      "step": 73
    },
    {
      "epoch": 0.8671875,
      "grad_norm": 0.36947256326675415,
      "learning_rate": 3.7e-06,
      "loss": 0.0666,
      "step": 74
    },
    {
      "epoch": 0.87890625,
      "grad_norm": 0.2516244351863861,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0618,
      "step": 75
    },
    {
      "epoch": 0.890625,
      "grad_norm": 0.27707308530807495,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0629,
      "step": 76
    },
    {
      "epoch": 0.90234375,
      "grad_norm": 0.31011566519737244,
      "learning_rate": 3.85e-06,
      "loss": 0.0639,
      "step": 77
    },
    {
      "epoch": 0.9140625,
      "grad_norm": 0.42964762449264526,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0575,
      "step": 78
    },
    {
      "epoch": 0.92578125,
      "grad_norm": 0.2871685326099396,
      "learning_rate": 3.95e-06,
      "loss": 0.056,
      "step": 79
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.2593545615673065,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0578,
      "step": 80
    },
    {
      "epoch": 0.94921875,
      "grad_norm": 0.27979350090026855,
      "learning_rate": 4.05e-06,
      "loss": 0.0582,
      "step": 81
    },
    {
      "epoch": 0.9609375,
      "grad_norm": 0.47774314880371094,
      "learning_rate": 4.1e-06,
      "loss": 0.0657,
      "step": 82
    },
    {
      "epoch": 0.97265625,
      "grad_norm": 0.46003803610801697,
      "learning_rate": 4.15e-06,
      "loss": 0.0591,
      "step": 83
    },
    {
      "epoch": 0.984375,
      "grad_norm": 0.28047817945480347,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0585,
      "step": 84
    },
    {
      "epoch": 0.99609375,
      "grad_norm": 0.3024386763572693,
      "learning_rate": 4.25e-06,
      "loss": 0.0597,
      "step": 85
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3024386763572693,
      "learning_rate": 4.3e-06,
      "loss": 0.0612,
      "step": 86
    },
    {
      "epoch": 1.01171875,
      "grad_norm": 1.1462472677230835,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0505,
      "step": 87
    },
    {
      "epoch": 1.0234375,
      "grad_norm": 0.28574419021606445,
      "learning_rate": 4.4e-06,
      "loss": 0.0575,
      "step": 88
    },
    {
      "epoch": 1.03515625,
      "grad_norm": 0.23733267188072205,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0567,
      "step": 89
    },
    {
      "epoch": 1.046875,
      "grad_norm": 0.25351643562316895,
      "learning_rate": 4.5e-06,
      "loss": 0.0595,
      "step": 90
    },
    {
      "epoch": 1.05859375,
      "grad_norm": 0.3286800980567932,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.056,
      "step": 91
    },
    {
      "epoch": 1.0703125,
      "grad_norm": 0.2708681523799896,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0541,
      "step": 92
    },
    {
      "epoch": 1.08203125,
      "grad_norm": 0.30789315700531006,
      "learning_rate": 4.65e-06,
      "loss": 0.0583,
      "step": 93
    },
    {
      "epoch": 1.09375,
      "grad_norm": 0.3800869882106781,
      "learning_rate": 4.7e-06,
      "loss": 0.059,
      "step": 94
    },
    {
      "epoch": 1.10546875,
      "grad_norm": 0.2756796181201935,
      "learning_rate": 4.75e-06,
      "loss": 0.0587,
      "step": 95
    },
    {
      "epoch": 1.1171875,
      "grad_norm": 0.29054075479507446,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0545,
      "step": 96
    },
    {
      "epoch": 1.12890625,
      "grad_norm": 0.24633410573005676,
      "learning_rate": 4.85e-06,
      "loss": 0.0576,
      "step": 97
    },
    {
      "epoch": 1.140625,
      "grad_norm": 0.23658645153045654,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0566,
      "step": 98
    },
    {
      "epoch": 1.15234375,
      "grad_norm": 0.2314174473285675,
      "learning_rate": 4.95e-06,
      "loss": 0.0535,
      "step": 99
    },
    {
      "epoch": 1.1640625,
      "grad_norm": 0.1932941973209381,
      "learning_rate": 5e-06,
      "loss": 0.0576,
      "step": 100
    },
    {
      "epoch": 1.17578125,
      "grad_norm": 0.22639156877994537,
      "learning_rate": 4.999926609487568e-06,
      "loss": 0.0519,
      "step": 101
    },
    {
      "epoch": 1.1875,
      "grad_norm": 0.23717834055423737,
      "learning_rate": 4.999706442259205e-06,
      "loss": 0.0487,
      "step": 102
    },
    {
      "epoch": 1.19921875,
      "grad_norm": 0.25690606236457825,
      "learning_rate": 4.999339511241458e-06,
      "loss": 0.0542,
      "step": 103
    },
    {
      "epoch": 1.2109375,
      "grad_norm": 0.28705868124961853,
      "learning_rate": 4.9988258379777334e-06,
      "loss": 0.0534,
      "step": 104
    },
    {
      "epoch": 1.22265625,
      "grad_norm": 0.24628855288028717,
      "learning_rate": 4.998165452627025e-06,
      "loss": 0.0536,
      "step": 105
    },
    {
      "epoch": 1.234375,
      "grad_norm": 0.38239291310310364,
      "learning_rate": 4.99735839396215e-06,
      "loss": 0.0545,
      "step": 106
    },
    {
      "epoch": 1.24609375,
      "grad_norm": 0.22199885547161102,
      "learning_rate": 4.996404709367466e-06,
      "loss": 0.0525,
      "step": 107
    },
    {
      "epoch": 1.2578125,
      "grad_norm": 0.2303021103143692,
      "learning_rate": 4.995304454836095e-06,
      "loss": 0.0544,
      "step": 108
    },
    {
      "epoch": 1.26953125,
      "grad_norm": 0.245191752910614,
      "learning_rate": 4.994057694966632e-06,
      "loss": 0.0521,
      "step": 109
    },
    {
      "epoch": 1.28125,
      "grad_norm": 0.267315149307251,
      "learning_rate": 4.992664502959351e-06,
      "loss": 0.0503,
      "step": 110
    },
    {
      "epoch": 1.29296875,
      "grad_norm": 0.23672722280025482,
      "learning_rate": 4.991124960611916e-06,
      "loss": 0.0471,
      "step": 111
    },
    {
      "epoch": 1.3046875,
      "grad_norm": 0.2500607371330261,
      "learning_rate": 4.989439158314566e-06,
      "loss": 0.0547,
      "step": 112
    },
    {
      "epoch": 1.31640625,
      "grad_norm": 0.27703896164894104,
      "learning_rate": 4.9876071950448185e-06,
      "loss": 0.0508,
      "step": 113
    },
    {
      "epoch": 1.328125,
      "grad_norm": 0.271287739276886,
      "learning_rate": 4.98562917836165e-06,
      "loss": 0.049,
      "step": 114
    },
    {
      "epoch": 1.33984375,
      "grad_norm": 0.25890663266181946,
      "learning_rate": 4.983505224399188e-06,
      "loss": 0.0509,
      "step": 115
    },
    {
      "epoch": 1.3515625,
      "grad_norm": 0.2870018780231476,
      "learning_rate": 4.9812354578598876e-06,
      "loss": 0.0518,
      "step": 116
    },
    {
      "epoch": 1.36328125,
      "grad_norm": 0.2725720703601837,
      "learning_rate": 4.978820012007213e-06,
      "loss": 0.047,
      "step": 117
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.19925664365291595,
      "learning_rate": 4.976259028657812e-06,
      "loss": 0.0493,
      "step": 118
    },
    {
      "epoch": 1.38671875,
      "grad_norm": 0.22143645584583282,
      "learning_rate": 4.973552658173186e-06,
      "loss": 0.0514,
      "step": 119
    },
    {
      "epoch": 1.3984375,
      "grad_norm": 0.22685761749744415,
      "learning_rate": 4.970701059450872e-06,
      "loss": 0.0525,
      "step": 120
    },
    {
      "epoch": 1.41015625,
      "grad_norm": 0.21121999621391296,
      "learning_rate": 4.9677043999151e-06,
      "loss": 0.051,
      "step": 121
    },
    {
      "epoch": 1.421875,
      "grad_norm": 0.23914238810539246,
      "learning_rate": 4.964562855506976e-06,
      "loss": 0.0469,
      "step": 122
    },
    {
      "epoch": 1.43359375,
      "grad_norm": 0.24409078061580658,
      "learning_rate": 4.961276610674141e-06,
      "loss": 0.0508,
      "step": 123
    },
    {
      "epoch": 1.4453125,
      "grad_norm": 0.23388482630252838,
      "learning_rate": 4.9578458583599495e-06,
      "loss": 0.051,
      "step": 124
    },
    {
      "epoch": 1.45703125,
      "grad_norm": 0.23801520466804504,
      "learning_rate": 4.954270799992138e-06,
      "loss": 0.0506,
      "step": 125
    },
    {
      "epoch": 1.46875,
      "grad_norm": 0.24522367119789124,
      "learning_rate": 4.950551645470998e-06,
      "loss": 0.0496,
      "step": 126
    },
    {
      "epoch": 1.48046875,
      "grad_norm": 0.22810965776443481,
      "learning_rate": 4.9466886131570565e-06,
      "loss": 0.0493,
      "step": 127
    },
    {
      "epoch": 1.4921875,
      "grad_norm": 0.21910196542739868,
      "learning_rate": 4.942681929858249e-06,
      "loss": 0.048,
      "step": 128
    },
    {
      "epoch": 1.50390625,
      "grad_norm": 0.22025763988494873,
      "learning_rate": 4.9385318308166065e-06,
      "loss": 0.0448,
      "step": 129
    },
    {
      "epoch": 1.515625,
      "grad_norm": 0.2766706049442291,
      "learning_rate": 4.934238559694448e-06,
      "loss": 0.0463,
      "step": 130
    },
    {
      "epoch": 1.52734375,
      "grad_norm": 0.22292804718017578,
      "learning_rate": 4.929802368560066e-06,
      "loss": 0.0433,
      "step": 131
    },
    {
      "epoch": 1.5390625,
      "grad_norm": 0.312200129032135,
      "learning_rate": 4.925223517872934e-06,
      "loss": 0.0477,
      "step": 132
    },
    {
      "epoch": 1.55078125,
      "grad_norm": 0.21522150933742523,
      "learning_rate": 4.920502276468408e-06,
      "loss": 0.0476,
      "step": 133
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.2599612772464752,
      "learning_rate": 4.915638921541952e-06,
      "loss": 0.0472,
      "step": 134
    },
    {
      "epoch": 1.57421875,
      "grad_norm": 0.2079230546951294,
      "learning_rate": 4.9106337386328524e-06,
      "loss": 0.0437,
      "step": 135
    },
    {
      "epoch": 1.5859375,
      "grad_norm": 0.23645609617233276,
      "learning_rate": 4.905487021607462e-06,
      "loss": 0.0418,
      "step": 136
    },
    {
      "epoch": 1.59765625,
      "grad_norm": 0.24720944464206696,
      "learning_rate": 4.900199072641937e-06,
      "loss": 0.0471,
      "step": 137
    },
    {
      "epoch": 1.609375,
      "grad_norm": 0.24982096254825592,
      "learning_rate": 4.894770202204509e-06,
      "loss": 0.0427,
      "step": 138
    },
    {
      "epoch": 1.62109375,
      "grad_norm": 0.2666509747505188,
      "learning_rate": 4.889200729037241e-06,
      "loss": 0.0457,
      "step": 139
    },
    {
      "epoch": 1.6328125,
      "grad_norm": 0.2664203345775604,
      "learning_rate": 4.883490980137327e-06,
      "loss": 0.0467,
      "step": 140
    },
    {
      "epoch": 1.64453125,
      "grad_norm": 0.250724732875824,
      "learning_rate": 4.8776412907378845e-06,
      "loss": 0.0389,
      "step": 141
    },
    {
      "epoch": 1.65625,
      "grad_norm": 0.3050763010978699,
      "learning_rate": 4.871652004288275e-06,
      "loss": 0.0447,
      "step": 142
    },
    {
      "epoch": 1.66796875,
      "grad_norm": 0.3108248710632324,
      "learning_rate": 4.865523472433942e-06,
      "loss": 0.0486,
      "step": 143
    },
    {
      "epoch": 1.6796875,
      "grad_norm": 0.24726933240890503,
      "learning_rate": 4.859256054995758e-06,
      "loss": 0.0425,
      "step": 144
    },
    {
      "epoch": 1.69140625,
      "grad_norm": 0.2529606819152832,
      "learning_rate": 4.8528501199489045e-06,
      "loss": 0.0468,
      "step": 145
    },
    {
      "epoch": 1.703125,
      "grad_norm": 0.2741091847419739,
      "learning_rate": 4.846306043401268e-06,
      "loss": 0.043,
      "step": 146
    },
    {
      "epoch": 1.71484375,
      "grad_norm": 0.22653311491012573,
      "learning_rate": 4.839624209571352e-06,
      "loss": 0.0436,
      "step": 147
    },
    {
      "epoch": 1.7265625,
      "grad_norm": 0.2258974015712738,
      "learning_rate": 4.832805010765724e-06,
      "loss": 0.0489,
      "step": 148
    },
    {
      "epoch": 1.73828125,
      "grad_norm": 0.2763097584247589,
      "learning_rate": 4.8258488473559794e-06,
      "loss": 0.0408,
      "step": 149
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.23488503694534302,
      "learning_rate": 4.8187561277552376e-06,
      "loss": 0.0396,
      "step": 150
    },
    {
      "epoch": 1.76171875,
      "grad_norm": 0.2815094590187073,
      "learning_rate": 4.811527268394157e-06,
      "loss": 0.0422,
      "step": 151
    },
    {
      "epoch": 1.7734375,
      "grad_norm": 0.2870430648326874,
      "learning_rate": 4.804162693696494e-06,
      "loss": 0.0417,
      "step": 152
    },
    {
      "epoch": 1.78515625,
      "grad_norm": 0.34226351976394653,
      "learning_rate": 4.796662836054176e-06,
      "loss": 0.0453,
      "step": 153
    },
    {
      "epoch": 1.796875,
      "grad_norm": 0.3671925961971283,
      "learning_rate": 4.789028135801919e-06,
      "loss": 0.043,
      "step": 154
    },
    {
      "epoch": 1.80859375,
      "grad_norm": 0.26639413833618164,
      "learning_rate": 4.7812590411913755e-06,
      "loss": 0.0437,
      "step": 155
    },
    {
      "epoch": 1.8203125,
      "grad_norm": 0.24375365674495697,
      "learning_rate": 4.773356008364812e-06,
      "loss": 0.0376,
      "step": 156
    },
    {
      "epoch": 1.83203125,
      "grad_norm": 0.23922622203826904,
      "learning_rate": 4.765319501328332e-06,
      "loss": 0.0459,
      "step": 157
    },
    {
      "epoch": 1.84375,
      "grad_norm": 0.21677494049072266,
      "learning_rate": 4.757149991924633e-06,
      "loss": 0.0429,
      "step": 158
    },
    {
      "epoch": 1.85546875,
      "grad_norm": 0.2717846632003784,
      "learning_rate": 4.748847959805297e-06,
      "loss": 0.0403,
      "step": 159
    },
    {
      "epoch": 1.8671875,
      "grad_norm": 0.23092707991600037,
      "learning_rate": 4.740413892402639e-06,
      "loss": 0.0417,
      "step": 160
    },
    {
      "epoch": 1.87890625,
      "grad_norm": 0.23728156089782715,
      "learning_rate": 4.731848284901082e-06,
      "loss": 0.0445,
      "step": 161
    },
    {
      "epoch": 1.890625,
      "grad_norm": 0.22474321722984314,
      "learning_rate": 4.723151640208084e-06,
      "loss": 0.0409,
      "step": 162
    },
    {
      "epoch": 1.90234375,
      "grad_norm": 0.22294431924819946,
      "learning_rate": 4.714324468924614e-06,
      "loss": 0.0387,
      "step": 163
    },
    {
      "epoch": 1.9140625,
      "grad_norm": 0.21790871024131775,
      "learning_rate": 4.705367289315172e-06,
      "loss": 0.0424,
      "step": 164
    },
    {
      "epoch": 1.92578125,
      "grad_norm": 0.24538278579711914,
      "learning_rate": 4.696280627277356e-06,
      "loss": 0.0399,
      "step": 165
    },
    {
      "epoch": 1.9375,
      "grad_norm": 0.2369818389415741,
      "learning_rate": 4.687065016310996e-06,
      "loss": 0.0386,
      "step": 166
    },
    {
      "epoch": 1.94921875,
      "grad_norm": 0.2826277017593384,
      "learning_rate": 4.6777209974868194e-06,
      "loss": 0.0404,
      "step": 167
    },
    {
      "epoch": 1.9609375,
      "grad_norm": 0.26429814100265503,
      "learning_rate": 4.668249119414692e-06,
      "loss": 0.0407,
      "step": 168
    },
    {
      "epoch": 1.97265625,
      "grad_norm": 0.2524208426475525,
      "learning_rate": 4.6586499382113985e-06,
      "loss": 0.037,
      "step": 169
    },
    {
      "epoch": 1.984375,
      "grad_norm": 0.22663331031799316,
      "learning_rate": 4.648924017468003e-06,
      "loss": 0.0337,
      "step": 170
    },
    {
      "epoch": 1.99609375,
      "grad_norm": 0.22759781777858734,
      "learning_rate": 4.6390719282167515e-06,
      "loss": 0.0359,
      "step": 171
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.22759781777858734,
      "learning_rate": 4.629094248897546e-06,
      "loss": 0.0577,
      "step": 172
    },
    {
      "epoch": 2.01171875,
      "grad_norm": 1.5158915519714355,
      "learning_rate": 4.618991565323987e-06,
      "loss": 0.0384,
      "step": 173
    },
    {
      "epoch": 2.0234375,
      "grad_norm": 0.27917784452438354,
      "learning_rate": 4.608764470648971e-06,
      "loss": 0.0321,
      "step": 174
    },
    {
      "epoch": 2.03515625,
      "grad_norm": 0.3492794632911682,
      "learning_rate": 4.598413565329876e-06,
      "loss": 0.0352,
      "step": 175
    },
    {
      "epoch": 2.046875,
      "grad_norm": 0.33379310369491577,
      "learning_rate": 4.587939457093296e-06,
      "loss": 0.0343,
      "step": 176
    },
    {
      "epoch": 2.05859375,
      "grad_norm": 0.3063555955886841,
      "learning_rate": 4.577342760899368e-06,
      "loss": 0.0336,
      "step": 177
    },
    {
      "epoch": 2.0703125,
      "grad_norm": 0.2801457345485687,
      "learning_rate": 4.566624098905665e-06,
      "loss": 0.0283,
      "step": 178
    },
    {
      "epoch": 2.08203125,
      "grad_norm": 0.2922702431678772,
      "learning_rate": 4.555784100430662e-06,
      "loss": 0.035,
      "step": 179
    },
    {
      "epoch": 2.09375,
      "grad_norm": 0.4648834764957428,
      "learning_rate": 4.544823401916794e-06,
      "loss": 0.0338,
      "step": 180
    },
    {
      "epoch": 2.10546875,
      "grad_norm": 0.25468185544013977,
      "learning_rate": 4.533742646893086e-06,
      "loss": 0.0264,
      "step": 181
    },
    {
      "epoch": 2.1171875,
      "grad_norm": 0.27900931239128113,
      "learning_rate": 4.522542485937369e-06,
      "loss": 0.0315,
      "step": 182
    },
    {
      "epoch": 2.12890625,
      "grad_norm": 0.25326547026634216,
      "learning_rate": 4.511223576638084e-06,
      "loss": 0.0274,
      "step": 183
    },
    {
      "epoch": 2.140625,
      "grad_norm": 0.24400267004966736,
      "learning_rate": 4.499786583555675e-06,
      "loss": 0.0276,
      "step": 184
    },
    {
      "epoch": 2.15234375,
      "grad_norm": 0.2513110637664795,
      "learning_rate": 4.4882321781835666e-06,
      "loss": 0.0328,
      "step": 185
    },
    {
      "epoch": 2.1640625,
      "grad_norm": 0.2765730619430542,
      "learning_rate": 4.476561038908745e-06,
      "loss": 0.0276,
      "step": 186
    },
    {
      "epoch": 2.17578125,
      "grad_norm": 0.2313976287841797,
      "learning_rate": 4.464773850971924e-06,
      "loss": 0.0271,
      "step": 187
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.23784014582633972,
      "learning_rate": 4.452871306427314e-06,
      "loss": 0.0339,
      "step": 188
    },
    {
      "epoch": 2.19921875,
      "grad_norm": 0.26526233553886414,
      "learning_rate": 4.440854104101988e-06,
      "loss": 0.0318,
      "step": 189
    },
    {
      "epoch": 2.2109375,
      "grad_norm": 0.27092573046684265,
      "learning_rate": 4.428722949554858e-06,
      "loss": 0.0295,
      "step": 190
    },
    {
      "epoch": 2.22265625,
      "grad_norm": 0.2370125949382782,
      "learning_rate": 4.416478555035241e-06,
      "loss": 0.0262,
      "step": 191
    },
    {
      "epoch": 2.234375,
      "grad_norm": 0.25447142124176025,
      "learning_rate": 4.404121639441047e-06,
      "loss": 0.0282,
      "step": 192
    },
    {
      "epoch": 2.24609375,
      "grad_norm": 0.3046242892742157,
      "learning_rate": 4.391652928276572e-06,
      "loss": 0.0296,
      "step": 193
    },
    {
      "epoch": 2.2578125,
      "grad_norm": 0.2992015480995178,
      "learning_rate": 4.379073153609896e-06,
      "loss": 0.0262,
      "step": 194
    },
    {
      "epoch": 2.26953125,
      "grad_norm": 0.2726775109767914,
      "learning_rate": 4.366383054029907e-06,
      "loss": 0.0247,
      "step": 195
    },
    {
      "epoch": 2.28125,
      "grad_norm": 0.2988302409648895,
      "learning_rate": 4.3535833746029335e-06,
      "loss": 0.03,
      "step": 196
    },
    {
      "epoch": 2.29296875,
      "grad_norm": 0.4626574218273163,
      "learning_rate": 4.340674866829001e-06,
      "loss": 0.0297,
      "step": 197
    },
    {
      "epoch": 2.3046875,
      "grad_norm": 0.26700106263160706,
      "learning_rate": 4.32765828859771e-06,
      "loss": 0.0282,
      "step": 198
    },
    {
      "epoch": 2.31640625,
      "grad_norm": 0.3087283968925476,
      "learning_rate": 4.314534404143738e-06,
      "loss": 0.0275,
      "step": 199
    },
    {
      "epoch": 2.328125,
      "grad_norm": 0.2655637264251709,
      "learning_rate": 4.3013039840019675e-06,
      "loss": 0.0289,
      "step": 200
    },
    {
      "epoch": 2.33984375,
      "grad_norm": 0.286205530166626,
      "learning_rate": 4.287967804962252e-06,
      "loss": 0.0287,
      "step": 201
    },
    {
      "epoch": 2.3515625,
      "grad_norm": 0.27568572759628296,
      "learning_rate": 4.274526650023801e-06,
      "loss": 0.0272,
      "step": 202
    },
    {
      "epoch": 2.36328125,
      "grad_norm": 0.2881161868572235,
      "learning_rate": 4.260981308349214e-06,
      "loss": 0.0286,
      "step": 203
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.2385960966348648,
      "learning_rate": 4.247332575218144e-06,
      "loss": 0.0227,
      "step": 204
    },
    {
      "epoch": 2.38671875,
      "grad_norm": 0.23624156415462494,
      "learning_rate": 4.233581251980604e-06,
      "loss": 0.0264,
      "step": 205
    },
    {
      "epoch": 2.3984375,
      "grad_norm": 0.24731102585792542,
      "learning_rate": 4.2197281460099245e-06,
      "loss": 0.0276,
      "step": 206
    },
    {
      "epoch": 2.41015625,
      "grad_norm": 0.2916982173919678,
      "learning_rate": 4.2057740706553415e-06,
      "loss": 0.0291,
      "step": 207
    },
    {
      "epoch": 2.421875,
      "grad_norm": 0.299566388130188,
      "learning_rate": 4.191719845194246e-06,
      "loss": 0.0237,
      "step": 208
    },
    {
      "epoch": 2.43359375,
      "grad_norm": 0.2720220685005188,
      "learning_rate": 4.177566294784085e-06,
      "loss": 0.0229,
      "step": 209
    },
    {
      "epoch": 2.4453125,
      "grad_norm": 0.24948938190937042,
      "learning_rate": 4.163314250413913e-06,
      "loss": 0.0227,
      "step": 210
    },
    {
      "epoch": 2.45703125,
      "grad_norm": 0.2563379406929016,
      "learning_rate": 4.148964548855603e-06,
      "loss": 0.022,
      "step": 211
    },
    {
      "epoch": 2.46875,
      "grad_norm": 0.285737544298172,
      "learning_rate": 4.134518032614713e-06,
      "loss": 0.026,
      "step": 212
    },
    {
      "epoch": 2.48046875,
      "grad_norm": 0.2400374859571457,
      "learning_rate": 4.119975549881029e-06,
      "loss": 0.0213,
      "step": 213
    },
    {
      "epoch": 2.4921875,
      "grad_norm": 0.30329281091690063,
      "learning_rate": 4.105337954478756e-06,
      "loss": 0.0218,
      "step": 214
    },
    {
      "epoch": 2.50390625,
      "grad_norm": 0.2331625074148178,
      "learning_rate": 4.0906061058164e-06,
      "loss": 0.0206,
      "step": 215
    },
    {
      "epoch": 2.515625,
      "grad_norm": 0.2650165557861328,
      "learning_rate": 4.075780868836296e-06,
      "loss": 0.0185,
      "step": 216
    },
    {
      "epoch": 2.52734375,
      "grad_norm": 0.27158695459365845,
      "learning_rate": 4.060863113963835e-06,
      "loss": 0.0216,
      "step": 217
    },
    {
      "epoch": 2.5390625,
      "grad_norm": 0.3077850043773651,
      "learning_rate": 4.045853717056358e-06,
      "loss": 0.0176,
      "step": 218
    },
    {
      "epoch": 2.55078125,
      "grad_norm": 0.4170614778995514,
      "learning_rate": 4.030753559351728e-06,
      "loss": 0.0177,
      "step": 219
    },
    {
      "epoch": 2.5625,
      "grad_norm": 0.3055117428302765,
      "learning_rate": 4.015563527416596e-06,
      "loss": 0.0199,
      "step": 220
    },
    {
      "epoch": 2.57421875,
      "grad_norm": 0.28157854080200195,
      "learning_rate": 4.000284513094342e-06,
      "loss": 0.0233,
      "step": 221
    },
    {
      "epoch": 2.5859375,
      "grad_norm": 0.27278608083724976,
      "learning_rate": 3.984917413452721e-06,
      "loss": 0.0222,
      "step": 222
    },
    {
      "epoch": 2.59765625,
      "grad_norm": 0.3474902808666229,
      "learning_rate": 3.969463130731183e-06,
      "loss": 0.0199,
      "step": 223
    },
    {
      "epoch": 2.609375,
      "grad_norm": 0.2641908824443817,
      "learning_rate": 3.953922572287915e-06,
      "loss": 0.0198,
      "step": 224
    },
    {
      "epoch": 2.62109375,
      "grad_norm": 0.26810458302497864,
      "learning_rate": 3.938296650546552e-06,
      "loss": 0.0212,
      "step": 225
    },
    {
      "epoch": 2.6328125,
      "grad_norm": 0.29748621582984924,
      "learning_rate": 3.9225862829426184e-06,
      "loss": 0.0167,
      "step": 226
    },
    {
      "epoch": 2.64453125,
      "grad_norm": 0.3216516077518463,
      "learning_rate": 3.906792391869657e-06,
      "loss": 0.0198,
      "step": 227
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.3365074396133423,
      "learning_rate": 3.890915904625075e-06,
      "loss": 0.0243,
      "step": 228
    },
    {
      "epoch": 2.66796875,
      "grad_norm": 0.3966096341609955,
      "learning_rate": 3.874957753355701e-06,
      "loss": 0.0201,
      "step": 229
    },
    {
      "epoch": 2.6796875,
      "grad_norm": 0.3851388394832611,
      "learning_rate": 3.858918875003053e-06,
      "loss": 0.0197,
      "step": 230
    },
    {
      "epoch": 2.69140625,
      "grad_norm": 0.2769019305706024,
      "learning_rate": 3.842800211248333e-06,
      "loss": 0.0156,
      "step": 231
    },
    {
      "epoch": 2.703125,
      "grad_norm": 0.27606338262557983,
      "learning_rate": 3.8266027084571335e-06,
      "loss": 0.0215,
      "step": 232
    },
    {
      "epoch": 2.71484375,
      "grad_norm": 0.2841218411922455,
      "learning_rate": 3.810327317623881e-06,
      "loss": 0.014,
      "step": 233
    },
    {
      "epoch": 2.7265625,
      "grad_norm": 0.2940453290939331,
      "learning_rate": 3.793974994315991e-06,
      "loss": 0.021,
      "step": 234
    },
    {
      "epoch": 2.73828125,
      "grad_norm": 0.26162609457969666,
      "learning_rate": 3.7775466986177763e-06,
      "loss": 0.0186,
      "step": 235
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.26991286873817444,
      "learning_rate": 3.7610433950740667e-06,
      "loss": 0.0231,
      "step": 236
    },
    {
      "epoch": 2.76171875,
      "grad_norm": 0.33351415395736694,
      "learning_rate": 3.7444660526335853e-06,
      "loss": 0.0138,
      "step": 237
    },
    {
      "epoch": 2.7734375,
      "grad_norm": 0.2652147114276886,
      "learning_rate": 3.7278156445920584e-06,
      "loss": 0.0174,
      "step": 238
    },
    {
      "epoch": 2.78515625,
      "grad_norm": 0.3918457627296448,
      "learning_rate": 3.711093148535068e-06,
      "loss": 0.0183,
      "step": 239
    },
    {
      "epoch": 2.796875,
      "grad_norm": 0.3130027949810028,
      "learning_rate": 3.6942995462806574e-06,
      "loss": 0.0153,
      "step": 240
    },
    {
      "epoch": 2.80859375,
      "grad_norm": 0.2556680142879486,
      "learning_rate": 3.6774358238216878e-06,
      "loss": 0.0142,
      "step": 241
    },
    {
      "epoch": 2.8203125,
      "grad_norm": 0.26903846859931946,
      "learning_rate": 3.660502971267945e-06,
      "loss": 0.0162,
      "step": 242
    },
    {
      "epoch": 2.83203125,
      "grad_norm": 0.2997351884841919,
      "learning_rate": 3.6435019827880093e-06,
      "loss": 0.0199,
      "step": 243
    },
    {
      "epoch": 2.84375,
      "grad_norm": 0.31552162766456604,
      "learning_rate": 3.626433856550886e-06,
      "loss": 0.016,
      "step": 244
    },
    {
      "epoch": 2.85546875,
      "grad_norm": 0.29252275824546814,
      "learning_rate": 3.6092995946673996e-06,
      "loss": 0.0156,
      "step": 245
    },
    {
      "epoch": 2.8671875,
      "grad_norm": 0.3055340349674225,
      "learning_rate": 3.5921002031313586e-06,
      "loss": 0.0214,
      "step": 246
    },
    {
      "epoch": 2.87890625,
      "grad_norm": 0.2900487780570984,
      "learning_rate": 3.574836691760489e-06,
      "loss": 0.0107,
      "step": 247
    },
    {
      "epoch": 2.890625,
      "grad_norm": 0.334410697221756,
      "learning_rate": 3.557510074137147e-06,
      "loss": 0.0163,
      "step": 248
    },
    {
      "epoch": 2.90234375,
      "grad_norm": 0.27932974696159363,
      "learning_rate": 3.540121367548811e-06,
      "loss": 0.0123,
      "step": 249
    },
    {
      "epoch": 2.9140625,
      "grad_norm": 0.25670287013053894,
      "learning_rate": 3.5226715929283507e-06,
      "loss": 0.0156,
      "step": 250
    },
    {
      "epoch": 2.92578125,
      "grad_norm": 0.3766632080078125,
      "learning_rate": 3.505161774794085e-06,
      "loss": 0.0104,
      "step": 251
    },
    {
      "epoch": 2.9375,
      "grad_norm": 0.353557288646698,
      "learning_rate": 3.487592941189636e-06,
      "loss": 0.0164,
      "step": 252
    },
    {
      "epoch": 2.94921875,
      "grad_norm": 0.2867110073566437,
      "learning_rate": 3.469966123623563e-06,
      "loss": 0.0112,
      "step": 253
    },
    {
      "epoch": 2.9609375,
      "grad_norm": 0.28520146012306213,
      "learning_rate": 3.4522823570088073e-06,
      "loss": 0.0146,
      "step": 254
    },
    {
      "epoch": 2.97265625,
      "grad_norm": 0.4236989915370941,
      "learning_rate": 3.434542679601922e-06,
      "loss": 0.013,
      "step": 255
    },
    {
      "epoch": 2.984375,
      "grad_norm": 0.24684889614582062,
      "learning_rate": 3.4167481329421204e-06,
      "loss": 0.0115,
      "step": 256
    },
    {
      "epoch": 2.99609375,
      "grad_norm": 0.27294921875,
      "learning_rate": 3.39889976179012e-06,
      "loss": 0.0156,
      "step": 257
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8994725942611694,
      "learning_rate": 3.380998614066805e-06,
      "loss": 0.015,
      "step": 258
    },
    {
      "epoch": 3.01171875,
      "grad_norm": 0.2117411494255066,
      "learning_rate": 3.363045740791698e-06,
      "loss": 0.0092,
      "step": 259
    },
    {
      "epoch": 3.0234375,
      "grad_norm": 0.19839300215244293,
      "learning_rate": 3.345042196021257e-06,
      "loss": 0.0064,
      "step": 260
    },
    {
      "epoch": 3.03515625,
      "grad_norm": 0.20257411897182465,
      "learning_rate": 3.326989036786981e-06,
      "loss": 0.0069,
      "step": 261
    },
    {
      "epoch": 3.046875,
      "grad_norm": 0.347488671541214,
      "learning_rate": 3.3088873230333562e-06,
      "loss": 0.0102,
      "step": 262
    },
    {
      "epoch": 3.05859375,
      "grad_norm": 0.43107420206069946,
      "learning_rate": 3.290738117555622e-06,
      "loss": 0.0092,
      "step": 263
    },
    {
      "epoch": 3.0703125,
      "grad_norm": 0.2854469418525696,
      "learning_rate": 3.272542485937369e-06,
      "loss": 0.0069,
      "step": 264
    },
    {
      "epoch": 3.08203125,
      "grad_norm": 0.3907378315925598,
      "learning_rate": 3.2543014964879814e-06,
      "loss": 0.0111,
      "step": 265
    },
    {
      "epoch": 3.09375,
      "grad_norm": 0.3163127303123474,
      "learning_rate": 3.2360162201799085e-06,
      "loss": 0.0123,
      "step": 266
    },
    {
      "epoch": 3.10546875,
      "grad_norm": 0.2967950999736786,
      "learning_rate": 3.21768773058579e-06,
      "loss": 0.0071,
      "step": 267
    },
    {
      "epoch": 3.1171875,
      "grad_norm": 0.22201873362064362,
      "learning_rate": 3.1993171038154203e-06,
      "loss": 0.009,
      "step": 268
    },
    {
      "epoch": 3.12890625,
      "grad_norm": 0.30213239789009094,
      "learning_rate": 3.180905418452569e-06,
      "loss": 0.0063,
      "step": 269
    },
    {
      "epoch": 3.140625,
      "grad_norm": 0.24428071081638336,
      "learning_rate": 3.162453755491655e-06,
      "loss": 0.0083,
      "step": 270
    },
    {
      "epoch": 3.15234375,
      "grad_norm": 0.18968743085861206,
      "learning_rate": 3.143963198274278e-06,
      "loss": 0.0087,
      "step": 271
    },
    {
      "epoch": 3.1640625,
      "grad_norm": 0.2680652141571045,
      "learning_rate": 3.125434832425613e-06,
      "loss": 0.0085,
      "step": 272
    },
    {
      "epoch": 3.17578125,
      "grad_norm": 0.250074565410614,
      "learning_rate": 3.1068697457906736e-06,
      "loss": 0.0074,
      "step": 273
    },
    {
      "epoch": 3.1875,
      "grad_norm": 0.19228896498680115,
      "learning_rate": 3.0882690283704355e-06,
      "loss": 0.0041,
      "step": 274
    },
    {
      "epoch": 3.19921875,
      "grad_norm": 0.19371174275875092,
      "learning_rate": 3.0696337722578444e-06,
      "loss": 0.0056,
      "step": 275
    },
    {
      "epoch": 3.2109375,
      "grad_norm": 0.3111586272716522,
      "learning_rate": 3.0509650715736977e-06,
      "loss": 0.0086,
      "step": 276
    },
    {
      "epoch": 3.22265625,
      "grad_norm": 0.31564322113990784,
      "learning_rate": 3.0322640224024024e-06,
      "loss": 0.0061,
      "step": 277
    },
    {
      "epoch": 3.234375,
      "grad_norm": 0.36250177025794983,
      "learning_rate": 3.0135317227276247e-06,
      "loss": 0.0089,
      "step": 278
    },
    {
      "epoch": 3.24609375,
      "grad_norm": 0.2839025855064392,
      "learning_rate": 2.994769272367822e-06,
      "loss": 0.0071,
      "step": 279
    },
    {
      "epoch": 3.2578125,
      "grad_norm": 0.30359339714050293,
      "learning_rate": 2.975977772911671e-06,
      "loss": 0.0065,
      "step": 280
    },
    {
      "epoch": 3.26953125,
      "grad_norm": 0.34792155027389526,
      "learning_rate": 2.9571583276533923e-06,
      "loss": 0.0053,
      "step": 281
    },
    {
      "epoch": 3.28125,
      "grad_norm": 0.3163566291332245,
      "learning_rate": 2.93831204152797e-06,
      "loss": 0.0103,
      "step": 282
    },
    {
      "epoch": 3.29296875,
      "grad_norm": 0.2554697096347809,
      "learning_rate": 2.9194400210462808e-06,
      "loss": 0.009,
      "step": 283
    },
    {
      "epoch": 3.3046875,
      "grad_norm": 0.2524597942829132,
      "learning_rate": 2.9005433742301274e-06,
      "loss": 0.0053,
      "step": 284
    },
    {
      "epoch": 3.31640625,
      "grad_norm": 0.27872219681739807,
      "learning_rate": 2.8816232105471864e-06,
      "loss": 0.0066,
      "step": 285
    },
    {
      "epoch": 3.328125,
      "grad_norm": 0.24400565028190613,
      "learning_rate": 2.8626806408458626e-06,
      "loss": 0.0067,
      "step": 286
    },
    {
      "epoch": 3.33984375,
      "grad_norm": 0.20240619778633118,
      "learning_rate": 2.843716777290074e-06,
      "loss": 0.0033,
      "step": 287
    },
    {
      "epoch": 3.3515625,
      "grad_norm": 0.18623533844947815,
      "learning_rate": 2.8247327332939512e-06,
      "loss": 0.0047,
      "step": 288
    },
    {
      "epoch": 3.36328125,
      "grad_norm": 0.219360813498497,
      "learning_rate": 2.805729623456469e-06,
      "loss": 0.0049,
      "step": 289
    },
    {
      "epoch": 3.375,
      "grad_norm": 0.2884356379508972,
      "learning_rate": 2.786708563496002e-06,
      "loss": 0.0047,
      "step": 290
    },
    {
      "epoch": 3.38671875,
      "grad_norm": 0.20841263234615326,
      "learning_rate": 2.7676706701848187e-06,
      "loss": 0.005,
      "step": 291
    },
    {
      "epoch": 3.3984375,
      "grad_norm": 0.2048211395740509,
      "learning_rate": 2.748617061283518e-06,
      "loss": 0.0049,
      "step": 292
    },
    {
      "epoch": 3.41015625,
      "grad_norm": 0.18965432047843933,
      "learning_rate": 2.7295488554753957e-06,
      "loss": 0.0044,
      "step": 293
    },
    {
      "epoch": 3.421875,
      "grad_norm": 0.4232686758041382,
      "learning_rate": 2.710467172300768e-06,
      "loss": 0.012,
      "step": 294
    },
    {
      "epoch": 3.43359375,
      "grad_norm": 0.5638787746429443,
      "learning_rate": 2.69137313209124e-06,
      "loss": 0.0035,
      "step": 295
    },
    {
      "epoch": 3.4453125,
      "grad_norm": 0.2795218825340271,
      "learning_rate": 2.672267855903927e-06,
      "loss": 0.0043,
      "step": 296
    },
    {
      "epoch": 3.45703125,
      "grad_norm": 0.2081572711467743,
      "learning_rate": 2.653152465455639e-06,
      "loss": 0.0051,
      "step": 297
    },
    {
      "epoch": 3.46875,
      "grad_norm": 0.22985902428627014,
      "learning_rate": 2.6340280830570142e-06,
      "loss": 0.0048,
      "step": 298
    },
    {
      "epoch": 3.48046875,
      "grad_norm": 0.2268473356962204,
      "learning_rate": 2.614895831546633e-06,
      "loss": 0.0056,
      "step": 299
    },
    {
      "epoch": 3.4921875,
      "grad_norm": 0.19264478981494904,
      "learning_rate": 2.595756834225089e-06,
      "loss": 0.0041,
      "step": 300
    },
    {
      "epoch": 3.50390625,
      "grad_norm": 0.26570984721183777,
      "learning_rate": 2.576612214789039e-06,
      "loss": 0.0067,
      "step": 301
    },
    {
      "epoch": 3.515625,
      "grad_norm": 0.1802910566329956,
      "learning_rate": 2.5574630972652263e-06,
      "loss": 0.0049,
      "step": 302
    },
    {
      "epoch": 3.52734375,
      "grad_norm": 0.1650409996509552,
      "learning_rate": 2.538310605944491e-06,
      "loss": 0.0044,
      "step": 303
    },
    {
      "epoch": 3.5390625,
      "grad_norm": 0.16457340121269226,
      "learning_rate": 2.5191558653157542e-06,
      "loss": 0.0028,
      "step": 304
    },
    {
      "epoch": 3.55078125,
      "grad_norm": 0.20663093030452728,
      "learning_rate": 2.5e-06,
      "loss": 0.0047,
      "step": 305
    },
    {
      "epoch": 3.5625,
      "grad_norm": 0.17303715646266937,
      "learning_rate": 2.480844134684246e-06,
      "loss": 0.005,
      "step": 306
    },
    {
      "epoch": 3.57421875,
      "grad_norm": 0.2537713348865509,
      "learning_rate": 2.4616893940555094e-06,
      "loss": 0.0076,
      "step": 307
    },
    {
      "epoch": 3.5859375,
      "grad_norm": 0.28013190627098083,
      "learning_rate": 2.4425369027347746e-06,
      "loss": 0.0052,
      "step": 308
    },
    {
      "epoch": 3.59765625,
      "grad_norm": 0.16120599210262299,
      "learning_rate": 2.423387785210962e-06,
      "loss": 0.0027,
      "step": 309
    },
    {
      "epoch": 3.609375,
      "grad_norm": 0.246893510222435,
      "learning_rate": 2.404243165774912e-06,
      "loss": 0.0042,
      "step": 310
    },
    {
      "epoch": 3.62109375,
      "grad_norm": 0.31807446479797363,
      "learning_rate": 2.3851041684533677e-06,
      "loss": 0.0046,
      "step": 311
    },
    {
      "epoch": 3.6328125,
      "grad_norm": 0.41118142008781433,
      "learning_rate": 2.3659719169429866e-06,
      "loss": 0.0036,
      "step": 312
    },
    {
      "epoch": 3.64453125,
      "grad_norm": 0.28581708669662476,
      "learning_rate": 2.346847534544362e-06,
      "loss": 0.0059,
      "step": 313
    },
    {
      "epoch": 3.65625,
      "grad_norm": 0.23896564543247223,
      "learning_rate": 2.3277321440960733e-06,
      "loss": 0.004,
      "step": 314
    },
    {
      "epoch": 3.66796875,
      "grad_norm": 0.3185984194278717,
      "learning_rate": 2.308626867908761e-06,
      "loss": 0.0064,
      "step": 315
    },
    {
      "epoch": 3.6796875,
      "grad_norm": 0.19877824187278748,
      "learning_rate": 2.2895328276992325e-06,
      "loss": 0.0047,
      "step": 316
    },
    {
      "epoch": 3.69140625,
      "grad_norm": 0.2414735108613968,
      "learning_rate": 2.270451144524605e-06,
      "loss": 0.0034,
      "step": 317
    },
    {
      "epoch": 3.703125,
      "grad_norm": 0.1455293744802475,
      "learning_rate": 2.251382938716482e-06,
      "loss": 0.0015,
      "step": 318
    },
    {
      "epoch": 3.71484375,
      "grad_norm": 0.1986742913722992,
      "learning_rate": 2.2323293298151817e-06,
      "loss": 0.0037,
      "step": 319
    },
    {
      "epoch": 3.7265625,
      "grad_norm": 0.2751258909702301,
      "learning_rate": 2.2132914365039993e-06,
      "loss": 0.0037,
      "step": 320
    },
    {
      "epoch": 3.73828125,
      "grad_norm": 0.18424548208713531,
      "learning_rate": 2.1942703765435317e-06,
      "loss": 0.0041,
      "step": 321
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.29543158411979675,
      "learning_rate": 2.1752672667060488e-06,
      "loss": 0.0069,
      "step": 322
    },
    {
      "epoch": 3.76171875,
      "grad_norm": 0.24201937019824982,
      "learning_rate": 2.1562832227099266e-06,
      "loss": 0.0043,
      "step": 323
    },
    {
      "epoch": 3.7734375,
      "grad_norm": 0.2203102856874466,
      "learning_rate": 2.137319359154138e-06,
      "loss": 0.0042,
      "step": 324
    },
    {
      "epoch": 3.78515625,
      "grad_norm": 0.17772577702999115,
      "learning_rate": 2.1183767894528135e-06,
      "loss": 0.003,
      "step": 325
    },
    {
      "epoch": 3.796875,
      "grad_norm": 0.16460777819156647,
      "learning_rate": 2.099456625769872e-06,
      "loss": 0.0023,
      "step": 326
    },
    {
      "epoch": 3.80859375,
      "grad_norm": 0.13881157338619232,
      "learning_rate": 2.08055997895372e-06,
      "loss": 0.003,
      "step": 327
    },
    {
      "epoch": 3.8203125,
      "grad_norm": 0.14287902414798737,
      "learning_rate": 2.0616879584720305e-06,
      "loss": 0.0028,
      "step": 328
    },
    {
      "epoch": 3.83203125,
      "grad_norm": 0.2369135171175003,
      "learning_rate": 2.042841672346608e-06,
      "loss": 0.0035,
      "step": 329
    },
    {
      "epoch": 3.84375,
      "grad_norm": 0.1415957510471344,
      "learning_rate": 2.024022227088329e-06,
      "loss": 0.0023,
      "step": 330
    },
    {
      "epoch": 3.85546875,
      "grad_norm": 0.14889901876449585,
      "learning_rate": 2.0052307276321793e-06,
      "loss": 0.003,
      "step": 331
    },
    {
      "epoch": 3.8671875,
      "grad_norm": 0.17904798686504364,
      "learning_rate": 1.9864682772723757e-06,
      "loss": 0.0042,
      "step": 332
    },
    {
      "epoch": 3.87890625,
      "grad_norm": 0.16420456767082214,
      "learning_rate": 1.967735977597598e-06,
      "loss": 0.0024,
      "step": 333
    },
    {
      "epoch": 3.890625,
      "grad_norm": 0.22299158573150635,
      "learning_rate": 1.9490349284263036e-06,
      "loss": 0.0028,
      "step": 334
    },
    {
      "epoch": 3.90234375,
      "grad_norm": 0.17052407562732697,
      "learning_rate": 1.930366227742157e-06,
      "loss": 0.0024,
      "step": 335
    },
    {
      "epoch": 3.9140625,
      "grad_norm": 0.12893564999103546,
      "learning_rate": 1.9117309716295658e-06,
      "loss": 0.0016,
      "step": 336
    },
    {
      "epoch": 3.92578125,
      "grad_norm": 0.21460793912410736,
      "learning_rate": 1.8931302542093274e-06,
      "loss": 0.0032,
      "step": 337
    },
    {
      "epoch": 3.9375,
      "grad_norm": 2.8890089988708496,
      "learning_rate": 1.8745651675743876e-06,
      "loss": 0.0024,
      "step": 338
    },
    {
      "epoch": 3.94921875,
      "grad_norm": 0.2956985533237457,
      "learning_rate": 1.8560368017257229e-06,
      "loss": 0.0042,
      "step": 339
    },
    {
      "epoch": 3.9609375,
      "grad_norm": 0.1472138911485672,
      "learning_rate": 1.8375462445083464e-06,
      "loss": 0.0019,
      "step": 340
    }
  ],
  "logging_steps": 1,
  "max_steps": 510,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 85,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.554882281186263e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}