{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.8720930232558137,
  "eval_steps": 500,
  "global_step": 112,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03488372093023256,
      "grad_norm": 43.01280212402344,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 2.7496,
      "step": 1
    },
    {
      "epoch": 0.06976744186046512,
      "grad_norm": 44.429630279541016,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 2.7809,
      "step": 2
    },
    {
      "epoch": 0.10465116279069768,
      "grad_norm": 42.754234313964844,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 2.6968,
      "step": 3
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 45.27891159057617,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.8076,
      "step": 4
    },
    {
      "epoch": 0.1744186046511628,
      "grad_norm": 43.920860290527344,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 2.7255,
      "step": 5
    },
    {
      "epoch": 0.20930232558139536,
      "grad_norm": 44.61790466308594,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 2.7178,
      "step": 6
    },
    {
      "epoch": 0.2441860465116279,
      "grad_norm": 44.28205871582031,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 2.7193,
      "step": 7
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 45.1516227722168,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.6846,
      "step": 8
    },
    {
      "epoch": 0.313953488372093,
      "grad_norm": 46.8723258972168,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 2.67,
      "step": 9
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 48.33848571777344,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.6368,
      "step": 10
    },
    {
      "epoch": 0.38372093023255816,
      "grad_norm": 46.772193908691406,
      "learning_rate": 5.5e-07,
      "loss": 2.5562,
      "step": 11
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 48.12417984008789,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.5362,
      "step": 12
    },
    {
      "epoch": 0.45348837209302323,
      "grad_norm": 49.0787239074707,
      "learning_rate": 6.5e-07,
      "loss": 2.3981,
      "step": 13
    },
    {
      "epoch": 0.4883720930232558,
      "grad_norm": 52.367183685302734,
      "learning_rate": 7.000000000000001e-07,
      "loss": 2.3837,
      "step": 14
    },
    {
      "epoch": 0.5232558139534884,
      "grad_norm": 53.57649230957031,
      "learning_rate": 7.5e-07,
      "loss": 2.2885,
      "step": 15
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 56.58591079711914,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.1447,
      "step": 16
    },
    {
      "epoch": 0.5930232558139535,
      "grad_norm": 55.064735412597656,
      "learning_rate": 8.500000000000001e-07,
      "loss": 2.0242,
      "step": 17
    },
    {
      "epoch": 0.627906976744186,
      "grad_norm": 52.61149597167969,
      "learning_rate": 9.000000000000001e-07,
      "loss": 1.9263,
      "step": 18
    },
    {
      "epoch": 0.6627906976744186,
      "grad_norm": 44.34925079345703,
      "learning_rate": 9.500000000000001e-07,
      "loss": 1.7466,
      "step": 19
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 36.627296447753906,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.6084,
      "step": 20
    },
    {
      "epoch": 0.7325581395348837,
      "grad_norm": 30.89563751220703,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 1.4701,
      "step": 21
    },
    {
      "epoch": 0.7674418604651163,
      "grad_norm": 29.17167091369629,
      "learning_rate": 1.1e-06,
      "loss": 1.4218,
      "step": 22
    },
    {
      "epoch": 0.8023255813953488,
      "grad_norm": 28.237022399902344,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 1.3014,
      "step": 23
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 28.778654098510742,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.1857,
      "step": 24
    },
    {
      "epoch": 0.872093023255814,
      "grad_norm": 29.368289947509766,
      "learning_rate": 1.25e-06,
      "loss": 1.082,
      "step": 25
    },
    {
      "epoch": 0.9069767441860465,
      "grad_norm": 28.46448516845703,
      "learning_rate": 1.3e-06,
      "loss": 0.8716,
      "step": 26
    },
    {
      "epoch": 0.9418604651162791,
      "grad_norm": 24.584318161010742,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.6848,
      "step": 27
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 21.166847229003906,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.5438,
      "step": 28
    },
    {
      "epoch": 1.0,
      "grad_norm": 21.166847229003906,
      "learning_rate": 1.45e-06,
      "loss": 0.4084,
      "step": 29
    },
    {
      "epoch": 1.0348837209302326,
      "grad_norm": 24.098440170288086,
      "learning_rate": 1.5e-06,
      "loss": 0.3545,
      "step": 30
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 12.6813325881958,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.2673,
      "step": 31
    },
    {
      "epoch": 1.1046511627906976,
      "grad_norm": 7.6524128913879395,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.2033,
      "step": 32
    },
    {
      "epoch": 1.1395348837209303,
      "grad_norm": 5.23082160949707,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.1887,
      "step": 33
    },
    {
      "epoch": 1.1744186046511629,
      "grad_norm": 4.729929447174072,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.1683,
      "step": 34
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 5.831579208374023,
      "learning_rate": 1.75e-06,
      "loss": 0.1643,
      "step": 35
    },
    {
      "epoch": 1.244186046511628,
      "grad_norm": 4.030057430267334,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.1528,
      "step": 36
    },
    {
      "epoch": 1.2790697674418605,
      "grad_norm": 3.560523509979248,
      "learning_rate": 1.85e-06,
      "loss": 0.1422,
      "step": 37
    },
    {
      "epoch": 1.3139534883720931,
      "grad_norm": 3.3749780654907227,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.1327,
      "step": 38
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 2.7184131145477295,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.1199,
      "step": 39
    },
    {
      "epoch": 1.3837209302325582,
      "grad_norm": 2.8681583404541016,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.1107,
      "step": 40
    },
    {
      "epoch": 1.4186046511627908,
      "grad_norm": 2.8731987476348877,
      "learning_rate": 2.05e-06,
      "loss": 0.1063,
      "step": 41
    },
    {
      "epoch": 1.4534883720930232,
      "grad_norm": 2.9522054195404053,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.0989,
      "step": 42
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 2.689038038253784,
      "learning_rate": 2.15e-06,
      "loss": 0.0949,
      "step": 43
    },
    {
      "epoch": 1.5232558139534884,
      "grad_norm": 2.6328952312469482,
      "learning_rate": 2.2e-06,
      "loss": 0.1006,
      "step": 44
    },
    {
      "epoch": 1.558139534883721,
      "grad_norm": 2.7056033611297607,
      "learning_rate": 2.25e-06,
      "loss": 0.0929,
      "step": 45
    },
    {
      "epoch": 1.5930232558139537,
      "grad_norm": 2.568206310272217,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.0881,
      "step": 46
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 2.714211940765381,
      "learning_rate": 2.35e-06,
      "loss": 0.096,
      "step": 47
    },
    {
      "epoch": 1.6627906976744184,
      "grad_norm": 2.3650710582733154,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.086,
      "step": 48
    },
    {
      "epoch": 1.697674418604651,
      "grad_norm": 2.1763498783111572,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0858,
      "step": 49
    },
    {
      "epoch": 1.7325581395348837,
      "grad_norm": 2.124727487564087,
      "learning_rate": 2.5e-06,
      "loss": 0.0843,
      "step": 50
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 1.641269564628601,
      "learning_rate": 2.55e-06,
      "loss": 0.0803,
      "step": 51
    },
    {
      "epoch": 1.802325581395349,
      "grad_norm": 1.5644842386245728,
      "learning_rate": 2.6e-06,
      "loss": 0.0779,
      "step": 52
    },
    {
      "epoch": 1.8372093023255816,
      "grad_norm": 1.1404681205749512,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0765,
      "step": 53
    },
    {
      "epoch": 1.872093023255814,
      "grad_norm": 0.8386123776435852,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0749,
      "step": 54
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 1.8363338708877563,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0737,
      "step": 55
    },
    {
      "epoch": 1.941860465116279,
      "grad_norm": 1.2105377912521362,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0762,
      "step": 56
    },
    {
      "epoch": 1.9767441860465116,
      "grad_norm": 1.5502218008041382,
      "learning_rate": 2.85e-06,
      "loss": 0.07,
      "step": 57
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0641449689865112,
      "learning_rate": 2.9e-06,
      "loss": 0.0679,
      "step": 58
    },
    {
      "epoch": 2.0348837209302326,
      "grad_norm": 0.9201306104660034,
      "learning_rate": 2.95e-06,
      "loss": 0.0677,
      "step": 59
    },
    {
      "epoch": 2.0697674418604652,
      "grad_norm": 0.5951386094093323,
      "learning_rate": 3e-06,
      "loss": 0.0654,
      "step": 60
    },
    {
      "epoch": 2.104651162790698,
      "grad_norm": 0.8307608962059021,
      "learning_rate": 3.05e-06,
      "loss": 0.0706,
      "step": 61
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 0.7110892534255981,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0665,
      "step": 62
    },
    {
      "epoch": 2.1744186046511627,
      "grad_norm": 0.6766234040260315,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0618,
      "step": 63
    },
    {
      "epoch": 2.2093023255813953,
      "grad_norm": 0.3967410922050476,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0644,
      "step": 64
    },
    {
      "epoch": 2.244186046511628,
      "grad_norm": 0.3713420331478119,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0607,
      "step": 65
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.5613359212875366,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0623,
      "step": 66
    },
    {
      "epoch": 2.313953488372093,
      "grad_norm": 0.5458635687828064,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0563,
      "step": 67
    },
    {
      "epoch": 2.3488372093023258,
      "grad_norm": 0.4196176528930664,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0547,
      "step": 68
    },
    {
      "epoch": 2.383720930232558,
      "grad_norm": 0.5757117867469788,
      "learning_rate": 3.45e-06,
      "loss": 0.0551,
      "step": 69
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.45777687430381775,
      "learning_rate": 3.5e-06,
      "loss": 0.0575,
      "step": 70
    },
    {
      "epoch": 2.453488372093023,
      "grad_norm": 0.5204553008079529,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.057,
      "step": 71
    },
    {
      "epoch": 2.488372093023256,
      "grad_norm": 0.6110821962356567,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0556,
      "step": 72
    },
    {
      "epoch": 2.5232558139534884,
      "grad_norm": 0.45246991515159607,
      "learning_rate": 3.65e-06,
      "loss": 0.0564,
      "step": 73
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.441976398229599,
      "learning_rate": 3.7e-06,
      "loss": 0.0534,
      "step": 74
    },
    {
      "epoch": 2.5930232558139537,
      "grad_norm": 0.46637651324272156,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0498,
      "step": 75
    },
    {
      "epoch": 2.6279069767441863,
      "grad_norm": 0.482038289308548,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0521,
      "step": 76
    },
    {
      "epoch": 2.6627906976744184,
      "grad_norm": 0.5914385318756104,
      "learning_rate": 3.85e-06,
      "loss": 0.0511,
      "step": 77
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.34532907605171204,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0527,
      "step": 78
    },
    {
      "epoch": 2.7325581395348837,
      "grad_norm": 0.35315006971359253,
      "learning_rate": 3.95e-06,
      "loss": 0.0486,
      "step": 79
    },
    {
      "epoch": 2.7674418604651163,
      "grad_norm": 0.4521324634552002,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0485,
      "step": 80
    },
    {
      "epoch": 2.802325581395349,
      "grad_norm": 0.49457868933677673,
      "learning_rate": 4.05e-06,
      "loss": 0.0525,
      "step": 81
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.7616601586341858,
      "learning_rate": 4.1e-06,
      "loss": 0.0528,
      "step": 82
    },
    {
      "epoch": 2.8720930232558137,
      "grad_norm": 0.4791123569011688,
      "learning_rate": 4.15e-06,
      "loss": 0.0512,
      "step": 83
    },
    {
      "epoch": 2.9069767441860463,
      "grad_norm": 0.38186997175216675,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0442,
      "step": 84
    },
    {
      "epoch": 2.941860465116279,
      "grad_norm": 0.32648414373397827,
      "learning_rate": 4.25e-06,
      "loss": 0.0432,
      "step": 85
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.410888671875,
      "learning_rate": 4.3e-06,
      "loss": 0.042,
      "step": 86
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4732547104358673,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0483,
      "step": 87
    },
    {
      "epoch": 3.0348837209302326,
      "grad_norm": 0.6372231841087341,
      "learning_rate": 4.4e-06,
      "loss": 0.0382,
      "step": 88
    },
    {
      "epoch": 3.0697674418604652,
      "grad_norm": 0.393078088760376,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0403,
      "step": 89
    },
    {
      "epoch": 3.104651162790698,
      "grad_norm": 0.4366248846054077,
      "learning_rate": 4.5e-06,
      "loss": 0.0367,
      "step": 90
    },
    {
      "epoch": 3.13953488372093,
      "grad_norm": 0.301724374294281,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0351,
      "step": 91
    },
    {
      "epoch": 3.1744186046511627,
      "grad_norm": 0.37665656208992004,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.03,
      "step": 92
    },
    {
      "epoch": 3.2093023255813953,
      "grad_norm": 0.581331729888916,
      "learning_rate": 4.65e-06,
      "loss": 0.0316,
      "step": 93
    },
    {
      "epoch": 3.244186046511628,
      "grad_norm": 0.44136878848075867,
      "learning_rate": 4.7e-06,
      "loss": 0.0312,
      "step": 94
    },
    {
      "epoch": 3.2790697674418605,
      "grad_norm": 0.7624006271362305,
      "learning_rate": 4.75e-06,
      "loss": 0.0323,
      "step": 95
    },
    {
      "epoch": 3.313953488372093,
      "grad_norm": 0.35294386744499207,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0286,
      "step": 96
    },
    {
      "epoch": 3.3488372093023258,
      "grad_norm": 0.6240035891532898,
      "learning_rate": 4.85e-06,
      "loss": 0.0289,
      "step": 97
    },
    {
      "epoch": 3.383720930232558,
      "grad_norm": 0.40580570697784424,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0259,
      "step": 98
    },
    {
      "epoch": 3.4186046511627906,
      "grad_norm": 0.2971636652946472,
      "learning_rate": 4.95e-06,
      "loss": 0.0251,
      "step": 99
    },
    {
      "epoch": 3.453488372093023,
      "grad_norm": 0.3758476972579956,
      "learning_rate": 5e-06,
      "loss": 0.0244,
      "step": 100
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.3845921754837036,
      "learning_rate": 4.997332437005932e-06,
      "loss": 0.0242,
      "step": 101
    },
    {
      "epoch": 3.5232558139534884,
      "grad_norm": 0.48894616961479187,
      "learning_rate": 4.989335440737587e-06,
      "loss": 0.0252,
      "step": 102
    },
    {
      "epoch": 3.558139534883721,
      "grad_norm": 0.36844325065612793,
      "learning_rate": 4.976026077188013e-06,
      "loss": 0.0197,
      "step": 103
    },
    {
      "epoch": 3.5930232558139537,
      "grad_norm": 0.2800253629684448,
      "learning_rate": 4.957432749209755e-06,
      "loss": 0.0187,
      "step": 104
    },
    {
      "epoch": 3.6279069767441863,
      "grad_norm": 0.34042924642562866,
      "learning_rate": 4.933595135901733e-06,
      "loss": 0.0182,
      "step": 105
    },
    {
      "epoch": 3.6627906976744184,
      "grad_norm": 0.33217477798461914,
      "learning_rate": 4.904564107932048e-06,
      "loss": 0.0166,
      "step": 106
    },
    {
      "epoch": 3.697674418604651,
      "grad_norm": 0.32867398858070374,
      "learning_rate": 4.870401618977415e-06,
      "loss": 0.017,
      "step": 107
    },
    {
      "epoch": 3.7325581395348837,
      "grad_norm": 0.24199356138706207,
      "learning_rate": 4.83118057351089e-06,
      "loss": 0.0138,
      "step": 108
    },
    {
      "epoch": 3.7674418604651163,
      "grad_norm": 0.3216392695903778,
      "learning_rate": 4.786984671220053e-06,
      "loss": 0.0171,
      "step": 109
    },
    {
      "epoch": 3.802325581395349,
      "grad_norm": 0.3574189841747284,
      "learning_rate": 4.737908228387656e-06,
      "loss": 0.0146,
      "step": 110
    },
    {
      "epoch": 3.8372093023255816,
      "grad_norm": 0.30931738018989563,
      "learning_rate": 4.684055976615924e-06,
      "loss": 0.0128,
      "step": 111
    },
    {
      "epoch": 3.8720930232558137,
      "grad_norm": 0.2074785828590393,
      "learning_rate": 4.625542839324036e-06,
      "loss": 0.0089,
      "step": 112
    }
  ],
  "logging_steps": 1,
  "max_steps": 168,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 28,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.521111556499177e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}