{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.8727272727272726,
  "eval_steps": 6,
  "global_step": 96,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04155844155844156,
      "grad_norm": 21.0,
      "learning_rate": 2.5e-06,
      "loss": 3.3357,
      "step": 1
    },
    {
      "epoch": 0.04155844155844156,
      "eval_loss": 4.24916934967041,
      "eval_runtime": 139.9881,
      "eval_samples_per_second": 1.586,
      "eval_steps_per_second": 1.586,
      "step": 1
    },
    {
      "epoch": 0.08311688311688312,
      "grad_norm": 20.625,
      "learning_rate": 5e-06,
      "loss": 3.2927,
      "step": 2
    },
    {
      "epoch": 0.12467532467532468,
      "grad_norm": 19.125,
      "learning_rate": 7.500000000000001e-06,
      "loss": 3.3517,
      "step": 3
    },
    {
      "epoch": 0.16623376623376623,
      "grad_norm": 10.0,
      "learning_rate": 1e-05,
      "loss": 3.1756,
      "step": 4
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 11.0,
      "learning_rate": 9.99708511183087e-06,
      "loss": 3.0295,
      "step": 5
    },
    {
      "epoch": 0.24935064935064935,
      "grad_norm": 7.34375,
      "learning_rate": 9.988343845952697e-06,
      "loss": 2.9892,
      "step": 6
    },
    {
      "epoch": 0.24935064935064935,
      "eval_loss": 3.628525733947754,
      "eval_runtime": 135.8923,
      "eval_samples_per_second": 1.634,
      "eval_steps_per_second": 1.634,
      "step": 6
    },
    {
      "epoch": 0.2909090909090909,
      "grad_norm": 5.34375,
      "learning_rate": 9.973786394290475e-06,
      "loss": 3.0567,
      "step": 7
    },
    {
      "epoch": 0.33246753246753247,
      "grad_norm": 3.9375,
      "learning_rate": 9.953429730181653e-06,
      "loss": 2.9789,
      "step": 8
    },
    {
      "epoch": 0.37402597402597404,
      "grad_norm": 3.109375,
      "learning_rate": 9.927297588585984e-06,
      "loss": 2.9467,
      "step": 9
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 2.875,
      "learning_rate": 9.895420438411616e-06,
      "loss": 2.8889,
      "step": 10
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 2.8125,
      "learning_rate": 9.857835446989708e-06,
      "loss": 2.9016,
      "step": 11
    },
    {
      "epoch": 0.4987012987012987,
      "grad_norm": 2.671875,
      "learning_rate": 9.814586436738998e-06,
      "loss": 2.7364,
      "step": 12
    },
    {
      "epoch": 0.4987012987012987,
      "eval_loss": 3.4675309658050537,
      "eval_runtime": 135.783,
      "eval_samples_per_second": 1.635,
      "eval_steps_per_second": 1.635,
      "step": 12
    },
    {
      "epoch": 0.5402597402597402,
      "grad_norm": 2.53125,
      "learning_rate": 9.765723834070805e-06,
      "loss": 2.7602,
      "step": 13
    },
    {
      "epoch": 0.5818181818181818,
      "grad_norm": 2.546875,
      "learning_rate": 9.711304610594104e-06,
      "loss": 2.8655,
      "step": 14
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 2.515625,
      "learning_rate": 9.651392216689167e-06,
      "loss": 2.7857,
      "step": 15
    },
    {
      "epoch": 0.6649350649350649,
      "grad_norm": 2.359375,
      "learning_rate": 9.586056507527266e-06,
      "loss": 2.6186,
      "step": 16
    },
    {
      "epoch": 0.7064935064935065,
      "grad_norm": 2.453125,
      "learning_rate": 9.515373661622665e-06,
      "loss": 2.6869,
      "step": 17
    },
    {
      "epoch": 0.7480519480519481,
      "grad_norm": 2.265625,
      "learning_rate": 9.439426092011877e-06,
      "loss": 2.7076,
      "step": 18
    },
    {
      "epoch": 0.7480519480519481,
      "eval_loss": 3.3928399085998535,
      "eval_runtime": 135.8425,
      "eval_samples_per_second": 1.634,
      "eval_steps_per_second": 1.634,
      "step": 18
    },
    {
      "epoch": 0.7896103896103897,
      "grad_norm": 2.40625,
      "learning_rate": 9.358302350163758e-06,
      "loss": 2.6393,
      "step": 19
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 2.234375,
      "learning_rate": 9.272097022732444e-06,
      "loss": 2.5202,
      "step": 20
    },
    {
      "epoch": 0.8727272727272727,
      "grad_norm": 2.25,
      "learning_rate": 9.180910621273555e-06,
      "loss": 2.8021,
      "step": 21
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 2.171875,
      "learning_rate": 9.08484946505221e-06,
      "loss": 2.679,
      "step": 22
    },
    {
      "epoch": 0.9558441558441558,
      "grad_norm": 2.21875,
      "learning_rate": 8.984025557079523e-06,
      "loss": 2.684,
      "step": 23
    },
    {
      "epoch": 0.9974025974025974,
      "grad_norm": 2.359375,
      "learning_rate": 8.8785564535221e-06,
      "loss": 2.757,
      "step": 24
    },
    {
      "epoch": 0.9974025974025974,
      "eval_loss": 3.348358631134033,
      "eval_runtime": 135.8159,
      "eval_samples_per_second": 1.635,
      "eval_steps_per_second": 1.635,
      "step": 24
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.4375,
      "learning_rate": 8.768565126636806e-06,
      "loss": 1.9296,
      "step": 25
    },
    {
      "epoch": 1.0415584415584416,
      "grad_norm": 4.34375,
      "learning_rate": 8.65417982139062e-06,
      "loss": 3.5077,
      "step": 26
    },
    {
      "epoch": 1.0831168831168831,
      "grad_norm": 2.296875,
      "learning_rate": 8.535533905932739e-06,
      "loss": 2.6161,
      "step": 27
    },
    {
      "epoch": 1.1246753246753247,
      "grad_norm": 2.1875,
      "learning_rate": 8.412765716093273e-06,
      "loss": 2.635,
      "step": 28
    },
    {
      "epoch": 1.1662337662337663,
      "grad_norm": 2.1875,
      "learning_rate": 8.286018394089864e-06,
      "loss": 2.626,
      "step": 29
    },
    {
      "epoch": 1.2077922077922079,
      "grad_norm": 2.125,
      "learning_rate": 8.155439721630265e-06,
      "loss": 2.5801,
      "step": 30
    },
    {
      "epoch": 1.2077922077922079,
      "eval_loss": 3.3286170959472656,
      "eval_runtime": 135.6246,
      "eval_samples_per_second": 1.637,
      "eval_steps_per_second": 1.637,
      "step": 30
    },
    {
      "epoch": 1.2493506493506494,
      "grad_norm": 2.109375,
      "learning_rate": 8.021181947605474e-06,
      "loss": 2.5571,
      "step": 31
    },
    {
      "epoch": 1.290909090909091,
      "grad_norm": 2.21875,
      "learning_rate": 7.883401610574338e-06,
      "loss": 2.629,
      "step": 32
    },
    {
      "epoch": 1.3324675324675326,
      "grad_norm": 2.078125,
      "learning_rate": 7.742259356246594e-06,
      "loss": 2.6276,
      "step": 33
    },
    {
      "epoch": 1.3740259740259742,
      "grad_norm": 2.109375,
      "learning_rate": 7.597919750177168e-06,
      "loss": 2.534,
      "step": 34
    },
    {
      "epoch": 1.4155844155844157,
      "grad_norm": 2.03125,
      "learning_rate": 7.450551085890087e-06,
      "loss": 2.6451,
      "step": 35
    },
    {
      "epoch": 1.457142857142857,
      "grad_norm": 2.234375,
      "learning_rate": 7.300325188655762e-06,
      "loss": 2.6156,
      "step": 36
    },
    {
      "epoch": 1.457142857142857,
      "eval_loss": 3.311096668243408,
      "eval_runtime": 135.7523,
      "eval_samples_per_second": 1.635,
      "eval_steps_per_second": 1.635,
      "step": 36
    },
    {
      "epoch": 1.4987012987012986,
      "grad_norm": 2.1875,
      "learning_rate": 7.147417215150411e-06,
      "loss": 2.5858,
      "step": 37
    },
    {
      "epoch": 1.5402597402597402,
      "grad_norm": 2.125,
      "learning_rate": 6.9920054492312086e-06,
      "loss": 2.4837,
      "step": 38
    },
    {
      "epoch": 1.5818181818181818,
      "grad_norm": 2.140625,
      "learning_rate": 6.834271094065284e-06,
      "loss": 2.5288,
      "step": 39
    },
    {
      "epoch": 1.6233766233766234,
      "grad_norm": 2.03125,
      "learning_rate": 6.674398060854931e-06,
      "loss": 2.5408,
      "step": 40
    },
    {
      "epoch": 1.664935064935065,
      "grad_norm": 2.078125,
      "learning_rate": 6.51257275440538e-06,
      "loss": 2.5974,
      "step": 41
    },
    {
      "epoch": 1.7064935064935065,
      "grad_norm": 2.109375,
      "learning_rate": 6.348983855785122e-06,
      "loss": 2.5308,
      "step": 42
    },
    {
      "epoch": 1.7064935064935065,
      "eval_loss": 3.2999422550201416,
      "eval_runtime": 135.66,
      "eval_samples_per_second": 1.636,
      "eval_steps_per_second": 1.636,
      "step": 42
    },
    {
      "epoch": 1.748051948051948,
      "grad_norm": 2.0625,
      "learning_rate": 6.183822102332234e-06,
      "loss": 2.4401,
      "step": 43
    },
    {
      "epoch": 1.7896103896103897,
      "grad_norm": 2.078125,
      "learning_rate": 6.0172800652631706e-06,
      "loss": 2.6279,
      "step": 44
    },
    {
      "epoch": 1.8311688311688312,
      "grad_norm": 2.109375,
      "learning_rate": 5.849551925143334e-06,
      "loss": 2.6906,
      "step": 45
    },
    {
      "epoch": 1.8727272727272726,
      "grad_norm": 2.171875,
      "learning_rate": 5.680833245481234e-06,
      "loss": 2.5528,
      "step": 46
    },
    {
      "epoch": 1.9142857142857141,
      "grad_norm": 2.171875,
      "learning_rate": 5.511320744710171e-06,
      "loss": 2.4684,
      "step": 47
    },
    {
      "epoch": 1.9558441558441557,
      "grad_norm": 2.03125,
      "learning_rate": 5.341212066823356e-06,
      "loss": 2.5481,
      "step": 48
    },
    {
      "epoch": 1.9558441558441557,
      "eval_loss": 3.2879936695098877,
      "eval_runtime": 135.9375,
      "eval_samples_per_second": 1.633,
      "eval_steps_per_second": 1.633,
      "step": 48
    },
    {
      "epoch": 1.9974025974025973,
      "grad_norm": 2.171875,
      "learning_rate": 5.17070555092984e-06,
      "loss": 2.4671,
      "step": 49
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.6875,
      "learning_rate": 5e-06,
      "loss": 1.5953,
      "step": 50
    },
    {
      "epoch": 2.0415584415584416,
      "grad_norm": 4.375,
      "learning_rate": 4.829294449070161e-06,
      "loss": 3.6037,
      "step": 51
    },
    {
      "epoch": 2.083116883116883,
      "grad_norm": 2.203125,
      "learning_rate": 4.6587879331766465e-06,
      "loss": 2.6118,
      "step": 52
    },
    {
      "epoch": 2.1246753246753247,
      "grad_norm": 2.125,
      "learning_rate": 4.488679255289829e-06,
      "loss": 2.4673,
      "step": 53
    },
    {
      "epoch": 2.1662337662337663,
      "grad_norm": 2.28125,
      "learning_rate": 4.319166754518768e-06,
      "loss": 2.5773,
      "step": 54
    },
    {
      "epoch": 2.1662337662337663,
      "eval_loss": 3.2839770317077637,
      "eval_runtime": 135.829,
      "eval_samples_per_second": 1.634,
      "eval_steps_per_second": 1.634,
      "step": 54
    },
    {
      "epoch": 2.207792207792208,
      "grad_norm": 2.046875,
      "learning_rate": 4.150448074856667e-06,
      "loss": 2.6372,
      "step": 55
    },
    {
      "epoch": 2.2493506493506494,
      "grad_norm": 1.96875,
      "learning_rate": 3.982719934736832e-06,
      "loss": 2.4622,
      "step": 56
    },
    {
      "epoch": 2.290909090909091,
      "grad_norm": 2.0625,
      "learning_rate": 3.816177897667767e-06,
      "loss": 2.578,
      "step": 57
    },
    {
      "epoch": 2.3324675324675326,
      "grad_norm": 2.0625,
      "learning_rate": 3.6510161442148783e-06,
      "loss": 2.5275,
      "step": 58
    },
    {
      "epoch": 2.374025974025974,
      "grad_norm": 2.15625,
      "learning_rate": 3.4874272455946217e-06,
      "loss": 2.5072,
      "step": 59
    },
    {
      "epoch": 2.4155844155844157,
      "grad_norm": 2.046875,
      "learning_rate": 3.3256019391450696e-06,
      "loss": 2.5269,
      "step": 60
    },
    {
      "epoch": 2.4155844155844157,
      "eval_loss": 3.2822012901306152,
      "eval_runtime": 136.0741,
      "eval_samples_per_second": 1.631,
      "eval_steps_per_second": 1.631,
      "step": 60
    },
    {
      "epoch": 2.4571428571428573,
      "grad_norm": 2.03125,
      "learning_rate": 3.1657289059347184e-06,
      "loss": 2.4748,
      "step": 61
    },
    {
      "epoch": 2.498701298701299,
      "grad_norm": 2.078125,
      "learning_rate": 3.007994550768793e-06,
      "loss": 2.5465,
      "step": 62
    },
    {
      "epoch": 2.5402597402597404,
      "grad_norm": 2.046875,
      "learning_rate": 2.8525827848495912e-06,
      "loss": 2.5334,
      "step": 63
    },
    {
      "epoch": 2.581818181818182,
      "grad_norm": 2.0625,
      "learning_rate": 2.6996748113442397e-06,
      "loss": 2.4747,
      "step": 64
    },
    {
      "epoch": 2.6233766233766236,
      "grad_norm": 1.9609375,
      "learning_rate": 2.5494489141099155e-06,
      "loss": 2.4145,
      "step": 65
    },
    {
      "epoch": 2.664935064935065,
      "grad_norm": 2.109375,
      "learning_rate": 2.4020802498228333e-06,
      "loss": 2.5418,
      "step": 66
    },
    {
      "epoch": 2.664935064935065,
      "eval_loss": 3.280576467514038,
      "eval_runtime": 135.9902,
      "eval_samples_per_second": 1.632,
      "eval_steps_per_second": 1.632,
      "step": 66
    },
    {
      "epoch": 2.7064935064935067,
      "grad_norm": 2.015625,
      "learning_rate": 2.2577406437534055e-06,
      "loss": 2.387,
      "step": 67
    },
    {
      "epoch": 2.7480519480519483,
      "grad_norm": 2.0625,
      "learning_rate": 2.1165983894256647e-06,
      "loss": 2.5978,
      "step": 68
    },
    {
      "epoch": 2.78961038961039,
      "grad_norm": 1.9921875,
      "learning_rate": 1.978818052394528e-06,
      "loss": 2.4508,
      "step": 69
    },
    {
      "epoch": 2.8311688311688314,
      "grad_norm": 1.96875,
      "learning_rate": 1.8445602783697375e-06,
      "loss": 2.5458,
      "step": 70
    },
    {
      "epoch": 2.8727272727272726,
      "grad_norm": 2.015625,
      "learning_rate": 1.7139816059101372e-06,
      "loss": 2.6041,
      "step": 71
    },
    {
      "epoch": 2.914285714285714,
      "grad_norm": 1.9453125,
      "learning_rate": 1.5872342839067305e-06,
      "loss": 2.4584,
      "step": 72
    },
    {
      "epoch": 2.914285714285714,
      "eval_loss": 3.2790873050689697,
      "eval_runtime": 135.938,
      "eval_samples_per_second": 1.633,
      "eval_steps_per_second": 1.633,
      "step": 72
    },
    {
      "epoch": 2.9558441558441557,
      "grad_norm": 2.0625,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 2.5471,
      "step": 73
    },
    {
      "epoch": 2.9974025974025973,
      "grad_norm": 1.96875,
      "learning_rate": 1.3458201786093795e-06,
      "loss": 2.455,
      "step": 74
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.71875,
      "learning_rate": 1.2314348733631958e-06,
      "loss": 1.7987,
      "step": 75
    },
    {
      "epoch": 3.0415584415584416,
      "grad_norm": 3.8125,
      "learning_rate": 1.1214435464779006e-06,
      "loss": 2.9485,
      "step": 76
    },
    {
      "epoch": 3.083116883116883,
      "grad_norm": 2.046875,
      "learning_rate": 1.0159744429204776e-06,
      "loss": 2.4781,
      "step": 77
    },
    {
      "epoch": 3.1246753246753247,
      "grad_norm": 2.0,
      "learning_rate": 9.151505349477901e-07,
      "loss": 2.6515,
      "step": 78
    },
    {
      "epoch": 3.1246753246753247,
      "eval_loss": 3.278913736343384,
      "eval_runtime": 135.8528,
      "eval_samples_per_second": 1.634,
      "eval_steps_per_second": 1.634,
      "step": 78
    },
    {
      "epoch": 3.1662337662337663,
      "grad_norm": 2.03125,
      "learning_rate": 8.19089378726447e-07,
      "loss": 2.5251,
      "step": 79
    },
    {
      "epoch": 3.207792207792208,
      "grad_norm": 1.9609375,
      "learning_rate": 7.279029772675572e-07,
      "loss": 2.5715,
      "step": 80
    },
    {
      "epoch": 3.2493506493506494,
      "grad_norm": 1.96875,
      "learning_rate": 6.416976498362432e-07,
      "loss": 2.5349,
      "step": 81
    },
    {
      "epoch": 3.290909090909091,
      "grad_norm": 1.953125,
      "learning_rate": 5.60573907988124e-07,
      "loss": 2.5464,
      "step": 82
    },
    {
      "epoch": 3.3324675324675326,
      "grad_norm": 1.921875,
      "learning_rate": 4.846263383773364e-07,
      "loss": 2.4217,
      "step": 83
    },
    {
      "epoch": 3.374025974025974,
      "grad_norm": 2.046875,
      "learning_rate": 4.139434924727359e-07,
      "loss": 2.4883,
      "step": 84
    },
    {
      "epoch": 3.374025974025974,
      "eval_loss": 3.278531551361084,
      "eval_runtime": 135.9158,
      "eval_samples_per_second": 1.633,
      "eval_steps_per_second": 1.633,
      "step": 84
    },
    {
      "epoch": 3.4155844155844157,
      "grad_norm": 2.03125,
      "learning_rate": 3.486077833108342e-07,
      "loss": 2.5032,
      "step": 85
    },
    {
      "epoch": 3.4571428571428573,
      "grad_norm": 2.03125,
      "learning_rate": 2.88695389405898e-07,
      "loss": 2.5042,
      "step": 86
    },
    {
      "epoch": 3.498701298701299,
      "grad_norm": 1.9453125,
      "learning_rate": 2.3427616592919587e-07,
      "loss": 2.4433,
      "step": 87
    },
    {
      "epoch": 3.5402597402597404,
      "grad_norm": 1.890625,
      "learning_rate": 1.8541356326100436e-07,
      "loss": 2.4906,
      "step": 88
    },
    {
      "epoch": 3.581818181818182,
      "grad_norm": 1.9296875,
      "learning_rate": 1.4216455301029274e-07,
      "loss": 2.4853,
      "step": 89
    },
    {
      "epoch": 3.6233766233766236,
      "grad_norm": 1.9140625,
      "learning_rate": 1.0457956158838545e-07,
      "loss": 2.4193,
      "step": 90
    },
    {
      "epoch": 3.6233766233766236,
      "eval_loss": 3.278684139251709,
      "eval_runtime": 135.8573,
      "eval_samples_per_second": 1.634,
      "eval_steps_per_second": 1.634,
      "step": 90
    },
    {
      "epoch": 3.664935064935065,
      "grad_norm": 2.03125,
      "learning_rate": 7.270241141401568e-08,
      "loss": 2.4591,
      "step": 91
    },
    {
      "epoch": 3.7064935064935067,
      "grad_norm": 1.90625,
      "learning_rate": 4.657026981834623e-08,
      "loss": 2.4648,
      "step": 92
    },
    {
      "epoch": 3.7480519480519483,
      "grad_norm": 1.9765625,
      "learning_rate": 2.6213605709525803e-08,
      "loss": 2.4175,
      "step": 93
    },
    {
      "epoch": 3.78961038961039,
      "grad_norm": 2.0,
      "learning_rate": 1.1656154047303691e-08,
      "loss": 2.6494,
      "step": 94
    },
    {
      "epoch": 3.8311688311688314,
      "grad_norm": 1.9921875,
      "learning_rate": 2.9148881691298812e-09,
      "loss": 2.5267,
      "step": 95
    },
    {
      "epoch": 3.8727272727272726,
      "grad_norm": 2.046875,
      "learning_rate": 0.0,
      "loss": 2.4337,
      "step": 96
    },
    {
      "epoch": 3.8727272727272726,
      "eval_loss": 3.2782182693481445,
      "eval_runtime": 136.1248,
      "eval_samples_per_second": 1.631,
      "eval_steps_per_second": 1.631,
      "step": 96
    }
  ],
  "logging_steps": 1,
  "max_steps": 96,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 24,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.024520953011896e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}