{
  "best_global_step": 90,
  "best_metric": 0.4230247437953949,
  "best_model_checkpoint": "/tmp/checkpoints_/job_dee14721-c20d-4532-9438-f037e42d05fa/checkpoint-90",
  "epoch": 2.0,
  "eval_steps": 10,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 5.87138032913208,
      "learning_rate": 0.0,
      "loss": 3.333970069885254,
      "step": 1
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": NaN,
      "learning_rate": 1e-05,
      "loss": 3.8522658348083496,
      "step": 2
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 6.316606044769287,
      "learning_rate": 1e-05,
      "loss": 3.6522226333618164,
      "step": 3
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 7.078700065612793,
      "learning_rate": 2e-05,
      "loss": 3.447444438934326,
      "step": 4
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 6.362159729003906,
      "learning_rate": 3e-05,
      "loss": 3.92345929145813,
      "step": 5
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 6.476008415222168,
      "learning_rate": 4e-05,
      "loss": 3.6273388862609863,
      "step": 6
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 5.623299598693848,
      "learning_rate": 5e-05,
      "loss": 3.505007266998291,
      "step": 7
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 4.786308288574219,
      "learning_rate": 4.961538461538462e-05,
      "loss": 3.331632137298584,
      "step": 8
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.265624523162842,
      "learning_rate": 4.923076923076924e-05,
      "loss": 3.1580982208251953,
      "step": 9
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 5.30989933013916,
      "learning_rate": 4.884615384615385e-05,
      "loss": 3.02443790435791,
      "step": 10
    },
    {
      "epoch": 0.2222222222222222,
      "eval_loss": 3.1489245891571045,
      "eval_runtime": 0.6942,
      "eval_samples_per_second": 28.81,
      "eval_steps_per_second": 7.202,
      "step": 10
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 5.131834506988525,
      "learning_rate": 4.846153846153846e-05,
      "loss": 2.9736461639404297,
      "step": 11
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 5.027902603149414,
      "learning_rate": 4.8076923076923084e-05,
      "loss": 3.3082339763641357,
      "step": 12
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 4.972473621368408,
      "learning_rate": 4.76923076923077e-05,
      "loss": 3.136728286743164,
      "step": 13
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 5.016367435455322,
      "learning_rate": 4.730769230769231e-05,
      "loss": 3.0399789810180664,
      "step": 14
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 5.018073558807373,
      "learning_rate": 4.692307692307693e-05,
      "loss": 2.875364303588867,
      "step": 15
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 4.275290012359619,
      "learning_rate": 4.653846153846154e-05,
      "loss": 2.5802812576293945,
      "step": 16
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 5.652186870574951,
      "learning_rate": 4.615384615384616e-05,
      "loss": 2.5892136096954346,
      "step": 17
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.4209089279174805,
      "learning_rate": 4.576923076923077e-05,
      "loss": 2.4496333599090576,
      "step": 18
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 5.391123294830322,
      "learning_rate": 4.538461538461539e-05,
      "loss": 2.800614833831787,
      "step": 19
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 5.081115245819092,
      "learning_rate": 4.5e-05,
      "loss": 2.662580728530884,
      "step": 20
    },
    {
      "epoch": 0.4444444444444444,
      "eval_loss": 2.5146267414093018,
      "eval_runtime": 0.6885,
      "eval_samples_per_second": 29.048,
      "eval_steps_per_second": 7.262,
      "step": 20
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 6.204960823059082,
      "learning_rate": 4.461538461538462e-05,
      "loss": 2.5791611671447754,
      "step": 21
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 5.024198532104492,
      "learning_rate": 4.423076923076923e-05,
      "loss": 2.480907678604126,
      "step": 22
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 5.818419933319092,
      "learning_rate": 4.384615384615385e-05,
      "loss": 2.482961654663086,
      "step": 23
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 4.510318279266357,
      "learning_rate": 4.346153846153846e-05,
      "loss": 2.291290283203125,
      "step": 24
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 5.408440113067627,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 2.196018695831299,
      "step": 25
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 5.395529270172119,
      "learning_rate": 4.269230769230769e-05,
      "loss": 2.4523730278015137,
      "step": 26
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.714017868041992,
      "learning_rate": 4.230769230769231e-05,
      "loss": 2.0109505653381348,
      "step": 27
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 5.9115118980407715,
      "learning_rate": 4.192307692307693e-05,
      "loss": 2.0749101638793945,
      "step": 28
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 4.3610992431640625,
      "learning_rate": 4.1538461538461544e-05,
      "loss": 1.9675894975662231,
      "step": 29
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 4.657850742340088,
      "learning_rate": 4.115384615384615e-05,
      "loss": 1.7926883697509766,
      "step": 30
    },
    {
      "epoch": 0.6666666666666666,
      "eval_loss": 2.016049861907959,
      "eval_runtime": 0.7002,
      "eval_samples_per_second": 28.564,
      "eval_steps_per_second": 7.141,
      "step": 30
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 5.570796489715576,
      "learning_rate": 4.0769230769230773e-05,
      "loss": 1.9156622886657715,
      "step": 31
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 4.96622371673584,
      "learning_rate": 4.038461538461539e-05,
      "loss": 1.678513765335083,
      "step": 32
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 5.127869606018066,
      "learning_rate": 4e-05,
      "loss": 1.8418490886688232,
      "step": 33
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 5.629909038543701,
      "learning_rate": 3.961538461538462e-05,
      "loss": 1.8050181865692139,
      "step": 34
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 5.332027912139893,
      "learning_rate": 3.923076923076923e-05,
      "loss": 1.5306370258331299,
      "step": 35
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.3822021484375,
      "learning_rate": 3.884615384615385e-05,
      "loss": 1.665320873260498,
      "step": 36
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 5.958496570587158,
      "learning_rate": 3.846153846153846e-05,
      "loss": 1.483096718788147,
      "step": 37
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 6.1535539627075195,
      "learning_rate": 3.807692307692308e-05,
      "loss": 1.7172764539718628,
      "step": 38
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 6.259358882904053,
      "learning_rate": 3.769230769230769e-05,
      "loss": 1.7732570171356201,
      "step": 39
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 4.751931667327881,
      "learning_rate": 3.730769230769231e-05,
      "loss": 1.2387514114379883,
      "step": 40
    },
    {
      "epoch": 0.8888888888888888,
      "eval_loss": 1.6059402227401733,
      "eval_runtime": 0.6877,
      "eval_samples_per_second": 29.081,
      "eval_steps_per_second": 7.27,
      "step": 40
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 7.033973217010498,
      "learning_rate": 3.692307692307693e-05,
      "loss": 1.7372736930847168,
      "step": 41
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 6.1621599197387695,
      "learning_rate": 3.653846153846154e-05,
      "loss": 1.569137692451477,
      "step": 42
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 4.52653169631958,
      "learning_rate": 3.615384615384615e-05,
      "loss": 1.3664658069610596,
      "step": 43
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 5.134096622467041,
      "learning_rate": 3.5769230769230774e-05,
      "loss": 1.2201919555664062,
      "step": 44
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.698665142059326,
      "learning_rate": 3.538461538461539e-05,
      "loss": 1.447831153869629,
      "step": 45
    },
    {
      "epoch": 1.0222222222222221,
      "grad_norm": 4.538883209228516,
      "learning_rate": 3.5e-05,
      "loss": 1.277827262878418,
      "step": 46
    },
    {
      "epoch": 1.0444444444444445,
      "grad_norm": 5.439610004425049,
      "learning_rate": 3.461538461538462e-05,
      "loss": 1.285160779953003,
      "step": 47
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 5.188867092132568,
      "learning_rate": 3.4230769230769234e-05,
      "loss": 0.9196145534515381,
      "step": 48
    },
    {
      "epoch": 1.0888888888888888,
      "grad_norm": 5.185536861419678,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.9245643019676208,
      "step": 49
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 6.820492744445801,
      "learning_rate": 3.346153846153846e-05,
      "loss": 1.1991827487945557,
      "step": 50
    },
    {
      "epoch": 1.1111111111111112,
      "eval_loss": 1.2328484058380127,
      "eval_runtime": 0.6853,
      "eval_samples_per_second": 29.185,
      "eval_steps_per_second": 7.296,
      "step": 50
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 6.76528263092041,
      "learning_rate": 3.307692307692308e-05,
      "loss": 1.1789758205413818,
      "step": 51
    },
    {
      "epoch": 1.1555555555555554,
      "grad_norm": 6.927876949310303,
      "learning_rate": 3.269230769230769e-05,
      "loss": 0.9650276303291321,
      "step": 52
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 5.535730838775635,
      "learning_rate": 3.230769230769231e-05,
      "loss": 0.9436954259872437,
      "step": 53
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.6094441413879395,
      "learning_rate": 3.192307692307692e-05,
      "loss": 0.9939174652099609,
      "step": 54
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 5.440947532653809,
      "learning_rate": 3.153846153846154e-05,
      "loss": 1.0225588083267212,
      "step": 55
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 5.308763027191162,
      "learning_rate": 3.115384615384615e-05,
      "loss": 1.1128517389297485,
      "step": 56
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 6.255494117736816,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 1.0173512697219849,
      "step": 57
    },
    {
      "epoch": 1.2888888888888888,
      "grad_norm": 5.410623550415039,
      "learning_rate": 3.0384615384615382e-05,
      "loss": 0.8253780603408813,
      "step": 58
    },
    {
      "epoch": 1.3111111111111111,
      "grad_norm": 6.577842712402344,
      "learning_rate": 3e-05,
      "loss": 0.919189453125,
      "step": 59
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 5.62638521194458,
      "learning_rate": 2.9615384615384616e-05,
      "loss": 0.8745652437210083,
      "step": 60
    },
    {
      "epoch": 1.3333333333333333,
      "eval_loss": 0.9308679699897766,
      "eval_runtime": 0.6899,
      "eval_samples_per_second": 28.99,
      "eval_steps_per_second": 7.248,
      "step": 60
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 5.577669620513916,
      "learning_rate": 2.9230769230769234e-05,
      "loss": 0.9851076006889343,
      "step": 61
    },
    {
      "epoch": 1.3777777777777778,
      "grad_norm": 6.061960220336914,
      "learning_rate": 2.8846153846153845e-05,
      "loss": 0.8584026098251343,
      "step": 62
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.552649021148682,
      "learning_rate": 2.846153846153846e-05,
      "loss": 0.7414474487304688,
      "step": 63
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 5.330775737762451,
      "learning_rate": 2.807692307692308e-05,
      "loss": 0.7163383960723877,
      "step": 64
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 5.08062744140625,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.6349257826805115,
      "step": 65
    },
    {
      "epoch": 1.4666666666666668,
      "grad_norm": 6.24244499206543,
      "learning_rate": 2.7307692307692305e-05,
      "loss": 0.734074592590332,
      "step": 66
    },
    {
      "epoch": 1.488888888888889,
      "grad_norm": 5.124273777008057,
      "learning_rate": 2.6923076923076923e-05,
      "loss": 0.57187420129776,
      "step": 67
    },
    {
      "epoch": 1.511111111111111,
      "grad_norm": 7.708939075469971,
      "learning_rate": 2.6538461538461538e-05,
      "loss": 0.6959559321403503,
      "step": 68
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 8.645610809326172,
      "learning_rate": 2.6153846153846157e-05,
      "loss": 0.8737912178039551,
      "step": 69
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 5.395994186401367,
      "learning_rate": 2.5769230769230768e-05,
      "loss": 0.7113362550735474,
      "step": 70
    },
    {
      "epoch": 1.5555555555555556,
      "eval_loss": 0.6864913702011108,
      "eval_runtime": 0.6965,
      "eval_samples_per_second": 28.716,
      "eval_steps_per_second": 7.179,
      "step": 70
    },
    {
      "epoch": 1.5777777777777777,
      "grad_norm": 5.260697841644287,
      "learning_rate": 2.5384615384615383e-05,
      "loss": 0.5419008731842041,
      "step": 71
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.789445400238037,
      "learning_rate": 2.5e-05,
      "loss": 0.6413639783859253,
      "step": 72
    },
    {
      "epoch": 1.6222222222222222,
      "grad_norm": 6.9997124671936035,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.5860044360160828,
      "step": 73
    },
    {
      "epoch": 1.6444444444444444,
      "grad_norm": 5.2596564292907715,
      "learning_rate": 2.423076923076923e-05,
      "loss": 0.5544954538345337,
      "step": 74
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 4.984217166900635,
      "learning_rate": 2.384615384615385e-05,
      "loss": 0.5622981786727905,
      "step": 75
    },
    {
      "epoch": 1.6888888888888889,
      "grad_norm": 5.373387813568115,
      "learning_rate": 2.3461538461538464e-05,
      "loss": 0.5327613949775696,
      "step": 76
    },
    {
      "epoch": 1.7111111111111112,
      "grad_norm": 3.9502971172332764,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.47186025977134705,
      "step": 77
    },
    {
      "epoch": 1.7333333333333334,
      "grad_norm": 5.415130615234375,
      "learning_rate": 2.2692307692307694e-05,
      "loss": 0.5782798528671265,
      "step": 78
    },
    {
      "epoch": 1.7555555555555555,
      "grad_norm": 6.2089104652404785,
      "learning_rate": 2.230769230769231e-05,
      "loss": 0.6664197444915771,
      "step": 79
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 5.193050861358643,
      "learning_rate": 2.1923076923076924e-05,
      "loss": 0.5752044916152954,
      "step": 80
    },
    {
      "epoch": 1.7777777777777777,
      "eval_loss": 0.5675191879272461,
      "eval_runtime": 0.7073,
      "eval_samples_per_second": 28.278,
      "eval_steps_per_second": 7.069,
      "step": 80
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.315133571624756,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4048479199409485,
      "step": 81
    },
    {
      "epoch": 1.8222222222222222,
      "grad_norm": 5.358142375946045,
      "learning_rate": 2.1153846153846154e-05,
      "loss": 0.552968442440033,
      "step": 82
    },
    {
      "epoch": 1.8444444444444446,
      "grad_norm": 6.099940299987793,
      "learning_rate": 2.0769230769230772e-05,
      "loss": 0.4713404178619385,
      "step": 83
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 6.3764848709106445,
      "learning_rate": 2.0384615384615387e-05,
      "loss": 0.44388440251350403,
      "step": 84
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 5.32869815826416,
      "learning_rate": 2e-05,
      "loss": 0.35722827911376953,
      "step": 85
    },
    {
      "epoch": 1.911111111111111,
      "grad_norm": 4.6434431076049805,
      "learning_rate": 1.9615384615384617e-05,
      "loss": 0.37209007143974304,
      "step": 86
    },
    {
      "epoch": 1.9333333333333333,
      "grad_norm": 5.760197639465332,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.3956947326660156,
      "step": 87
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 4.9346489906311035,
      "learning_rate": 1.8846153846153846e-05,
      "loss": 0.34232836961746216,
      "step": 88
    },
    {
      "epoch": 1.9777777777777779,
      "grad_norm": 6.571290969848633,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5473036170005798,
      "step": 89
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.914135932922363,
      "learning_rate": 1.8076923076923076e-05,
      "loss": 0.43153804540634155,
      "step": 90
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4230247437953949,
      "eval_runtime": 0.7059,
      "eval_samples_per_second": 28.331,
      "eval_steps_per_second": 7.083,
      "step": 90
    }
  ],
  "logging_steps": 1,
  "max_steps": 135,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 93362349857280.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}