{
  "best_metric": 0.6440326571464539,
  "best_model_checkpoint": "data/Gemma-2-2B_task-2_120-samples_config-1_auto/checkpoint-55",
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 132,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.7323388457298279,
      "learning_rate": 1.818181818181818e-06,
      "loss": 0.9734,
      "step": 1
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.9116188287734985,
      "learning_rate": 3.636363636363636e-06,
      "loss": 1.1167,
      "step": 2
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.7275955677032471,
      "learning_rate": 7.272727272727272e-06,
      "loss": 0.9397,
      "step": 4
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.721298336982727,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.9957,
      "step": 6
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.7969232201576233,
      "learning_rate": 1.4545454545454545e-05,
      "loss": 1.0424,
      "step": 8
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.6864991188049316,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.9373,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9925083518028259,
      "eval_runtime": 9.6526,
      "eval_samples_per_second": 2.486,
      "eval_steps_per_second": 2.486,
      "step": 11
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.7458894848823547,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 0.9672,
      "step": 12
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.6642060279846191,
      "learning_rate": 2.5454545454545454e-05,
      "loss": 0.9042,
      "step": 14
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6158398985862732,
      "learning_rate": 2.909090909090909e-05,
      "loss": 0.9204,
      "step": 16
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5651478171348572,
      "learning_rate": 3.272727272727273e-05,
      "loss": 0.9285,
      "step": 18
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.6330985426902771,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.8125,
      "step": 20
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4895676374435425,
      "learning_rate": 4e-05,
      "loss": 0.7244,
      "step": 22
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8189184665679932,
      "eval_runtime": 9.3856,
      "eval_samples_per_second": 2.557,
      "eval_steps_per_second": 2.557,
      "step": 22
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.46822482347488403,
      "learning_rate": 4.3636363636363636e-05,
      "loss": 0.7252,
      "step": 24
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.5026049613952637,
      "learning_rate": 4.7272727272727275e-05,
      "loss": 0.7829,
      "step": 26
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.4652740955352783,
      "learning_rate": 5.090909090909091e-05,
      "loss": 0.7031,
      "step": 28
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.49133890867233276,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.6953,
      "step": 30
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.4810543656349182,
      "learning_rate": 5.818181818181818e-05,
      "loss": 0.6269,
      "step": 32
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7204152941703796,
      "eval_runtime": 9.3876,
      "eval_samples_per_second": 2.557,
      "eval_steps_per_second": 2.557,
      "step": 33
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.4971282184123993,
      "learning_rate": 6.181818181818182e-05,
      "loss": 0.7309,
      "step": 34
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.4785713851451874,
      "learning_rate": 6.545454545454546e-05,
      "loss": 0.6176,
      "step": 36
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.45341768860816956,
      "learning_rate": 6.90909090909091e-05,
      "loss": 0.6018,
      "step": 38
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.5038592219352722,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.6105,
      "step": 40
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.5255274772644043,
      "learning_rate": 7.636363636363637e-05,
      "loss": 0.6159,
      "step": 42
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4701821506023407,
      "learning_rate": 8e-05,
      "loss": 0.588,
      "step": 44
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.6696464419364929,
      "eval_runtime": 9.3942,
      "eval_samples_per_second": 2.555,
      "eval_steps_per_second": 2.555,
      "step": 44
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5235864520072937,
      "learning_rate": 8.363636363636364e-05,
      "loss": 0.5666,
      "step": 46
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.4793376922607422,
      "learning_rate": 8.727272727272727e-05,
      "loss": 0.5587,
      "step": 48
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.4552352726459503,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.5214,
      "step": 50
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.5394156575202942,
      "learning_rate": 9.454545454545455e-05,
      "loss": 0.5624,
      "step": 52
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.5040570497512817,
      "learning_rate": 9.818181818181818e-05,
      "loss": 0.526,
      "step": 54
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.6440326571464539,
      "eval_runtime": 9.3919,
      "eval_samples_per_second": 2.555,
      "eval_steps_per_second": 2.555,
      "step": 55
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4954177141189575,
      "learning_rate": 9.999899300364532e-05,
      "loss": 0.4931,
      "step": 56
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.5694147944450378,
      "learning_rate": 9.99909372761763e-05,
      "loss": 0.5079,
      "step": 58
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.5428325533866882,
      "learning_rate": 9.997482711915927e-05,
      "loss": 0.4285,
      "step": 60
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.5923798680305481,
      "learning_rate": 9.99506651282272e-05,
      "loss": 0.4427,
      "step": 62
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.6247926354408264,
      "learning_rate": 9.991845519630678e-05,
      "loss": 0.4559,
      "step": 64
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6745326519012451,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.4669,
      "step": 66
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.6452058553695679,
      "eval_runtime": 9.3854,
      "eval_samples_per_second": 2.557,
      "eval_steps_per_second": 2.557,
      "step": 66
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.6631373167037964,
      "learning_rate": 9.982991356370404e-05,
      "loss": 0.3826,
      "step": 68
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.5684555768966675,
      "learning_rate": 9.977359612865423e-05,
      "loss": 0.3885,
      "step": 70
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.7757788896560669,
      "learning_rate": 9.970925928158274e-05,
      "loss": 0.3673,
      "step": 72
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.7346789240837097,
      "learning_rate": 9.963691338830044e-05,
      "loss": 0.3335,
      "step": 74
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.83540278673172,
      "learning_rate": 9.955657010501806e-05,
      "loss": 0.3682,
      "step": 76
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.7008591294288635,
      "eval_runtime": 9.394,
      "eval_samples_per_second": 2.555,
      "eval_steps_per_second": 2.555,
      "step": 77
    },
    {
      "epoch": 7.090909090909091,
      "grad_norm": 0.803167998790741,
      "learning_rate": 9.946824237646824e-05,
      "loss": 0.3355,
      "step": 78
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 0.7409563064575195,
      "learning_rate": 9.937194443381972e-05,
      "loss": 0.2621,
      "step": 80
    },
    {
      "epoch": 7.454545454545454,
      "grad_norm": 0.8879055380821228,
      "learning_rate": 9.926769179238466e-05,
      "loss": 0.226,
      "step": 82
    },
    {
      "epoch": 7.636363636363637,
      "grad_norm": 0.918166995048523,
      "learning_rate": 9.915550124911866e-05,
      "loss": 0.2492,
      "step": 84
    },
    {
      "epoch": 7.818181818181818,
      "grad_norm": 1.237692952156067,
      "learning_rate": 9.903539087991462e-05,
      "loss": 0.2737,
      "step": 86
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.1434404850006104,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.2714,
      "step": 88
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.8049829602241516,
      "eval_runtime": 9.3942,
      "eval_samples_per_second": 2.555,
      "eval_steps_per_second": 2.555,
      "step": 88
    },
    {
      "epoch": 8.181818181818182,
      "grad_norm": 0.8012731671333313,
      "learning_rate": 9.877148934427037e-05,
      "loss": 0.1763,
      "step": 90
    },
    {
      "epoch": 8.363636363636363,
      "grad_norm": 0.9968366026878357,
      "learning_rate": 9.862774069706346e-05,
      "loss": 0.1549,
      "step": 92
    },
    {
      "epoch": 8.545454545454545,
      "grad_norm": 0.9959315657615662,
      "learning_rate": 9.847615725553456e-05,
      "loss": 0.1541,
      "step": 94
    },
    {
      "epoch": 8.727272727272727,
      "grad_norm": 1.1761013269424438,
      "learning_rate": 9.831676344247342e-05,
      "loss": 0.143,
      "step": 96
    },
    {
      "epoch": 8.909090909090908,
      "grad_norm": 2.0634052753448486,
      "learning_rate": 9.814958493905963e-05,
      "loss": 0.1657,
      "step": 98
    },
    {
      "epoch": 9.0,
      "eval_loss": 1.0493817329406738,
      "eval_runtime": 9.3892,
      "eval_samples_per_second": 2.556,
      "eval_steps_per_second": 2.556,
      "step": 99
    },
    {
      "epoch": 9.090909090909092,
      "grad_norm": 1.3311676979064941,
      "learning_rate": 9.797464868072488e-05,
      "loss": 0.16,
      "step": 100
    },
    {
      "epoch": 9.272727272727273,
      "grad_norm": 0.9883970022201538,
      "learning_rate": 9.779198285281325e-05,
      "loss": 0.089,
      "step": 102
    },
    {
      "epoch": 9.454545454545455,
      "grad_norm": 1.2211483716964722,
      "learning_rate": 9.760161688604008e-05,
      "loss": 0.0928,
      "step": 104
    },
    {
      "epoch": 9.636363636363637,
      "grad_norm": 1.2108052968978882,
      "learning_rate": 9.740358145174998e-05,
      "loss": 0.1084,
      "step": 106
    },
    {
      "epoch": 9.818181818181818,
      "grad_norm": 1.1194443702697754,
      "learning_rate": 9.719790845697533e-05,
      "loss": 0.0833,
      "step": 108
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.244794249534607,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.076,
      "step": 110
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.158414363861084,
      "eval_runtime": 9.3848,
      "eval_samples_per_second": 2.557,
      "eval_steps_per_second": 2.557,
      "step": 110
    },
    {
      "epoch": 10.181818181818182,
      "grad_norm": 1.0360634326934814,
      "learning_rate": 9.676378356149734e-05,
      "loss": 0.0498,
      "step": 112
    },
    {
      "epoch": 10.363636363636363,
      "grad_norm": 0.8835354447364807,
      "learning_rate": 9.653540160603956e-05,
      "loss": 0.0532,
      "step": 114
    },
    {
      "epoch": 10.545454545454545,
      "grad_norm": 1.203364610671997,
      "learning_rate": 9.629952196931901e-05,
      "loss": 0.0473,
      "step": 116
    },
    {
      "epoch": 10.727272727272727,
      "grad_norm": 1.461548089981079,
      "learning_rate": 9.60561826557425e-05,
      "loss": 0.0644,
      "step": 118
    },
    {
      "epoch": 10.909090909090908,
      "grad_norm": 1.0692554712295532,
      "learning_rate": 9.580542287160348e-05,
      "loss": 0.0507,
      "step": 120
    },
    {
      "epoch": 11.0,
      "eval_loss": 1.344495415687561,
      "eval_runtime": 9.3858,
      "eval_samples_per_second": 2.557,
      "eval_steps_per_second": 2.557,
      "step": 121
    },
    {
      "epoch": 11.090909090909092,
      "grad_norm": 0.7460653781890869,
      "learning_rate": 9.554728301876526e-05,
      "loss": 0.0349,
      "step": 122
    },
    {
      "epoch": 11.272727272727273,
      "grad_norm": 0.9139772653579712,
      "learning_rate": 9.528180468815155e-05,
      "loss": 0.0266,
      "step": 124
    },
    {
      "epoch": 11.454545454545455,
      "grad_norm": 0.912947952747345,
      "learning_rate": 9.50090306530454e-05,
      "loss": 0.0336,
      "step": 126
    },
    {
      "epoch": 11.636363636363637,
      "grad_norm": 1.0267078876495361,
      "learning_rate": 9.472900486219769e-05,
      "loss": 0.0353,
      "step": 128
    },
    {
      "epoch": 11.818181818181818,
      "grad_norm": 0.9120265245437622,
      "learning_rate": 9.444177243274618e-05,
      "loss": 0.0298,
      "step": 130
    },
    {
      "epoch": 12.0,
      "grad_norm": 1.4009398221969604,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.0346,
      "step": 132
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.344093918800354,
      "eval_runtime": 9.4052,
      "eval_samples_per_second": 2.552,
      "eval_steps_per_second": 2.552,
      "step": 132
    },
    {
      "epoch": 12.0,
      "step": 132,
      "total_flos": 3.550877504556237e+16,
      "train_loss": 0.42980391238674975,
      "train_runtime": 1517.2103,
      "train_samples_per_second": 2.9,
      "train_steps_per_second": 0.363
    }
  ],
  "logging_steps": 2,
  "max_steps": 550,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 25,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 7,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.550877504556237e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}