{
  "best_metric": 1.1082465648651123,
  "best_model_checkpoint": "./outputs/llava-mistral/RLAIF-V-Cosi-q0_25/checkpoint-200",
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 216,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 27.170432023765155,
      "learning_rate": 5e-07,
      "loss": 1.6754,
      "step": 5
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 14.247587906112043,
      "learning_rate": 1e-06,
      "loss": 1.5482,
      "step": 10
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 7.671787021030683,
      "learning_rate": 9.985471028179154e-07,
      "loss": 1.3299,
      "step": 15
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 6.453004152776314,
      "learning_rate": 9.94196854912548e-07,
      "loss": 1.2749,
      "step": 20
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 5.369919876340136,
      "learning_rate": 9.869745381355905e-07,
      "loss": 1.2236,
      "step": 25
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 5.219833369218594,
      "learning_rate": 9.769221256218162e-07,
      "loss": 1.1995,
      "step": 30
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 5.172890690456561,
      "learning_rate": 9.64098037858483e-07,
      "loss": 1.1886,
      "step": 35
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.842256114253865,
      "learning_rate": 9.485768031694871e-07,
      "loss": 1.1867,
      "step": 40
    },
    {
      "epoch": 0.625,
      "grad_norm": 4.960268983249933,
      "learning_rate": 9.304486245873971e-07,
      "loss": 1.1606,
      "step": 45
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 5.100919525621712,
      "learning_rate": 9.098188556305262e-07,
      "loss": 1.1561,
      "step": 50
    },
    {
      "epoch": 0.6944444444444444,
      "eval_loss": 1.1514836549758911,
      "eval_runtime": 36.3445,
      "eval_samples_per_second": 56.212,
      "eval_steps_per_second": 0.88,
      "step": 50
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 4.954408629703573,
      "learning_rate": 8.868073880316123e-07,
      "loss": 1.1461,
      "step": 55
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 5.005034060908683,
      "learning_rate": 8.615479549763755e-07,
      "loss": 1.1405,
      "step": 60
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 4.953280817436457,
      "learning_rate": 8.341873539012443e-07,
      "loss": 1.1317,
      "step": 65
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 4.812467946627597,
      "learning_rate": 8.048845933670271e-07,
      "loss": 1.1163,
      "step": 70
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 5.832648715058256,
      "learning_rate": 7.738099689665539e-07,
      "loss": 1.0551,
      "step": 75
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 5.011097198106542,
      "learning_rate": 7.41144073636728e-07,
      "loss": 1.0026,
      "step": 80
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 5.207816191466898,
      "learning_rate": 7.070767481266492e-07,
      "loss": 1.0002,
      "step": 85
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.288582074627771,
      "learning_rate": 6.718059777212565e-07,
      "loss": 0.9809,
      "step": 90
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 5.272544049813405,
      "learning_rate": 6.355367416322778e-07,
      "loss": 0.982,
      "step": 95
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 5.34229662194457,
      "learning_rate": 5.984798217433531e-07,
      "loss": 0.9725,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "eval_loss": 1.1154539585113525,
      "eval_runtime": 35.7418,
      "eval_samples_per_second": 57.16,
      "eval_steps_per_second": 0.895,
      "step": 100
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 5.332872312807783,
      "learning_rate": 5.608505776324157e-07,
      "loss": 0.9741,
      "step": 105
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 5.039881296049518,
      "learning_rate": 5.228676949903973e-07,
      "loss": 0.9787,
      "step": 110
    },
    {
      "epoch": 1.5972222222222223,
      "grad_norm": 5.272528296120645,
      "learning_rate": 4.847519147099294e-07,
      "loss": 0.9577,
      "step": 115
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 6.108766118631545,
      "learning_rate": 4.46724750030062e-07,
      "loss": 0.9597,
      "step": 120
    },
    {
      "epoch": 1.7361111111111112,
      "grad_norm": 5.206098606588932,
      "learning_rate": 4.0900719919241935e-07,
      "loss": 0.9648,
      "step": 125
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 5.197849839562957,
      "learning_rate": 3.7181846109031e-07,
      "loss": 0.947,
      "step": 130
    },
    {
      "epoch": 1.875,
      "grad_norm": 5.3111405062375905,
      "learning_rate": 3.353746613749093e-07,
      "loss": 0.9594,
      "step": 135
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 5.445921545754994,
      "learning_rate": 2.9988759642186093e-07,
      "loss": 0.9702,
      "step": 140
    },
    {
      "epoch": 2.013888888888889,
      "grad_norm": 5.933633048275964,
      "learning_rate": 2.655635024578483e-07,
      "loss": 0.9456,
      "step": 145
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 5.549115678624608,
      "learning_rate": 2.3260185700046292e-07,
      "loss": 0.8717,
      "step": 150
    },
    {
      "epoch": 2.0833333333333335,
      "eval_loss": 1.1032688617706299,
      "eval_runtime": 35.642,
      "eval_samples_per_second": 57.32,
      "eval_steps_per_second": 0.898,
      "step": 150
    },
    {
      "epoch": 2.1527777777777777,
      "grad_norm": 5.425494474442028,
      "learning_rate": 2.0119421957691218e-07,
      "loss": 0.8643,
      "step": 155
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.818398592523753,
      "learning_rate": 1.7152311845883094e-07,
      "loss": 0.8643,
      "step": 160
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 5.873632446353599,
      "learning_rate": 1.4376098988303404e-07,
      "loss": 0.8696,
      "step": 165
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 5.4343592118456785,
      "learning_rate": 1.1806917592302761e-07,
      "loss": 0.8444,
      "step": 170
    },
    {
      "epoch": 2.4305555555555554,
      "grad_norm": 5.333034391959962,
      "learning_rate": 9.459698683523204e-08,
      "loss": 0.8694,
      "step": 175
    },
    {
      "epoch": 2.5,
      "grad_norm": 5.655910129037808,
      "learning_rate": 7.348083332917926e-08,
      "loss": 0.8773,
      "step": 180
    },
    {
      "epoch": 2.5694444444444446,
      "grad_norm": 5.278576061447754,
      "learning_rate": 5.484343380457124e-08,
      "loss": 0.8622,
      "step": 185
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 5.493429379631295,
      "learning_rate": 3.879310116241041e-08,
      "loss": 0.8657,
      "step": 190
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 5.821505606490213,
      "learning_rate": 2.5423113334966218e-08,
      "loss": 0.8589,
      "step": 195
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 5.676459483240144,
      "learning_rate": 1.4811171192794624e-08,
      "loss": 0.8608,
      "step": 200
    },
    {
      "epoch": 2.7777777777777777,
      "eval_loss": 1.1082465648651123,
      "eval_runtime": 35.5876,
      "eval_samples_per_second": 57.408,
      "eval_steps_per_second": 0.899,
      "step": 200
    },
    {
      "epoch": 2.8472222222222223,
      "grad_norm": 5.585848892693237,
      "learning_rate": 7.018946979234997e-09,
      "loss": 0.8773,
      "step": 205
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 5.591393456061243,
      "learning_rate": 2.0917258966953734e-09,
      "loss": 0.8621,
      "step": 210
    },
    {
      "epoch": 2.986111111111111,
      "grad_norm": 5.629015994718156,
      "learning_rate": 5.814292768108187e-11,
      "loss": 0.8677,
      "step": 215
    },
    {
      "epoch": 3.0,
      "step": 216,
      "total_flos": 1273399955423232.0,
      "train_loss": 1.028549515262798,
      "train_runtime": 3153.0326,
      "train_samples_per_second": 17.493,
      "train_steps_per_second": 0.069
    }
  ],
  "logging_steps": 5,
  "max_steps": 216,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1273399955423232.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}