{
  "best_metric": 0.8977844146136806,
  "best_model_checkpoint": "../../checkpoint/unlearn/nlvr2/vilt/salun/6.0/42/checkpoint-13500",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 13500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 3.2584547996520996,
      "learning_rate": 9.62962962962963e-06,
      "loss": 0.3282,
      "step": 500
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 4.812547206878662,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.3197,
      "step": 1000
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.050705909729004,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.3219,
      "step": 1500
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 3.562553882598877,
      "learning_rate": 8.518518518518519e-06,
      "loss": 0.3194,
      "step": 2000
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 5.053567409515381,
      "learning_rate": 8.148148148148148e-06,
      "loss": 0.3079,
      "step": 2500
    },
    {
      "df_accuracy": 0.854215003866976,
      "dt_accuracy": 0.7513606416499571,
      "epoch": 1.0,
      "eval_accuracy": 0.7513606416499571,
      "eval_loss": 0.5662743449211121,
      "eval_runtime": 113.7346,
      "eval_samples_per_second": 61.389,
      "eval_steps_per_second": 0.967,
      "eval_unlearn_overall_accuracy": 0.6449160678015313,
      "step": 2700,
      "unlearn_overall_accuracy": 0.6449160678015313,
      "unlearn_time": null
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 1.7998104095458984,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.2708,
      "step": 3000
    },
    {
      "epoch": 1.2962962962962963,
      "grad_norm": 2.322847366333008,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.2454,
      "step": 3500
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 2.8383140563964844,
      "learning_rate": 7.0370370370370375e-06,
      "loss": 0.2361,
      "step": 4000
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 2.5046894550323486,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2486,
      "step": 4500
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 4.192043781280518,
      "learning_rate": 6.296296296296297e-06,
      "loss": 0.2432,
      "step": 5000
    },
    {
      "df_accuracy": 0.7016627996906419,
      "dt_accuracy": 0.7433400171870524,
      "epoch": 2.0,
      "eval_accuracy": 0.7433400171870524,
      "eval_loss": 0.6402536630630493,
      "eval_runtime": 108.1015,
      "eval_samples_per_second": 64.587,
      "eval_steps_per_second": 1.018,
      "eval_unlearn_overall_accuracy": 0.7676276016043365,
      "step": 5400,
      "unlearn_overall_accuracy": 0.7676276016043365,
      "unlearn_time": null
    },
    {
      "epoch": 2.037037037037037,
      "grad_norm": 4.436009407043457,
      "learning_rate": 5.925925925925926e-06,
      "loss": 0.2304,
      "step": 5500
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 3.5263144969940186,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.1807,
      "step": 6000
    },
    {
      "epoch": 2.4074074074074074,
      "grad_norm": 3.2321693897247314,
      "learning_rate": 5.185185185185185e-06,
      "loss": 0.188,
      "step": 6500
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 3.6127853393554688,
      "learning_rate": 4.814814814814815e-06,
      "loss": 0.1931,
      "step": 7000
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 4.856908798217773,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.1923,
      "step": 7500
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 8.728678703308105,
      "learning_rate": 4.074074074074074e-06,
      "loss": 0.1905,
      "step": 8000
    },
    {
      "df_accuracy": 0.5862335653518949,
      "dt_accuracy": 0.7396161558292753,
      "epoch": 3.0,
      "eval_accuracy": 0.7396161558292753,
      "eval_loss": 0.7980125546455383,
      "eval_runtime": 102.947,
      "eval_samples_per_second": 67.821,
      "eval_steps_per_second": 1.069,
      "eval_unlearn_overall_accuracy": 0.845758066951303,
      "step": 8100,
      "unlearn_overall_accuracy": 0.845758066951303,
      "unlearn_time": null
    },
    {
      "epoch": 3.148148148148148,
      "grad_norm": 0.8636759519577026,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1548,
      "step": 8500
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.8238606452941895,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.1483,
      "step": 9000
    },
    {
      "epoch": 3.5185185185185186,
      "grad_norm": 3.636707067489624,
      "learning_rate": 2.962962962962963e-06,
      "loss": 0.1477,
      "step": 9500
    },
    {
      "epoch": 3.7037037037037037,
      "grad_norm": 0.8585140705108643,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 0.1434,
      "step": 10000
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.986074209213257,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.1569,
      "step": 10500
    },
    {
      "df_accuracy": 0.5175947409126064,
      "dt_accuracy": 0.7354626181609853,
      "epoch": 4.0,
      "eval_accuracy": 0.7354626181609853,
      "eval_loss": 0.8833993673324585,
      "eval_runtime": 109.8215,
      "eval_samples_per_second": 63.576,
      "eval_steps_per_second": 1.002,
      "eval_unlearn_overall_accuracy": 0.8863248174250627,
      "step": 10800,
      "unlearn_overall_accuracy": 0.8863248174250627,
      "unlearn_time": null
    },
    {
      "epoch": 4.074074074074074,
      "grad_norm": 2.4611103534698486,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.1421,
      "step": 11000
    },
    {
      "epoch": 4.2592592592592595,
      "grad_norm": 3.0824742317199707,
      "learning_rate": 1.4814814814814815e-06,
      "loss": 0.1283,
      "step": 11500
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 2.9113352298736572,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.1198,
      "step": 12000
    },
    {
      "epoch": 4.62962962962963,
      "grad_norm": 0.5261613130569458,
      "learning_rate": 7.407407407407407e-07,
      "loss": 0.1196,
      "step": 12500
    },
    {
      "epoch": 4.814814814814815,
      "grad_norm": 4.700664520263672,
      "learning_rate": 3.7037037037037036e-07,
      "loss": 0.1336,
      "step": 13000
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.13147497177124,
      "learning_rate": 0.0,
      "loss": 0.1273,
      "step": 13500
    },
    {
      "df_accuracy": 0.49593967517401394,
      "dt_accuracy": 0.7327413348610713,
      "epoch": 5.0,
      "eval_accuracy": 0.7327413348610713,
      "eval_loss": 0.9442616701126099,
      "eval_runtime": 113.4763,
      "eval_samples_per_second": 61.528,
      "eval_steps_per_second": 0.969,
      "eval_unlearn_overall_accuracy": 0.8977844146136806,
      "step": 13500,
      "unlearn_overall_accuracy": 0.8977844146136806,
      "unlearn_time": null
    },
    {
      "epoch": 5.0,
      "step": 13500,
      "total_flos": 8453448251631240.0,
      "train_loss": 0.20511870829264323,
      "train_runtime": 11255.3092,
      "train_samples_per_second": 38.37,
      "train_steps_per_second": 1.199
    }
  ],
  "logging_steps": 500,
  "max_steps": 13500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8453448251631240.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}