{
  "best_metric": 0.510191798210144,
  "best_model_checkpoint": "./Vit-GPT2-COCO2017Flickr-85k-09/checkpoint-4000",
  "epoch": 1.4930484277316414,
  "eval_steps": 500,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09330969487729775,
      "grad_norm": 1.1049598455429077,
      "learning_rate": 4.8444693293517484e-05,
      "loss": 0.2429,
      "step": 500
    },
    {
      "epoch": 0.09330969487729775,
      "eval_gen_len": 11.738,
      "eval_loss": 0.5351373553276062,
      "eval_rouge1": 39.4446,
      "eval_rouge2": 14.1599,
      "eval_rougeL": 35.6123,
      "eval_rougeLsum": 35.5846,
      "eval_runtime": 463.8252,
      "eval_samples_per_second": 8.624,
      "eval_steps_per_second": 2.156,
      "step": 500
    },
    {
      "epoch": 0.1866193897545955,
      "grad_norm": 1.1426328420639038,
      "learning_rate": 4.6889386587034965e-05,
      "loss": 0.2537,
      "step": 1000
    },
    {
      "epoch": 0.1866193897545955,
      "eval_gen_len": 12.34875,
      "eval_loss": 0.5301018357276917,
      "eval_rouge1": 39.5332,
      "eval_rouge2": 14.4745,
      "eval_rougeL": 35.644,
      "eval_rougeLsum": 35.6159,
      "eval_runtime": 451.0336,
      "eval_samples_per_second": 8.869,
      "eval_steps_per_second": 2.217,
      "step": 1000
    },
    {
      "epoch": 0.27992908463189325,
      "grad_norm": 1.017817735671997,
      "learning_rate": 4.5334079880552446e-05,
      "loss": 0.2564,
      "step": 1500
    },
    {
      "epoch": 0.27992908463189325,
      "eval_gen_len": 12.2455,
      "eval_loss": 0.5198264122009277,
      "eval_rouge1": 39.8297,
      "eval_rouge2": 14.555,
      "eval_rougeL": 35.8598,
      "eval_rougeLsum": 35.8344,
      "eval_runtime": 446.7054,
      "eval_samples_per_second": 8.954,
      "eval_steps_per_second": 2.239,
      "step": 1500
    },
    {
      "epoch": 0.373238779509191,
      "grad_norm": 0.9865152835845947,
      "learning_rate": 4.3778773174069934e-05,
      "loss": 0.2585,
      "step": 2000
    },
    {
      "epoch": 0.373238779509191,
      "eval_gen_len": 11.8575,
      "eval_loss": 0.5207428932189941,
      "eval_rouge1": 39.4558,
      "eval_rouge2": 14.0496,
      "eval_rougeL": 35.5597,
      "eval_rougeLsum": 35.526,
      "eval_runtime": 441.2453,
      "eval_samples_per_second": 9.065,
      "eval_steps_per_second": 2.266,
      "step": 2000
    },
    {
      "epoch": 0.4665484743864888,
      "grad_norm": 0.9201194047927856,
      "learning_rate": 4.222346646758741e-05,
      "loss": 0.2579,
      "step": 2500
    },
    {
      "epoch": 0.4665484743864888,
      "eval_gen_len": 11.97325,
      "eval_loss": 0.5187843441963196,
      "eval_rouge1": 39.1359,
      "eval_rouge2": 14.125,
      "eval_rougeL": 35.4068,
      "eval_rougeLsum": 35.3709,
      "eval_runtime": 456.9413,
      "eval_samples_per_second": 8.754,
      "eval_steps_per_second": 2.188,
      "step": 2500
    },
    {
      "epoch": 0.5598581692637865,
      "grad_norm": 1.011809229850769,
      "learning_rate": 4.066815976110489e-05,
      "loss": 0.2588,
      "step": 3000
    },
    {
      "epoch": 0.5598581692637865,
      "eval_gen_len": 12.278,
      "eval_loss": 0.5195733904838562,
      "eval_rouge1": 39.0831,
      "eval_rouge2": 14.0658,
      "eval_rougeL": 35.4608,
      "eval_rougeLsum": 35.4283,
      "eval_runtime": 453.7444,
      "eval_samples_per_second": 8.816,
      "eval_steps_per_second": 2.204,
      "step": 3000
    },
    {
      "epoch": 0.6531678641410843,
      "grad_norm": 0.9602178931236267,
      "learning_rate": 3.911285305462237e-05,
      "loss": 0.2618,
      "step": 3500
    },
    {
      "epoch": 0.6531678641410843,
      "eval_gen_len": 11.99425,
      "eval_loss": 0.5193876028060913,
      "eval_rouge1": 39.751,
      "eval_rouge2": 14.443,
      "eval_rougeL": 36.076,
      "eval_rougeLsum": 36.0475,
      "eval_runtime": 451.6054,
      "eval_samples_per_second": 8.857,
      "eval_steps_per_second": 2.214,
      "step": 3500
    },
    {
      "epoch": 0.746477559018382,
      "grad_norm": 1.0043072700500488,
      "learning_rate": 3.755754634813985e-05,
      "loss": 0.2579,
      "step": 4000
    },
    {
      "epoch": 0.746477559018382,
      "eval_gen_len": 12.05125,
      "eval_loss": 0.510191798210144,
      "eval_rouge1": 39.7601,
      "eval_rouge2": 14.5095,
      "eval_rougeL": 36.0252,
      "eval_rougeLsum": 35.9857,
      "eval_runtime": 450.2804,
      "eval_samples_per_second": 8.883,
      "eval_steps_per_second": 2.221,
      "step": 4000
    },
    {
      "epoch": 0.8397872538956798,
      "grad_norm": 0.9865091443061829,
      "learning_rate": 3.600223964165734e-05,
      "loss": 0.2569,
      "step": 4500
    },
    {
      "epoch": 0.8397872538956798,
      "eval_gen_len": 11.64825,
      "eval_loss": 0.5199217796325684,
      "eval_rouge1": 39.398,
      "eval_rouge2": 13.8871,
      "eval_rougeL": 35.7218,
      "eval_rougeLsum": 35.6911,
      "eval_runtime": 440.2487,
      "eval_samples_per_second": 9.086,
      "eval_steps_per_second": 2.271,
      "step": 4500
    },
    {
      "epoch": 0.9330969487729776,
      "grad_norm": 1.1134284734725952,
      "learning_rate": 3.444693293517482e-05,
      "loss": 0.253,
      "step": 5000
    },
    {
      "epoch": 0.9330969487729776,
      "eval_gen_len": 12.01975,
      "eval_loss": 0.5200353860855103,
      "eval_rouge1": 39.8951,
      "eval_rouge2": 14.4146,
      "eval_rougeL": 35.883,
      "eval_rougeLsum": 35.8507,
      "eval_runtime": 445.4978,
      "eval_samples_per_second": 8.979,
      "eval_steps_per_second": 2.245,
      "step": 5000
    },
    {
      "epoch": 1.0264066436502752,
      "grad_norm": 0.8591415286064148,
      "learning_rate": 3.28916262286923e-05,
      "loss": 0.2361,
      "step": 5500
    },
    {
      "epoch": 1.0264066436502752,
      "eval_gen_len": 12.183,
      "eval_loss": 0.560513973236084,
      "eval_rouge1": 39.3352,
      "eval_rouge2": 14.2234,
      "eval_rougeL": 35.3107,
      "eval_rougeLsum": 35.2772,
      "eval_runtime": 451.5632,
      "eval_samples_per_second": 8.858,
      "eval_steps_per_second": 2.215,
      "step": 5500
    },
    {
      "epoch": 1.119716338527573,
      "grad_norm": 0.9471074342727661,
      "learning_rate": 3.1336319522209783e-05,
      "loss": 0.2,
      "step": 6000
    },
    {
      "epoch": 1.119716338527573,
      "eval_gen_len": 11.85975,
      "eval_loss": 0.5701556205749512,
      "eval_rouge1": 39.2184,
      "eval_rouge2": 14.0096,
      "eval_rougeL": 35.5475,
      "eval_rougeLsum": 35.5042,
      "eval_runtime": 443.3267,
      "eval_samples_per_second": 9.023,
      "eval_steps_per_second": 2.256,
      "step": 6000
    },
    {
      "epoch": 1.2130260334048708,
      "grad_norm": 1.184536099433899,
      "learning_rate": 2.978101281572726e-05,
      "loss": 0.2034,
      "step": 6500
    },
    {
      "epoch": 1.2130260334048708,
      "eval_gen_len": 11.878,
      "eval_loss": 0.5543492436408997,
      "eval_rouge1": 39.7118,
      "eval_rouge2": 14.2757,
      "eval_rougeL": 35.7613,
      "eval_rougeLsum": 35.7316,
      "eval_runtime": 448.2166,
      "eval_samples_per_second": 8.924,
      "eval_steps_per_second": 2.231,
      "step": 6500
    },
    {
      "epoch": 1.3063357282821686,
      "grad_norm": 0.9648805260658264,
      "learning_rate": 2.8225706109244742e-05,
      "loss": 0.1968,
      "step": 7000
    },
    {
      "epoch": 1.3063357282821686,
      "eval_gen_len": 12.1725,
      "eval_loss": 0.5584425330162048,
      "eval_rouge1": 39.1847,
      "eval_rouge2": 13.9003,
      "eval_rougeL": 35.3962,
      "eval_rougeLsum": 35.3713,
      "eval_runtime": 447.4882,
      "eval_samples_per_second": 8.939,
      "eval_steps_per_second": 2.235,
      "step": 7000
    },
    {
      "epoch": 1.3996454231594662,
      "grad_norm": 1.0540575981140137,
      "learning_rate": 2.6670399402762224e-05,
      "loss": 0.1986,
      "step": 7500
    },
    {
      "epoch": 1.3996454231594662,
      "eval_gen_len": 11.8395,
      "eval_loss": 0.5572038292884827,
      "eval_rouge1": 39.4428,
      "eval_rouge2": 14.2672,
      "eval_rougeL": 35.7359,
      "eval_rougeLsum": 35.7093,
      "eval_runtime": 437.4159,
      "eval_samples_per_second": 9.145,
      "eval_steps_per_second": 2.286,
      "step": 7500
    },
    {
      "epoch": 1.4930484277316414,
      "grad_norm": 1.1180176734924316,
      "learning_rate": 2.5115092696279708e-05,
      "loss": 0.1988,
      "step": 8000
    },
    {
      "epoch": 1.4930484277316414,
      "eval_gen_len": 11.99325,
      "eval_loss": 0.555167019367218,
      "eval_rouge1": 39.2719,
      "eval_rouge2": 14.0411,
      "eval_rougeL": 35.482,
      "eval_rougeLsum": 35.4833,
      "eval_runtime": 466.4795,
      "eval_samples_per_second": 8.575,
      "eval_steps_per_second": 2.144,
      "step": 8000
    }
  ],
  "logging_steps": 500,
  "max_steps": 16074,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.3099017135594144e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}