trainer_state.json

{
  "best_global_step": 3000,
  "best_metric": 0.42,
  "best_model_checkpoint": "models/facebook/bart-base_30_5e-05/checkpoint-3000",
  "epoch": 12.096774193548388,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.0161290322580645,
      "grad_norm": 2.546293258666992,
      "learning_rate": 4.911573288058857e-05,
      "loss": 1.0786,
      "step": 500
    },
    {
      "epoch": 2.0161290322580645,
      "eval_gen_len": 19.5664,
      "eval_loss": 0.7127345204353333,
      "eval_rouge1": 0.3751,
      "eval_rouge2": 0.2929,
      "eval_rougeL": 0.3602,
      "eval_rougeLsum": 0.3651,
      "eval_runtime": 31.5466,
      "eval_samples_per_second": 35.82,
      "eval_steps_per_second": 1.141,
      "step": 500
    },
    {
      "epoch": 4.032258064516129,
      "grad_norm": 2.3489909172058105,
      "learning_rate": 4.557866440294284e-05,
      "loss": 0.6215,
      "step": 1000
    },
    {
      "epoch": 4.032258064516129,
      "eval_gen_len": 19.5982,
      "eval_loss": 0.7023738026618958,
      "eval_rouge1": 0.4106,
      "eval_rouge2": 0.3292,
      "eval_rougeL": 0.3958,
      "eval_rougeLsum": 0.4011,
      "eval_runtime": 31.0784,
      "eval_samples_per_second": 36.36,
      "eval_steps_per_second": 1.158,
      "step": 1000
    },
    {
      "epoch": 6.048387096774194,
      "grad_norm": 1.7153141498565674,
      "learning_rate": 4.204159592529712e-05,
      "loss": 0.46,
      "step": 1500
    },
    {
      "epoch": 6.048387096774194,
      "eval_gen_len": 19.0673,
      "eval_loss": 0.719135582447052,
      "eval_rouge1": 0.4193,
      "eval_rouge2": 0.34,
      "eval_rougeL": 0.4054,
      "eval_rougeLsum": 0.4105,
      "eval_runtime": 30.991,
      "eval_samples_per_second": 36.462,
      "eval_steps_per_second": 1.162,
      "step": 1500
    },
    {
      "epoch": 8.064516129032258,
      "grad_norm": 1.6871064901351929,
      "learning_rate": 3.850452744765139e-05,
      "loss": 0.3447,
      "step": 2000
    },
    {
      "epoch": 8.064516129032258,
      "eval_gen_len": 19.5212,
      "eval_loss": 0.7822473049163818,
      "eval_rouge1": 0.4084,
      "eval_rouge2": 0.3259,
      "eval_rougeL": 0.3945,
      "eval_rougeLsum": 0.3995,
      "eval_runtime": 31.1298,
      "eval_samples_per_second": 36.3,
      "eval_steps_per_second": 1.156,
      "step": 2000
    },
    {
      "epoch": 10.080645161290322,
      "grad_norm": 1.867733359336853,
      "learning_rate": 3.4967458970005664e-05,
      "loss": 0.2613,
      "step": 2500
    },
    {
      "epoch": 10.080645161290322,
      "eval_gen_len": 19.3071,
      "eval_loss": 0.8234670758247375,
      "eval_rouge1": 0.4259,
      "eval_rouge2": 0.3463,
      "eval_rougeL": 0.4125,
      "eval_rougeLsum": 0.4174,
      "eval_runtime": 31.0182,
      "eval_samples_per_second": 36.43,
      "eval_steps_per_second": 1.161,
      "step": 2500
    },
    {
      "epoch": 12.096774193548388,
      "grad_norm": 1.452479362487793,
      "learning_rate": 3.1444538766270515e-05,
      "loss": 0.1986,
      "step": 3000
    },
    {
      "epoch": 12.096774193548388,
      "eval_gen_len": 19.1496,
      "eval_loss": 0.8831787109375,
      "eval_rouge1": 0.4338,
      "eval_rouge2": 0.3544,
      "eval_rougeL": 0.42,
      "eval_rougeLsum": 0.4256,
      "eval_runtime": 31.0219,
      "eval_samples_per_second": 36.426,
      "eval_steps_per_second": 1.16,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 7440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "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": 5.674088748496896e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}