checkpoint-7500/trainer_state.json

{
  "best_metric": 1.3103234767913818,
  "best_model_checkpoint": "./kozh_mbartLarge_50p_run3_tokenize/checkpoint-7500",
  "epoch": 1.9952114924181963,
  "eval_steps": 7500,
  "global_step": 7500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 2.5913,
      "step": 500
    },
    {
      "epoch": 0.27,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.8919,
      "step": 1000
    },
    {
      "epoch": 0.4,
      "learning_rate": 5e-05,
      "loss": 1.7345,
      "step": 1500
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.930728733721253e-05,
      "loss": 1.6451,
      "step": 2000
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.861457467442505e-05,
      "loss": 1.5539,
      "step": 2500
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.792186201163758e-05,
      "loss": 1.512,
      "step": 3000
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.72291493488501e-05,
      "loss": 1.4671,
      "step": 3500
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.6536436686062625e-05,
      "loss": 1.417,
      "step": 4000
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.5843724023275146e-05,
      "loss": 1.3449,
      "step": 4500
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.515101136048767e-05,
      "loss": 1.2543,
      "step": 5000
    },
    {
      "epoch": 1.46,
      "learning_rate": 4.4458298697700194e-05,
      "loss": 1.1794,
      "step": 5500
    },
    {
      "epoch": 1.6,
      "learning_rate": 4.376558603491272e-05,
      "loss": 1.1359,
      "step": 6000
    },
    {
      "epoch": 1.73,
      "learning_rate": 4.307287337212524e-05,
      "loss": 1.1075,
      "step": 6500
    },
    {
      "epoch": 1.86,
      "learning_rate": 4.238016070933777e-05,
      "loss": 1.1079,
      "step": 7000
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.168744804655029e-05,
      "loss": 1.0808,
      "step": 7500
    },
    {
      "epoch": 2.0,
      "eval_bleu": 51.0184,
      "eval_gen_len": 13.3575,
      "eval_loss": 1.3103234767913818,
      "eval_runtime": 1625.4898,
      "eval_samples_per_second": 18.499,
      "eval_steps_per_second": 1.157,
      "step": 7500
    }
  ],
  "logging_steps": 500,
  "max_steps": 37590,
  "num_train_epochs": 10,
  "save_steps": 7500,
  "total_flos": 1.04022228860928e+18,
  "trial_name": null,
  "trial_params": null
}