{
  "best_metric": 27.8224,
  "best_model_checkpoint": "./ko-en_mbartLarge_exp20p_batch/checkpoint-8000",
  "epoch": 3.712727694628147,
  "eval_steps": 4000,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "learning_rate": 1.25e-05,
      "loss": 1.5446,
      "step": 500
    },
    {
      "epoch": 0.46,
      "learning_rate": 2.5e-05,
      "loss": 1.343,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.2767,
      "step": 1500
    },
    {
      "epoch": 0.93,
      "learning_rate": 5e-05,
      "loss": 1.242,
      "step": 2000
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.9995663025798074e-05,
      "loss": 1.1213,
      "step": 2500
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.998261883224581e-05,
      "loss": 1.0725,
      "step": 3000
    },
    {
      "epoch": 1.62,
      "learning_rate": 4.996087194586878e-05,
      "loss": 1.0628,
      "step": 3500
    },
    {
      "epoch": 1.86,
      "learning_rate": 4.99304299421927e-05,
      "loss": 1.0603,
      "step": 4000
    },
    {
      "epoch": 1.86,
      "eval_bleu": 27.7215,
      "eval_gen_len": 18.8456,
      "eval_loss": 1.1473960876464844,
      "eval_runtime": 1471.0712,
      "eval_samples_per_second": 11.717,
      "eval_steps_per_second": 1.465,
      "step": 4000
    },
    {
      "epoch": 2.09,
      "learning_rate": 4.989130342568659e-05,
      "loss": 0.9588,
      "step": 4500
    },
    {
      "epoch": 2.32,
      "learning_rate": 4.984371446700271e-05,
      "loss": 0.8217,
      "step": 5000
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.978741887280011e-05,
      "loss": 0.8436,
      "step": 5500
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.972238441511973e-05,
      "loss": 0.8488,
      "step": 6000
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.964873791724829e-05,
      "loss": 0.8471,
      "step": 6500
    },
    {
      "epoch": 3.25,
      "learning_rate": 4.956650503393613e-05,
      "loss": 0.6385,
      "step": 7000
    },
    {
      "epoch": 3.48,
      "learning_rate": 4.947571441100011e-05,
      "loss": 0.6568,
      "step": 7500
    },
    {
      "epoch": 3.71,
      "learning_rate": 4.937660479537447e-05,
      "loss": 0.6774,
      "step": 8000
    },
    {
      "epoch": 3.71,
      "eval_bleu": 27.8224,
      "eval_gen_len": 18.6267,
      "eval_loss": 1.2035210132598877,
      "eval_runtime": 1442.7545,
      "eval_samples_per_second": 11.947,
      "eval_steps_per_second": 1.494,
      "step": 8000
    }
  ],
  "logging_steps": 500,
  "max_steps": 86160,
  "num_train_epochs": 40,
  "save_steps": 4000,
  "total_flos": 1.109570441183232e+18,
  "trial_name": null,
  "trial_params": null
}