trainer_state.json

{
  "best_metric": 9.7241,
  "best_model_checkpoint": "saved/tobyoki-pairwise-wo_space/bart-base-japanese/BaseModel/checkpoint-3200",
  "epoch": 5.0,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.62,
      "learning_rate": 4.375e-05,
      "loss": 2.1248,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 13.69,
      "eval_loss": 1.9658905267715454,
      "eval_rouge1": 4.6735,
      "eval_rouge2": 1.3973,
      "eval_rougeL": 3.3399,
      "eval_rougeLsum": 3.6754,
      "eval_runtime": 20.4931,
      "eval_samples_per_second": 4.88,
      "eval_steps_per_second": 4.88,
      "step": 800
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.3046,
      "step": 1000
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.125e-05,
      "loss": 0.9881,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 13.09,
      "eval_loss": 2.0469071865081787,
      "eval_rouge1": 4.4604,
      "eval_rouge2": 0.6802,
      "eval_rougeL": 3.3883,
      "eval_rougeLsum": 3.5799,
      "eval_runtime": 46.7994,
      "eval_samples_per_second": 2.137,
      "eval_steps_per_second": 2.137,
      "step": 1600
    },
    {
      "epoch": 2.5,
      "learning_rate": 2.5e-05,
      "loss": 0.7473,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 22.42,
      "eval_loss": 2.3957085609436035,
      "eval_rouge1": 7.1848,
      "eval_rouge2": 1.4395,
      "eval_rougeL": 5.7287,
      "eval_rougeLsum": 6.234,
      "eval_runtime": 53.6886,
      "eval_samples_per_second": 1.863,
      "eval_steps_per_second": 1.863,
      "step": 2400
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.591,
      "step": 2500
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.25e-05,
      "loss": 0.5128,
      "step": 3000
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 24.21,
      "eval_loss": 2.401291608810425,
      "eval_rouge1": 9.7241,
      "eval_rouge2": 1.7342,
      "eval_rougeL": 7.7361,
      "eval_rougeLsum": 8.3869,
      "eval_runtime": 52.513,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 3200
    },
    {
      "epoch": 4.38,
      "learning_rate": 6.25e-06,
      "loss": 0.4408,
      "step": 3500
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0,
      "loss": 0.4115,
      "step": 4000
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 25.78,
      "eval_loss": 2.4646646976470947,
      "eval_rouge1": 9.203,
      "eval_rouge2": 1.326,
      "eval_rougeL": 7.3269,
      "eval_rougeLsum": 7.9726,
      "eval_runtime": 58.8067,
      "eval_samples_per_second": 1.7,
      "eval_steps_per_second": 1.7,
      "step": 4000
    },
    {
      "epoch": 5.0,
      "step": 4000,
      "total_flos": 675929268264960.0,
      "train_loss": 0.8901088180541992,
      "train_runtime": 815.2717,
      "train_samples_per_second": 4.906,
      "train_steps_per_second": 4.906
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 5,
  "total_flos": 675929268264960.0,
  "trial_name": null,
  "trial_params": null
}