checkpoint-1240/trainer_state.json

{
  "best_metric": 5.6794586181640625,
  "best_model_checkpoint": "bill_sum_finetune_test_gpt2_3/checkpoint-496",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_gen_len": 25.0,
      "eval_loss": 5.72403621673584,
      "eval_rouge1": 0.1624,
      "eval_rouge2": 0.0178,
      "eval_rougeL": 0.1416,
      "eval_rougeLsum": 0.1514,
      "eval_runtime": 5.8118,
      "eval_samples_per_second": 42.672,
      "eval_steps_per_second": 10.668,
      "step": 248
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 25.0,
      "eval_loss": 5.6794586181640625,
      "eval_rouge1": 0.1736,
      "eval_rouge2": 0.0188,
      "eval_rougeL": 0.1512,
      "eval_rougeLsum": 0.1618,
      "eval_runtime": 5.905,
      "eval_samples_per_second": 41.998,
      "eval_steps_per_second": 10.499,
      "step": 496
    },
    {
      "epoch": 2.02,
      "grad_norm": 1349184.375,
      "learning_rate": 3.580645161290323e-05,
      "loss": 5.8134,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 25.0,
      "eval_loss": 5.689182281494141,
      "eval_rouge1": 0.1636,
      "eval_rouge2": 0.0174,
      "eval_rougeL": 0.1451,
      "eval_rougeLsum": 0.156,
      "eval_runtime": 6.059,
      "eval_samples_per_second": 40.931,
      "eval_steps_per_second": 10.233,
      "step": 744
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 25.0,
      "eval_loss": 5.705079078674316,
      "eval_rouge1": 0.1625,
      "eval_rouge2": 0.0178,
      "eval_rougeL": 0.1444,
      "eval_rougeLsum": 0.1549,
      "eval_runtime": 5.7853,
      "eval_samples_per_second": 42.868,
      "eval_steps_per_second": 10.717,
      "step": 992
    },
    {
      "epoch": 4.03,
      "grad_norm": 1369836.5,
      "learning_rate": 1.1612903225806451e-05,
      "loss": 5.3454,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 25.0,
      "eval_loss": 5.706656455993652,
      "eval_rouge1": 0.1625,
      "eval_rouge2": 0.0178,
      "eval_rougeL": 0.1444,
      "eval_rougeLsum": 0.1549,
      "eval_runtime": 5.7722,
      "eval_samples_per_second": 42.965,
      "eval_steps_per_second": 10.741,
      "step": 1240
    }
  ],
  "logging_steps": 500,
  "max_steps": 1240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 50472230400000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}