trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_bleu": 58.9649,
      "eval_gen_len": 19.44,
      "eval_loss": 0.6476399302482605,
      "eval_runtime": 7.9533,
      "eval_samples_per_second": 12.573,
      "eval_steps_per_second": 6.287,
      "step": 400
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.6755,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_bleu": 59.7239,
      "eval_gen_len": 19.21,
      "eval_loss": 0.6683109998703003,
      "eval_runtime": 7.5813,
      "eval_samples_per_second": 13.19,
      "eval_steps_per_second": 6.595,
      "step": 800
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2568,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_bleu": 59.3781,
      "eval_gen_len": 19.22,
      "eval_loss": 0.7284716963768005,
      "eval_runtime": 7.5691,
      "eval_samples_per_second": 13.212,
      "eval_steps_per_second": 6.606,
      "step": 1200
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.25e-05,
      "loss": 0.1116,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_bleu": 60.1615,
      "eval_gen_len": 19.14,
      "eval_loss": 0.8120526075363159,
      "eval_runtime": 7.4229,
      "eval_samples_per_second": 13.472,
      "eval_steps_per_second": 6.736,
      "step": 1600
    },
    {
      "epoch": 5.0,
      "learning_rate": 1e-05,
      "loss": 0.048,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_bleu": 58.8865,
      "eval_gen_len": 19.34,
      "eval_loss": 0.8653954267501831,
      "eval_runtime": 8.7147,
      "eval_samples_per_second": 11.475,
      "eval_steps_per_second": 5.737,
      "step": 2000
    },
    {
      "epoch": 6.0,
      "eval_bleu": 58.2232,
      "eval_gen_len": 19.07,
      "eval_loss": 0.8903241157531738,
      "eval_runtime": 8.0705,
      "eval_samples_per_second": 12.391,
      "eval_steps_per_second": 6.195,
      "step": 2400
    },
    {
      "epoch": 6.25,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.0178,
      "step": 2500
    },
    {
      "epoch": 7.0,
      "eval_bleu": 58.0224,
      "eval_gen_len": 19.22,
      "eval_loss": 0.9194047451019287,
      "eval_runtime": 8.0477,
      "eval_samples_per_second": 12.426,
      "eval_steps_per_second": 6.213,
      "step": 2800
    },
    {
      "epoch": 7.5,
      "learning_rate": 5e-06,
      "loss": 0.0087,
      "step": 3000
    },
    {
      "epoch": 8.0,
      "eval_bleu": 58.8822,
      "eval_gen_len": 19.29,
      "eval_loss": 0.930671751499176,
      "eval_runtime": 7.9933,
      "eval_samples_per_second": 12.51,
      "eval_steps_per_second": 6.255,
      "step": 3200
    },
    {
      "epoch": 8.75,
      "learning_rate": 2.5e-06,
      "loss": 0.0039,
      "step": 3500
    },
    {
      "epoch": 9.0,
      "eval_bleu": 59.1049,
      "eval_gen_len": 19.18,
      "eval_loss": 0.9447950124740601,
      "eval_runtime": 7.8845,
      "eval_samples_per_second": 12.683,
      "eval_steps_per_second": 6.342,
      "step": 3600
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.0,
      "loss": 0.0027,
      "step": 4000
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 10,
  "total_flos": 199212586500096.0,
  "trial_name": null,
  "trial_params": null
}