trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 2352,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_bleu": 13.50642145973043,
      "eval_loss": 1.9098774194717407,
      "eval_runtime": 18.3505,
      "eval_samples_per_second": 1.199,
      "eval_steps_per_second": 1.199,
      "step": 196
    },
    {
      "epoch": 2.0,
      "eval_bleu": 15.388992406642911,
      "eval_loss": 1.5389823913574219,
      "eval_runtime": 17.9949,
      "eval_samples_per_second": 1.223,
      "eval_steps_per_second": 1.223,
      "step": 392
    },
    {
      "epoch": 2.5510204081632653,
      "grad_norm": 2.0644779205322266,
      "learning_rate": 7.891156462585034e-05,
      "loss": 2.4223,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_bleu": 16.361746081538808,
      "eval_loss": 1.4969135522842407,
      "eval_runtime": 18.102,
      "eval_samples_per_second": 1.215,
      "eval_steps_per_second": 1.215,
      "step": 588
    },
    {
      "epoch": 4.0,
      "eval_bleu": 16.192662011002973,
      "eval_loss": 1.558351993560791,
      "eval_runtime": 18.0485,
      "eval_samples_per_second": 1.219,
      "eval_steps_per_second": 1.219,
      "step": 784
    },
    {
      "epoch": 5.0,
      "eval_bleu": 20.388609774897613,
      "eval_loss": 1.6301382780075073,
      "eval_runtime": 17.8816,
      "eval_samples_per_second": 1.23,
      "eval_steps_per_second": 1.23,
      "step": 980
    },
    {
      "epoch": 5.1020408163265305,
      "grad_norm": 3.387509822845459,
      "learning_rate": 5.7653061224489805e-05,
      "loss": 0.5615,
      "step": 1000
    },
    {
      "epoch": 6.0,
      "eval_bleu": 20.820547846052477,
      "eval_loss": 1.6597435474395752,
      "eval_runtime": 18.4966,
      "eval_samples_per_second": 1.189,
      "eval_steps_per_second": 1.189,
      "step": 1176
    },
    {
      "epoch": 7.0,
      "eval_bleu": 20.333903309139888,
      "eval_loss": 1.710249662399292,
      "eval_runtime": 17.8589,
      "eval_samples_per_second": 1.232,
      "eval_steps_per_second": 1.232,
      "step": 1372
    },
    {
      "epoch": 7.653061224489796,
      "grad_norm": 2.134899377822876,
      "learning_rate": 3.639455782312925e-05,
      "loss": 0.2154,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "eval_bleu": 19.92413260385578,
      "eval_loss": 1.7504385709762573,
      "eval_runtime": 19.3501,
      "eval_samples_per_second": 1.137,
      "eval_steps_per_second": 1.137,
      "step": 1568
    },
    {
      "epoch": 9.0,
      "eval_bleu": 21.495754342436005,
      "eval_loss": 1.7941405773162842,
      "eval_runtime": 18.5215,
      "eval_samples_per_second": 1.188,
      "eval_steps_per_second": 1.188,
      "step": 1764
    },
    {
      "epoch": 10.0,
      "eval_bleu": 22.143240405081055,
      "eval_loss": 1.8230061531066895,
      "eval_runtime": 19.4583,
      "eval_samples_per_second": 1.131,
      "eval_steps_per_second": 1.131,
      "step": 1960
    },
    {
      "epoch": 10.204081632653061,
      "grad_norm": 0.1605977714061737,
      "learning_rate": 1.5136054421768709e-05,
      "loss": 0.1043,
      "step": 2000
    },
    {
      "epoch": 11.0,
      "eval_bleu": 21.334082933744533,
      "eval_loss": 1.8319286108016968,
      "eval_runtime": 18.1893,
      "eval_samples_per_second": 1.21,
      "eval_steps_per_second": 1.21,
      "step": 2156
    }
  ],
  "logging_steps": 500,
  "max_steps": 2352,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 637128753020928.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}