trainer_state.json

{
  "best_metric": 1.9379221200942993,
  "best_model_checkpoint": "./pretrain_en_zh_ai_challenger_output/checkpoint-781250",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 781250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.5240626335144043,
      "learning_rate": 4e-05,
      "loss": 2.2886,
      "step": 156250
    },
    {
      "epoch": 1.0,
      "eval_bleu": 65.1489,
      "eval_gen_len": 12.8416,
      "eval_loss": 2.096710443496704,
      "eval_meteor": 0.1576,
      "eval_runtime": 196.7178,
      "eval_samples_per_second": 40.667,
      "eval_steps_per_second": 0.635,
      "step": 156250
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.2112767696380615,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 312500
    },
    {
      "epoch": 2.0,
      "eval_bleu": 65.1489,
      "eval_gen_len": 12.7059,
      "eval_loss": 2.0353355407714844,
      "eval_meteor": 0.1591,
      "eval_runtime": 198.1849,
      "eval_samples_per_second": 40.366,
      "eval_steps_per_second": 0.631,
      "step": 312500
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.3523008823394775,
      "learning_rate": 2e-05,
      "loss": 2.05,
      "step": 468750
    },
    {
      "epoch": 3.0,
      "eval_bleu": 65.1489,
      "eval_gen_len": 12.7766,
      "eval_loss": 1.987796664237976,
      "eval_meteor": 0.1612,
      "eval_runtime": 201.5597,
      "eval_samples_per_second": 39.69,
      "eval_steps_per_second": 0.62,
      "step": 468750
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.332340717315674,
      "learning_rate": 1e-05,
      "loss": 1.9868,
      "step": 625000
    },
    {
      "epoch": 4.0,
      "eval_bleu": 65.1489,
      "eval_gen_len": 12.7332,
      "eval_loss": 1.9567323923110962,
      "eval_meteor": 0.1619,
      "eval_runtime": 189.3875,
      "eval_samples_per_second": 42.241,
      "eval_steps_per_second": 0.66,
      "step": 625000
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.409123420715332,
      "learning_rate": 0.0,
      "loss": 1.9345,
      "step": 781250
    },
    {
      "epoch": 5.0,
      "eval_bleu": 65.1489,
      "eval_gen_len": 12.7575,
      "eval_loss": 1.9379221200942993,
      "eval_meteor": 0.1615,
      "eval_runtime": 205.2631,
      "eval_samples_per_second": 38.974,
      "eval_steps_per_second": 0.609,
      "step": 781250
    },
    {
      "epoch": 5.0,
      "step": 781250,
      "total_flos": 6.787036896905134e+17,
      "train_loss": 2.07831184,
      "train_runtime": 195294.7701,
      "train_samples_per_second": 256.023,
      "train_steps_per_second": 4.0
    }
  ],
  "logging_steps": 500,
  "max_steps": 781250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 6.787036896905134e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}