trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 570,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.5840111374855042,
      "learning_rate": 1.8280701754385966e-05,
      "loss": 1.1383,
      "mean_token_accuracy": 0.7269072496891021,
      "num_tokens": 118106.0,
      "step": 50
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.2792308032512665,
      "learning_rate": 1.6526315789473686e-05,
      "loss": 0.9902,
      "mean_token_accuracy": 0.7435696792602539,
      "num_tokens": 234972.0,
      "step": 100
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.2899998426437378,
      "learning_rate": 1.4771929824561405e-05,
      "loss": 0.9532,
      "mean_token_accuracy": 0.7494280987977981,
      "num_tokens": 353907.0,
      "step": 150
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.30872082710266113,
      "learning_rate": 1.3017543859649125e-05,
      "loss": 0.919,
      "mean_token_accuracy": 0.7515459644794464,
      "num_tokens": 471655.0,
      "step": 200
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.44872578978538513,
      "learning_rate": 1.1263157894736843e-05,
      "loss": 0.892,
      "mean_token_accuracy": 0.7596375972032547,
      "num_tokens": 591976.0,
      "step": 250
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.37252089381217957,
      "learning_rate": 9.508771929824562e-06,
      "loss": 0.8696,
      "mean_token_accuracy": 0.7662665128707886,
      "num_tokens": 708164.0,
      "step": 300
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.4809369146823883,
      "learning_rate": 7.754385964912282e-06,
      "loss": 0.8524,
      "mean_token_accuracy": 0.7682689487934112,
      "num_tokens": 827826.0,
      "step": 350
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.4046804904937744,
      "learning_rate": 6e-06,
      "loss": 0.8722,
      "mean_token_accuracy": 0.7630033561587334,
      "num_tokens": 944070.0,
      "step": 400
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 0.45627760887145996,
      "learning_rate": 4.245614035087719e-06,
      "loss": 0.8453,
      "mean_token_accuracy": 0.7691507586836814,
      "num_tokens": 1065956.0,
      "step": 450
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.5287462472915649,
      "learning_rate": 2.4912280701754387e-06,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7708389762043953,
      "num_tokens": 1182307.0,
      "step": 500
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 0.4097572863101959,
      "learning_rate": 7.368421052631579e-07,
      "loss": 0.8365,
      "mean_token_accuracy": 0.7708183565735817,
      "num_tokens": 1299425.0,
      "step": 550
    }
  ],
  "logging_steps": 50,
  "max_steps": 570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 15,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.012775913382912e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}