last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7933884297520661,
  "eval_steps": 4,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06611570247933884,
      "grad_norm": 7.201543807983398,
      "learning_rate": 1e-05,
      "loss": 4.515,
      "step": 1
    },
    {
      "epoch": 0.06611570247933884,
      "eval_loss": 4.704464912414551,
      "eval_runtime": 14.7628,
      "eval_samples_per_second": 1.761,
      "eval_steps_per_second": 0.881,
      "step": 1
    },
    {
      "epoch": 0.1322314049586777,
      "grad_norm": 7.516412734985352,
      "learning_rate": 2e-05,
      "loss": 5.1072,
      "step": 2
    },
    {
      "epoch": 0.19834710743801653,
      "grad_norm": 7.386772155761719,
      "learning_rate": 3e-05,
      "loss": 4.8947,
      "step": 3
    },
    {
      "epoch": 0.2644628099173554,
      "grad_norm": 8.112839698791504,
      "learning_rate": 4e-05,
      "loss": 4.5637,
      "step": 4
    },
    {
      "epoch": 0.2644628099173554,
      "eval_loss": 4.352102756500244,
      "eval_runtime": 4.2849,
      "eval_samples_per_second": 6.068,
      "eval_steps_per_second": 3.034,
      "step": 4
    },
    {
      "epoch": 0.3305785123966942,
      "grad_norm": 7.6727705001831055,
      "learning_rate": 5e-05,
      "loss": 4.5121,
      "step": 5
    },
    {
      "epoch": 0.39669421487603307,
      "grad_norm": 7.3341240882873535,
      "learning_rate": 6e-05,
      "loss": 4.0657,
      "step": 6
    },
    {
      "epoch": 0.4628099173553719,
      "grad_norm": 7.578364372253418,
      "learning_rate": 7e-05,
      "loss": 3.9716,
      "step": 7
    },
    {
      "epoch": 0.5289256198347108,
      "grad_norm": 6.718101978302002,
      "learning_rate": 8e-05,
      "loss": 3.4257,
      "step": 8
    },
    {
      "epoch": 0.5289256198347108,
      "eval_loss": 3.077786445617676,
      "eval_runtime": 4.3038,
      "eval_samples_per_second": 6.041,
      "eval_steps_per_second": 3.021,
      "step": 8
    },
    {
      "epoch": 0.5950413223140496,
      "grad_norm": 4.271751403808594,
      "learning_rate": 9e-05,
      "loss": 2.8224,
      "step": 9
    },
    {
      "epoch": 0.6611570247933884,
      "grad_norm": 4.623085975646973,
      "learning_rate": 0.0001,
      "loss": 3.1332,
      "step": 10
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 7.002574920654297,
      "learning_rate": 9.330127018922194e-05,
      "loss": 2.8715,
      "step": 11
    },
    {
      "epoch": 0.7933884297520661,
      "grad_norm": 10.354272842407227,
      "learning_rate": 7.500000000000001e-05,
      "loss": 3.1827,
      "step": 12
    },
    {
      "epoch": 0.7933884297520661,
      "eval_loss": 2.666003942489624,
      "eval_runtime": 4.2925,
      "eval_samples_per_second": 6.057,
      "eval_steps_per_second": 3.029,
      "step": 12
    }
  ],
  "logging_steps": 1,
  "max_steps": 16,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 4,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.362185870809498e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}