trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9774436090225564,
  "eval_steps": 500,
  "global_step": 99,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3007518796992481,
      "grad_norm": 6.465223184133898,
      "learning_rate": 5e-06,
      "loss": 0.8179,
      "step": 10
    },
    {
      "epoch": 0.6015037593984962,
      "grad_norm": 10.978444836397953,
      "learning_rate": 5e-06,
      "loss": 0.6978,
      "step": 20
    },
    {
      "epoch": 0.9022556390977443,
      "grad_norm": 15.247842813330424,
      "learning_rate": 5e-06,
      "loss": 0.6638,
      "step": 30
    },
    {
      "epoch": 0.9924812030075187,
      "eval_loss": 0.6395203471183777,
      "eval_runtime": 23.4207,
      "eval_samples_per_second": 38.257,
      "eval_steps_per_second": 0.598,
      "step": 33
    },
    {
      "epoch": 1.2030075187969924,
      "grad_norm": 2.1050038369400648,
      "learning_rate": 5e-06,
      "loss": 0.6771,
      "step": 40
    },
    {
      "epoch": 1.5037593984962405,
      "grad_norm": 3.3978718073083334,
      "learning_rate": 5e-06,
      "loss": 0.5989,
      "step": 50
    },
    {
      "epoch": 1.8045112781954886,
      "grad_norm": 2.9934882379275236,
      "learning_rate": 5e-06,
      "loss": 0.5863,
      "step": 60
    },
    {
      "epoch": 1.9849624060150375,
      "eval_loss": 0.6017050743103027,
      "eval_runtime": 23.2997,
      "eval_samples_per_second": 38.455,
      "eval_steps_per_second": 0.601,
      "step": 66
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 1.251564777495961,
      "learning_rate": 5e-06,
      "loss": 0.6073,
      "step": 70
    },
    {
      "epoch": 2.406015037593985,
      "grad_norm": 1.1824873260326954,
      "learning_rate": 5e-06,
      "loss": 0.5368,
      "step": 80
    },
    {
      "epoch": 2.706766917293233,
      "grad_norm": 0.9602307006188483,
      "learning_rate": 5e-06,
      "loss": 0.5238,
      "step": 90
    },
    {
      "epoch": 2.9774436090225564,
      "eval_loss": 0.5830685496330261,
      "eval_runtime": 22.3026,
      "eval_samples_per_second": 40.175,
      "eval_steps_per_second": 0.628,
      "step": 99
    },
    {
      "epoch": 2.9774436090225564,
      "step": 99,
      "total_flos": 165619307642880.0,
      "train_loss": 0.6243041982554426,
      "train_runtime": 3527.761,
      "train_samples_per_second": 14.47,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 99,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 165619307642880.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}