last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.008594757198109154,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008594757198109154,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 3.5906,
      "step": 1
    },
    {
      "epoch": 0.0008594757198109154,
      "eval_loss": 4.292478561401367,
      "eval_runtime": 51.5431,
      "eval_samples_per_second": 4.753,
      "eval_steps_per_second": 4.753,
      "step": 1
    },
    {
      "epoch": 0.0017189514396218307,
      "grad_norm": 3.34263014793396,
      "learning_rate": 2e-05,
      "loss": 2.9538,
      "step": 2
    },
    {
      "epoch": 0.002578427159432746,
      "grad_norm": 6.1069183349609375,
      "learning_rate": 4e-05,
      "loss": 4.841,
      "step": 3
    },
    {
      "epoch": 0.002578427159432746,
      "eval_loss": 4.275597095489502,
      "eval_runtime": 51.1295,
      "eval_samples_per_second": 4.792,
      "eval_steps_per_second": 4.792,
      "step": 3
    },
    {
      "epoch": 0.0034379028792436614,
      "grad_norm": 7.2223992347717285,
      "learning_rate": 6e-05,
      "loss": 4.2997,
      "step": 4
    },
    {
      "epoch": 0.004297378599054577,
      "grad_norm": 4.632462978363037,
      "learning_rate": 8e-05,
      "loss": 4.3137,
      "step": 5
    },
    {
      "epoch": 0.005156854318865492,
      "grad_norm": 5.050290584564209,
      "learning_rate": 0.0001,
      "loss": 3.8527,
      "step": 6
    },
    {
      "epoch": 0.005156854318865492,
      "eval_loss": 3.8723323345184326,
      "eval_runtime": 51.1071,
      "eval_samples_per_second": 4.794,
      "eval_steps_per_second": 4.794,
      "step": 6
    },
    {
      "epoch": 0.006016330038676407,
      "grad_norm": 4.985762119293213,
      "learning_rate": 0.00012,
      "loss": 3.5523,
      "step": 7
    },
    {
      "epoch": 0.006875805758487323,
      "grad_norm": 5.663036823272705,
      "learning_rate": 0.00014,
      "loss": 3.8084,
      "step": 8
    },
    {
      "epoch": 0.007735281478298238,
      "grad_norm": 5.977775573730469,
      "learning_rate": 0.00016,
      "loss": 2.787,
      "step": 9
    },
    {
      "epoch": 0.007735281478298238,
      "eval_loss": 2.5680861473083496,
      "eval_runtime": 51.2453,
      "eval_samples_per_second": 4.781,
      "eval_steps_per_second": 4.781,
      "step": 9
    },
    {
      "epoch": 0.008594757198109154,
      "grad_norm": 6.121236324310303,
      "learning_rate": 0.00018,
      "loss": 2.1664,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7418872835604480.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}