trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.976,
  "eval_steps": 500,
  "global_step": 93,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 14.676734872106179,
      "learning_rate": 5e-06,
      "loss": 1.2289,
      "step": 10
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.598401054075069,
      "learning_rate": 5e-06,
      "loss": 1.1085,
      "step": 20
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.323246185800818,
      "learning_rate": 5e-06,
      "loss": 1.0621,
      "step": 30
    },
    {
      "epoch": 0.992,
      "eval_loss": 1.0408542156219482,
      "eval_runtime": 23.3229,
      "eval_samples_per_second": 36.059,
      "eval_steps_per_second": 0.6,
      "step": 31
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.634281663066763,
      "learning_rate": 5e-06,
      "loss": 1.0841,
      "step": 40
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5489112716620952,
      "learning_rate": 5e-06,
      "loss": 0.9873,
      "step": 50
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.6581760731689397,
      "learning_rate": 5e-06,
      "loss": 0.9711,
      "step": 60
    },
    {
      "epoch": 1.984,
      "eval_loss": 0.9961642026901245,
      "eval_runtime": 22.996,
      "eval_samples_per_second": 36.572,
      "eval_steps_per_second": 0.609,
      "step": 62
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.4201489917654573,
      "learning_rate": 5e-06,
      "loss": 1.008,
      "step": 70
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.2773128474148137,
      "learning_rate": 5e-06,
      "loss": 0.9094,
      "step": 80
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.2888597776168031,
      "learning_rate": 5e-06,
      "loss": 0.9089,
      "step": 90
    },
    {
      "epoch": 2.976,
      "eval_loss": 0.9772511720657349,
      "eval_runtime": 21.6003,
      "eval_samples_per_second": 38.935,
      "eval_steps_per_second": 0.648,
      "step": 93
    },
    {
      "epoch": 2.976,
      "step": 93,
      "total_flos": 155569084170240.0,
      "train_loss": 1.0259817133667648,
      "train_runtime": 3317.6169,
      "train_samples_per_second": 14.443,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 93,
  "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": 155569084170240.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}