trainer_state.json

{
  "best_metric": 0.8682666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0859/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 140.69654846191406,
      "learning_rate": 7.714285714285714e-05,
      "loss": 2.757,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7381333333333333,
      "eval_loss": 1.055326223373413,
      "eval_runtime": 12.4314,
      "eval_samples_per_second": 301.656,
      "eval_steps_per_second": 4.746,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 144.78509521484375,
      "learning_rate": 6.428571428571429e-05,
      "loss": 0.7665,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8194666666666667,
      "eval_loss": 0.600494384765625,
      "eval_runtime": 12.7941,
      "eval_samples_per_second": 293.104,
      "eval_steps_per_second": 4.612,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 93.71951293945312,
      "learning_rate": 5.142857142857143e-05,
      "loss": 0.4756,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8450666666666666,
      "eval_loss": 0.4990963637828827,
      "eval_runtime": 12.5212,
      "eval_samples_per_second": 299.492,
      "eval_steps_per_second": 4.712,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 69.63736724853516,
      "learning_rate": 3.857142857142857e-05,
      "loss": 0.3483,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.4565032422542572,
      "eval_runtime": 12.4527,
      "eval_samples_per_second": 301.139,
      "eval_steps_per_second": 4.738,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 97.76102447509766,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.2673,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.864,
      "eval_loss": 0.45005273818969727,
      "eval_runtime": 12.3823,
      "eval_samples_per_second": 302.852,
      "eval_steps_per_second": 4.765,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 157.81263732910156,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 0.2186,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8682666666666666,
      "eval_loss": 0.43268465995788574,
      "eval_runtime": 12.6496,
      "eval_samples_per_second": 296.453,
      "eval_steps_per_second": 4.664,
      "step": 1998
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9058575437312e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}