trainer_state.json

{
  "best_metric": 0.8474666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0675/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": 115.87236785888672,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 3.7311,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4856,
      "eval_loss": 3.266024589538574,
      "eval_runtime": 8.0589,
      "eval_samples_per_second": 465.322,
      "eval_steps_per_second": 7.321,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 131.90057373046875,
      "learning_rate": 3.571428571428572e-05,
      "loss": 2.0812,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7202666666666667,
      "eval_loss": 1.2061145305633545,
      "eval_runtime": 8.0134,
      "eval_samples_per_second": 467.966,
      "eval_steps_per_second": 7.363,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 116.20150756835938,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.9714,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7994666666666667,
      "eval_loss": 0.7403157353401184,
      "eval_runtime": 8.1792,
      "eval_samples_per_second": 458.483,
      "eval_steps_per_second": 7.213,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 137.74749755859375,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.6965,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8237333333333333,
      "eval_loss": 0.6111857295036316,
      "eval_runtime": 8.6819,
      "eval_samples_per_second": 431.935,
      "eval_steps_per_second": 6.796,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 105.71757507324219,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.6022,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.844,
      "eval_loss": 0.5303592085838318,
      "eval_runtime": 7.9693,
      "eval_samples_per_second": 470.556,
      "eval_steps_per_second": 7.403,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 142.519287109375,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.5448,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.5124239325523376,
      "eval_runtime": 7.909,
      "eval_samples_per_second": 474.142,
      "eval_steps_per_second": 7.46,
      "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
}