trainer_state.json

{
  "best_metric": 0.8776,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0002/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 105.36347198486328,
      "learning_rate": 5e-05,
      "loss": 3.6968,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5034666666666666,
      "eval_loss": 3.105559825897217,
      "eval_runtime": 8.6446,
      "eval_samples_per_second": 433.795,
      "eval_steps_per_second": 6.825,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 101.71227264404297,
      "learning_rate": 5e-05,
      "loss": 1.6888,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7496,
      "eval_loss": 0.9344081282615662,
      "eval_runtime": 7.7113,
      "eval_samples_per_second": 486.301,
      "eval_steps_per_second": 7.651,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 177.6521453857422,
      "learning_rate": 5e-05,
      "loss": 0.7294,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8232,
      "eval_loss": 0.5934777855873108,
      "eval_runtime": 7.6659,
      "eval_samples_per_second": 489.179,
      "eval_steps_per_second": 7.696,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 156.4947967529297,
      "learning_rate": 5e-05,
      "loss": 0.4938,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8490666666666666,
      "eval_loss": 0.49322909116744995,
      "eval_runtime": 7.6777,
      "eval_samples_per_second": 488.43,
      "eval_steps_per_second": 7.685,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 231.85943603515625,
      "learning_rate": 5e-05,
      "loss": 0.3837,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.4583577513694763,
      "eval_runtime": 7.7253,
      "eval_samples_per_second": 485.418,
      "eval_steps_per_second": 7.637,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 65.17208099365234,
      "learning_rate": 5e-05,
      "loss": 0.295,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.4200899004936218,
      "eval_runtime": 7.6121,
      "eval_samples_per_second": 492.635,
      "eval_steps_per_second": 7.751,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 95.08432006835938,
      "learning_rate": 5e-05,
      "loss": 0.249,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.4055381715297699,
      "eval_runtime": 7.7896,
      "eval_samples_per_second": 481.413,
      "eval_steps_per_second": 7.574,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}