trainer_state.json

{
  "best_metric": 0.8853333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0277/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": 89.16474151611328,
      "learning_rate": 6.222222222222222e-05,
      "loss": 3.179,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6426666666666667,
      "eval_loss": 1.6181687116622925,
      "eval_runtime": 7.395,
      "eval_samples_per_second": 507.101,
      "eval_steps_per_second": 7.978,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 155.06361389160156,
      "learning_rate": 5.444444444444444e-05,
      "loss": 0.9462,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.812,
      "eval_loss": 0.6519731879234314,
      "eval_runtime": 7.16,
      "eval_samples_per_second": 523.747,
      "eval_steps_per_second": 8.24,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 318.2770080566406,
      "learning_rate": 4.666666666666666e-05,
      "loss": 0.5599,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5054487586021423,
      "eval_runtime": 7.1591,
      "eval_samples_per_second": 523.812,
      "eval_steps_per_second": 8.241,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 140.73019409179688,
      "learning_rate": 3.8888888888888884e-05,
      "loss": 0.41,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.4326043725013733,
      "eval_runtime": 7.1173,
      "eval_samples_per_second": 526.882,
      "eval_steps_per_second": 8.29,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 77.55543518066406,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.3247,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8773333333333333,
      "eval_loss": 0.40490350127220154,
      "eval_runtime": 7.1115,
      "eval_samples_per_second": 527.313,
      "eval_steps_per_second": 8.296,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 104.19577026367188,
      "learning_rate": 2.333333333333333e-05,
      "loss": 0.2754,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8802666666666666,
      "eval_loss": 0.38352170586586,
      "eval_runtime": 7.1482,
      "eval_samples_per_second": 524.607,
      "eval_steps_per_second": 8.254,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 94.6551284790039,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.24,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8853333333333333,
      "eval_loss": 0.37526604533195496,
      "eval_runtime": 7.0773,
      "eval_samples_per_second": 529.86,
      "eval_steps_per_second": 8.336,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}