trainer_state.json

{
  "best_metric": 0.8872,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0668/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": 63.71931838989258,
      "learning_rate": 9e-05,
      "loss": 2.7333,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7488,
      "eval_loss": 0.8967779874801636,
      "eval_runtime": 8.089,
      "eval_samples_per_second": 463.591,
      "eval_steps_per_second": 7.294,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 84.29446411132812,
      "learning_rate": 9e-05,
      "loss": 0.6649,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8381333333333333,
      "eval_loss": 0.5393460988998413,
      "eval_runtime": 8.1783,
      "eval_samples_per_second": 458.531,
      "eval_steps_per_second": 7.214,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 268.5426940917969,
      "learning_rate": 9e-05,
      "loss": 0.4043,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8586666666666667,
      "eval_loss": 0.4365205764770508,
      "eval_runtime": 8.2799,
      "eval_samples_per_second": 452.906,
      "eval_steps_per_second": 7.126,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 96.0676040649414,
      "learning_rate": 9e-05,
      "loss": 0.2639,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8738666666666667,
      "eval_loss": 0.41490378975868225,
      "eval_runtime": 8.4571,
      "eval_samples_per_second": 443.414,
      "eval_steps_per_second": 6.976,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 141.33059692382812,
      "learning_rate": 9e-05,
      "loss": 0.1801,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8738666666666667,
      "eval_loss": 0.4219409227371216,
      "eval_runtime": 8.0287,
      "eval_samples_per_second": 467.073,
      "eval_steps_per_second": 7.349,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 199.8916778564453,
      "learning_rate": 9e-05,
      "loss": 0.1228,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.41972851753234863,
      "eval_runtime": 8.1146,
      "eval_samples_per_second": 462.13,
      "eval_steps_per_second": 7.271,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 65.55452728271484,
      "learning_rate": 9e-05,
      "loss": 0.0894,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.4328984320163727,
      "eval_runtime": 8.1041,
      "eval_samples_per_second": 462.728,
      "eval_steps_per_second": 7.28,
      "step": 2331
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}