trainer_state.json

{
  "best_metric": 0.8432,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0984/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 71.05562591552734,
      "learning_rate": 3e-05,
      "loss": 3.8154,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4648,
      "eval_loss": 3.6025619506835938,
      "eval_runtime": 8.0644,
      "eval_samples_per_second": 465.009,
      "eval_steps_per_second": 7.316,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 204.97000122070312,
      "learning_rate": 3e-05,
      "loss": 2.793,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6610666666666667,
      "eval_loss": 1.683272123336792,
      "eval_runtime": 8.0171,
      "eval_samples_per_second": 467.75,
      "eval_steps_per_second": 7.359,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 127.4854736328125,
      "learning_rate": 3e-05,
      "loss": 1.2146,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7754666666666666,
      "eval_loss": 0.8469237089157104,
      "eval_runtime": 8.0743,
      "eval_samples_per_second": 464.434,
      "eval_steps_per_second": 7.307,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 194.16033935546875,
      "learning_rate": 3e-05,
      "loss": 0.7547,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8205333333333333,
      "eval_loss": 0.5997509360313416,
      "eval_runtime": 7.9842,
      "eval_samples_per_second": 469.68,
      "eval_steps_per_second": 7.39,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 142.62339782714844,
      "learning_rate": 3e-05,
      "loss": 0.5762,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8432,
      "eval_loss": 0.5175759792327881,
      "eval_runtime": 8.1188,
      "eval_samples_per_second": 461.893,
      "eval_steps_per_second": 7.267,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 4.088214619776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}