trainer_state.json

{
  "best_metric": 0.8808,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0034/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": 72.13731384277344,
      "learning_rate": 7.2e-05,
      "loss": 2.9612,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7346666666666667,
      "eval_loss": 1.2076894044876099,
      "eval_runtime": 17.1758,
      "eval_samples_per_second": 218.331,
      "eval_steps_per_second": 3.435,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 102.64971923828125,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.7505,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.4950296878814697,
      "eval_runtime": 17.558,
      "eval_samples_per_second": 213.577,
      "eval_steps_per_second": 3.36,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 93.6780776977539,
      "learning_rate": 3.6e-05,
      "loss": 0.4307,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.40632206201553345,
      "eval_runtime": 17.2439,
      "eval_samples_per_second": 217.469,
      "eval_steps_per_second": 3.422,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 125.46685791015625,
      "learning_rate": 1.8e-05,
      "loss": 0.3387,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.39488154649734497,
      "eval_runtime": 17.2854,
      "eval_samples_per_second": 216.946,
      "eval_steps_per_second": 3.413,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 95.16941833496094,
      "learning_rate": 0.0,
      "loss": 0.2909,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.38486093282699585,
      "eval_runtime": 15.1783,
      "eval_samples_per_second": 247.064,
      "eval_steps_per_second": 3.887,
      "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
}