trainer_state.json

{
  "best_metric": 0.8504,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0070/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": 171.94830322265625,
      "learning_rate": 5e-05,
      "loss": 3.6395,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4928,
      "eval_loss": 2.9956977367401123,
      "eval_runtime": 13.5713,
      "eval_samples_per_second": 276.319,
      "eval_steps_per_second": 4.347,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 129.51856994628906,
      "learning_rate": 5e-05,
      "loss": 1.6623,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7544,
      "eval_loss": 0.9169057011604309,
      "eval_runtime": 13.2249,
      "eval_samples_per_second": 283.556,
      "eval_steps_per_second": 4.461,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 102.16647338867188,
      "learning_rate": 5e-05,
      "loss": 0.7645,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8149333333333333,
      "eval_loss": 0.6371418237686157,
      "eval_runtime": 12.9856,
      "eval_samples_per_second": 288.782,
      "eval_steps_per_second": 4.543,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 188.4789276123047,
      "learning_rate": 5e-05,
      "loss": 0.56,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8304,
      "eval_loss": 0.5535508990287781,
      "eval_runtime": 13.2174,
      "eval_samples_per_second": 283.717,
      "eval_steps_per_second": 4.464,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 79.81751251220703,
      "learning_rate": 5e-05,
      "loss": 0.4461,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8504,
      "eval_loss": 0.4851263463497162,
      "eval_runtime": 12.968,
      "eval_samples_per_second": 289.173,
      "eval_steps_per_second": 4.55,
      "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
}