trainer_state.json

{
  "best_metric": 0.8570666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0305/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": 131.26731872558594,
      "learning_rate": 3e-05,
      "loss": 3.8445,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4224,
      "eval_loss": 3.694094181060791,
      "eval_runtime": 8.1415,
      "eval_samples_per_second": 460.605,
      "eval_steps_per_second": 7.247,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 82.4087142944336,
      "learning_rate": 3e-05,
      "loss": 2.9885,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6264,
      "eval_loss": 2.1510961055755615,
      "eval_runtime": 8.0969,
      "eval_samples_per_second": 463.138,
      "eval_steps_per_second": 7.287,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 241.3684844970703,
      "learning_rate": 3e-05,
      "loss": 1.3093,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7848,
      "eval_loss": 0.8788554668426514,
      "eval_runtime": 8.0405,
      "eval_samples_per_second": 466.388,
      "eval_steps_per_second": 7.338,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 184.693115234375,
      "learning_rate": 3e-05,
      "loss": 0.6916,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8288,
      "eval_loss": 0.5998766422271729,
      "eval_runtime": 8.1273,
      "eval_samples_per_second": 461.409,
      "eval_steps_per_second": 7.259,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 346.4747314453125,
      "learning_rate": 3e-05,
      "loss": 0.4846,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.5146359801292419,
      "eval_runtime": 8.03,
      "eval_samples_per_second": 466.997,
      "eval_steps_per_second": 7.347,
      "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
}