trainer_state.json

{
  "best_metric": 0.8917333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0238/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": 108.97286987304688,
      "learning_rate": 8.75e-05,
      "loss": 2.8677,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7274666666666667,
      "eval_loss": 1.0429983139038086,
      "eval_runtime": 13.0402,
      "eval_samples_per_second": 287.571,
      "eval_steps_per_second": 4.524,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 91.49505615234375,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.7803,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8328,
      "eval_loss": 0.5610671639442444,
      "eval_runtime": 12.9416,
      "eval_samples_per_second": 289.763,
      "eval_steps_per_second": 4.559,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 101.4925308227539,
      "learning_rate": 6.25e-05,
      "loss": 0.4858,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8602666666666666,
      "eval_loss": 0.4613882899284363,
      "eval_runtime": 13.0168,
      "eval_samples_per_second": 288.09,
      "eval_steps_per_second": 4.533,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 67.69094848632812,
      "learning_rate": 5e-05,
      "loss": 0.3616,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.3972567319869995,
      "eval_runtime": 13.1546,
      "eval_samples_per_second": 285.071,
      "eval_steps_per_second": 4.485,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 45.42450714111328,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.2994,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8856,
      "eval_loss": 0.3840443789958954,
      "eval_runtime": 13.1264,
      "eval_samples_per_second": 285.684,
      "eval_steps_per_second": 4.495,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 61.80015182495117,
      "learning_rate": 2.5e-05,
      "loss": 0.2519,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.38566136360168457,
      "eval_runtime": 13.295,
      "eval_samples_per_second": 282.061,
      "eval_steps_per_second": 4.438,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 110.72581481933594,
      "learning_rate": 1.25e-05,
      "loss": 0.2109,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8917333333333334,
      "eval_loss": 0.3625173270702362,
      "eval_runtime": 12.8746,
      "eval_samples_per_second": 291.271,
      "eval_steps_per_second": 4.583,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}