trainer_state.json

{
  "best_metric": 0.8805333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0045/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 71.15617370605469,
      "learning_rate": 0.0001,
      "loss": 2.8264,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7298666666666667,
      "eval_loss": 0.9967694282531738,
      "eval_runtime": 9.7685,
      "eval_samples_per_second": 383.886,
      "eval_steps_per_second": 6.04,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 85.65457916259766,
      "learning_rate": 0.0001,
      "loss": 0.7083,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8394666666666667,
      "eval_loss": 0.5425072312355042,
      "eval_runtime": 8.9677,
      "eval_samples_per_second": 418.167,
      "eval_steps_per_second": 6.579,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 61.90450668334961,
      "learning_rate": 0.0001,
      "loss": 0.4438,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8605333333333334,
      "eval_loss": 0.4732644855976105,
      "eval_runtime": 8.7513,
      "eval_samples_per_second": 428.506,
      "eval_steps_per_second": 6.742,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 64.16138458251953,
      "learning_rate": 0.0001,
      "loss": 0.3153,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.4290253818035126,
      "eval_runtime": 8.5597,
      "eval_samples_per_second": 438.1,
      "eval_steps_per_second": 6.893,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 47.978939056396484,
      "learning_rate": 0.0001,
      "loss": 0.2312,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4343804717063904,
      "eval_runtime": 8.7172,
      "eval_samples_per_second": 430.184,
      "eval_steps_per_second": 6.768,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 103.427734375,
      "learning_rate": 0.0001,
      "loss": 0.1831,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8749333333333333,
      "eval_loss": 0.45398589968681335,
      "eval_runtime": 8.9042,
      "eval_samples_per_second": 421.147,
      "eval_steps_per_second": 6.626,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 70.57642364501953,
      "learning_rate": 0.0001,
      "loss": 0.1321,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4624747931957245,
      "eval_runtime": 8.8348,
      "eval_samples_per_second": 424.458,
      "eval_steps_per_second": 6.678,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 80.42642974853516,
      "learning_rate": 0.0001,
      "loss": 0.1085,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4588300883769989,
      "eval_runtime": 8.4075,
      "eval_samples_per_second": 446.031,
      "eval_steps_per_second": 7.018,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 36.23624801635742,
      "learning_rate": 0.0001,
      "loss": 0.0815,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8805333333333333,
      "eval_loss": 0.48218756914138794,
      "eval_runtime": 8.6351,
      "eval_samples_per_second": 434.273,
      "eval_steps_per_second": 6.833,
      "step": 2997
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 7.3587863155968e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}