trainer_state.json

{
  "best_metric": 0.9069333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0785/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": 51.99765396118164,
      "learning_rate": 0.0004849231551964771,
      "loss": 1.3505,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8373333333333334,
      "eval_loss": 0.552812397480011,
      "eval_runtime": 15.1686,
      "eval_samples_per_second": 247.222,
      "eval_steps_per_second": 3.89,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 20.645851135253906,
      "learning_rate": 0.0004415111107797445,
      "loss": 0.4543,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.46782806515693665,
      "eval_runtime": 15.1675,
      "eval_samples_per_second": 247.239,
      "eval_steps_per_second": 3.89,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 30.825429916381836,
      "learning_rate": 0.000375,
      "loss": 0.2601,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.4789236783981323,
      "eval_runtime": 15.159,
      "eval_samples_per_second": 247.378,
      "eval_steps_per_second": 3.892,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 40.26069641113281,
      "learning_rate": 0.00029341204441673266,
      "loss": 0.1539,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.4734314978122711,
      "eval_runtime": 15.2095,
      "eval_samples_per_second": 246.556,
      "eval_steps_per_second": 3.879,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 28.817977905273438,
      "learning_rate": 0.00020658795558326743,
      "loss": 0.0917,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.4449414014816284,
      "eval_runtime": 14.3723,
      "eval_samples_per_second": 260.918,
      "eval_steps_per_second": 4.105,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 25.196884155273438,
      "learning_rate": 0.00012500000000000006,
      "loss": 0.0513,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9,
      "eval_loss": 0.4346129596233368,
      "eval_runtime": 14.1601,
      "eval_samples_per_second": 264.828,
      "eval_steps_per_second": 4.167,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 28.46689224243164,
      "learning_rate": 5.848888922025553e-05,
      "loss": 0.0346,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9069333333333334,
      "eval_loss": 0.41077345609664917,
      "eval_runtime": 14.4835,
      "eval_samples_per_second": 258.916,
      "eval_steps_per_second": 4.074,
      "step": 2331
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}