trainer_state.json

{
  "best_metric": 0.8674666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0592/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": 95.26696014404297,
      "learning_rate": 4.849231551964771e-05,
      "loss": 3.6477,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5056,
      "eval_loss": 2.936440944671631,
      "eval_runtime": 13.0219,
      "eval_samples_per_second": 287.977,
      "eval_steps_per_second": 4.531,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 83.31976318359375,
      "learning_rate": 4.415111107797445e-05,
      "loss": 1.713,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7541333333333333,
      "eval_loss": 0.9426048994064331,
      "eval_runtime": 12.9917,
      "eval_samples_per_second": 288.646,
      "eval_steps_per_second": 4.541,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 152.92611694335938,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.7748,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8152,
      "eval_loss": 0.623187780380249,
      "eval_runtime": 13.0436,
      "eval_samples_per_second": 287.498,
      "eval_steps_per_second": 4.523,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 145.85153198242188,
      "learning_rate": 2.9341204441673266e-05,
      "loss": 0.5818,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8432,
      "eval_loss": 0.512504518032074,
      "eval_runtime": 13.1999,
      "eval_samples_per_second": 284.093,
      "eval_steps_per_second": 4.47,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 76.13021850585938,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.4895,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8536,
      "eval_loss": 0.4737875759601593,
      "eval_runtime": 12.9268,
      "eval_samples_per_second": 290.095,
      "eval_steps_per_second": 4.564,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 191.00933837890625,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.4341,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8618666666666667,
      "eval_loss": 0.4479192793369293,
      "eval_runtime": 13.1291,
      "eval_samples_per_second": 285.626,
      "eval_steps_per_second": 4.494,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 62.442134857177734,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.3999,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.4456882178783417,
      "eval_runtime": 13.0378,
      "eval_samples_per_second": 287.625,
      "eval_steps_per_second": 4.525,
      "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
}