trainer_state.json

{
  "best_metric": 0.8776,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0203/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": 62.63008117675781,
      "learning_rate": 5e-05,
      "loss": 3.6795,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4437333333333333,
      "eval_loss": 3.111616849899292,
      "eval_runtime": 14.1428,
      "eval_samples_per_second": 265.153,
      "eval_steps_per_second": 4.172,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 186.1746826171875,
      "learning_rate": 5e-05,
      "loss": 1.865,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7218666666666667,
      "eval_loss": 1.1138335466384888,
      "eval_runtime": 13.4416,
      "eval_samples_per_second": 278.985,
      "eval_steps_per_second": 4.389,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 86.97637176513672,
      "learning_rate": 5e-05,
      "loss": 0.8874,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8192,
      "eval_loss": 0.6838088631629944,
      "eval_runtime": 13.5916,
      "eval_samples_per_second": 275.905,
      "eval_steps_per_second": 4.341,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 282.0440979003906,
      "learning_rate": 5e-05,
      "loss": 0.6109,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.5649511814117432,
      "eval_runtime": 13.5247,
      "eval_samples_per_second": 277.27,
      "eval_steps_per_second": 4.362,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 102.1645736694336,
      "learning_rate": 5e-05,
      "loss": 0.4833,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.47892311215400696,
      "eval_runtime": 13.6121,
      "eval_samples_per_second": 275.489,
      "eval_steps_per_second": 4.334,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 68.2385482788086,
      "learning_rate": 5e-05,
      "loss": 0.3958,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.46053510904312134,
      "eval_runtime": 13.6191,
      "eval_samples_per_second": 275.348,
      "eval_steps_per_second": 4.332,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 107.06715393066406,
      "learning_rate": 5e-05,
      "loss": 0.3363,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.42865774035453796,
      "eval_runtime": 13.3986,
      "eval_samples_per_second": 279.879,
      "eval_steps_per_second": 4.403,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}