trainer_state.json

{
  "best_metric": 0.9325333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0161/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": 39.31516647338867,
      "learning_rate": 0.0002625,
      "loss": 1.3943,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.43095889687538147,
      "eval_runtime": 14.3967,
      "eval_samples_per_second": 260.477,
      "eval_steps_per_second": 4.098,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 38.080204010009766,
      "learning_rate": 0.000225,
      "loss": 0.3222,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8896,
      "eval_loss": 0.36768344044685364,
      "eval_runtime": 14.1941,
      "eval_samples_per_second": 264.194,
      "eval_steps_per_second": 4.157,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 30.242080688476562,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.1631,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9184,
      "eval_loss": 0.29511919617652893,
      "eval_runtime": 14.2468,
      "eval_samples_per_second": 263.216,
      "eval_steps_per_second": 4.141,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 35.199363708496094,
      "learning_rate": 0.00015,
      "loss": 0.0914,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9189333333333334,
      "eval_loss": 0.3066985607147217,
      "eval_runtime": 14.1655,
      "eval_samples_per_second": 264.728,
      "eval_steps_per_second": 4.165,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 35.92844772338867,
      "learning_rate": 0.0001125,
      "loss": 0.0389,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.2976543605327606,
      "eval_runtime": 14.1314,
      "eval_samples_per_second": 265.366,
      "eval_steps_per_second": 4.175,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 36.62189865112305,
      "learning_rate": 7.5e-05,
      "loss": 0.0275,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9285333333333333,
      "eval_loss": 0.3020574748516083,
      "eval_runtime": 14.389,
      "eval_samples_per_second": 260.616,
      "eval_steps_per_second": 4.1,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 98.71747589111328,
      "learning_rate": 3.75e-05,
      "loss": 0.0149,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.29278460144996643,
      "eval_runtime": 14.149,
      "eval_samples_per_second": 265.036,
      "eval_steps_per_second": 4.17,
      "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
}