trainer_state.json

{
  "best_metric": 0.8634920634920635,
  "best_model_checkpoint": "dinov2-base-finetuned-har/checkpoint-42",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 42,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 44.80964660644531,
      "learning_rate": 4.324324324324325e-05,
      "loss": 2.2183,
      "step": 10
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 20.761051177978516,
      "learning_rate": 2.9729729729729733e-05,
      "loss": 1.1574,
      "step": 20
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 28.136566162109375,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.8864,
      "step": 30
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 15.380955696105957,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.8084,
      "step": 40
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8634920634920635,
      "eval_loss": 0.44639232754707336,
      "eval_runtime": 28.1795,
      "eval_samples_per_second": 67.07,
      "eval_steps_per_second": 1.065,
      "step": 42
    },
    {
      "epoch": 1.0,
      "step": 42,
      "total_flos": 1.0941516662243328e+18,
      "train_loss": 1.2471325397491455,
      "train_runtime": 397.8778,
      "train_samples_per_second": 26.918,
      "train_steps_per_second": 0.106
    }
  ],
  "logging_steps": 10,
  "max_steps": 42,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1.0941516662243328e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}