{
  "best_global_step": 6454,
  "best_metric": 0.9517362117767334,
  "best_model_checkpoint": "siglip2-finetune-full/checkpoint-6454",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 6454,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1549426712116517,
      "grad_norm": 55.33502197265625,
      "learning_rate": 2.7891942535915053e-06,
      "loss": 1.9844,
      "step": 500
    },
    {
      "epoch": 0.3098853424233034,
      "grad_norm": 37.46580505371094,
      "learning_rate": 2.5549656464709556e-06,
      "loss": 1.4259,
      "step": 1000
    },
    {
      "epoch": 0.4648280136349551,
      "grad_norm": 41.121490478515625,
      "learning_rate": 2.320737039350406e-06,
      "loss": 1.3044,
      "step": 1500
    },
    {
      "epoch": 0.6197706848466068,
      "grad_norm": 37.83391189575195,
      "learning_rate": 2.0865084322298563e-06,
      "loss": 1.2348,
      "step": 2000
    },
    {
      "epoch": 0.7747133560582584,
      "grad_norm": 41.89641571044922,
      "learning_rate": 1.8522798251093069e-06,
      "loss": 1.1889,
      "step": 2500
    },
    {
      "epoch": 0.9296560272699101,
      "grad_norm": 38.50979995727539,
      "learning_rate": 1.6180512179887572e-06,
      "loss": 1.1496,
      "step": 3000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6981598062953995,
      "eval_loss": 1.0367238521575928,
      "eval_model_preparation_time": 0.0038,
      "eval_runtime": 1328.8007,
      "eval_samples_per_second": 77.702,
      "eval_steps_per_second": 9.713,
      "step": 3227
    },
    {
      "epoch": 1.0845986984815619,
      "grad_norm": 35.213375091552734,
      "learning_rate": 1.3838226108682073e-06,
      "loss": 1.088,
      "step": 3500
    },
    {
      "epoch": 1.2395413696932136,
      "grad_norm": 41.03507995605469,
      "learning_rate": 1.1495940037476577e-06,
      "loss": 1.0959,
      "step": 4000
    },
    {
      "epoch": 1.394484040904865,
      "grad_norm": 39.10642623901367,
      "learning_rate": 9.15365396627108e-07,
      "loss": 1.0571,
      "step": 4500
    },
    {
      "epoch": 1.5494267121165168,
      "grad_norm": 40.350914001464844,
      "learning_rate": 6.811367895065585e-07,
      "loss": 1.0547,
      "step": 5000
    },
    {
      "epoch": 1.7043693833281686,
      "grad_norm": 44.81325912475586,
      "learning_rate": 4.4690818238600874e-07,
      "loss": 1.0296,
      "step": 5500
    },
    {
      "epoch": 1.8593120545398203,
      "grad_norm": 38.73249435424805,
      "learning_rate": 2.126795752654591e-07,
      "loss": 1.0197,
      "step": 6000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7182566585956417,
      "eval_loss": 0.9517362117767334,
      "eval_model_preparation_time": 0.0038,
      "eval_runtime": 1326.8357,
      "eval_samples_per_second": 77.817,
      "eval_steps_per_second": 9.728,
      "step": 6454
    }
  ],
  "logging_steps": 500,
  "max_steps": 6454,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 1.7301390177770496e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}