{
  "best_metric": 2.9360151290893555,
  "best_model_checkpoint": "traffic_sign_detection/checkpoint-3255",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 3255,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.7680491551459293,
      "grad_norm": 1.5009950399398804,
      "learning_rate": 1.7191887675507019e-06,
      "loss": 3.6834,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3687851275399914,
      "eval_loss": 3.585068702697754,
      "eval_runtime": 23.4653,
      "eval_samples_per_second": 98.571,
      "eval_steps_per_second": 12.359,
      "step": 651
    },
    {
      "epoch": 1.5360983102918588,
      "grad_norm": 1.6353492736816406,
      "learning_rate": 1.4071762870514821e-06,
      "loss": 3.3809,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6502377864245569,
      "eval_loss": 3.3017969131469727,
      "eval_runtime": 23.2776,
      "eval_samples_per_second": 99.366,
      "eval_steps_per_second": 12.458,
      "step": 1302
    },
    {
      "epoch": 2.3041474654377883,
      "grad_norm": 1.6631314754486084,
      "learning_rate": 1.0951638065522622e-06,
      "loss": 3.1341,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7423259835711198,
      "eval_loss": 3.092193365097046,
      "eval_runtime": 23.2504,
      "eval_samples_per_second": 99.482,
      "eval_steps_per_second": 12.473,
      "step": 1953
    },
    {
      "epoch": 3.0721966205837172,
      "grad_norm": 1.7302379608154297,
      "learning_rate": 7.831513260530422e-07,
      "loss": 2.9474,
      "step": 2000
    },
    {
      "epoch": 3.8402457757296466,
      "grad_norm": 1.7401925325393677,
      "learning_rate": 4.7113884555382213e-07,
      "loss": 2.825,
      "step": 2500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7851275399913532,
      "eval_loss": 2.9739248752593994,
      "eval_runtime": 23.2681,
      "eval_samples_per_second": 99.407,
      "eval_steps_per_second": 12.463,
      "step": 2604
    },
    {
      "epoch": 4.6082949308755765,
      "grad_norm": 1.737073540687561,
      "learning_rate": 1.5912636505460217e-07,
      "loss": 2.7523,
      "step": 3000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7980977086035452,
      "eval_loss": 2.9360151290893555,
      "eval_runtime": 23.295,
      "eval_samples_per_second": 99.292,
      "eval_steps_per_second": 12.449,
      "step": 3255
    }
  ],
  "logging_steps": 500,
  "max_steps": 3255,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 8.068862243415214e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}