reward_model_train_final / eval_results.json
shirwu's picture
Training in progress, step 1
44f8035 verified
{
"epoch": 0.0044444444444444444,
"eval_accuracy": 0.5934065934065934,
"eval_loss": 0.6923828125,
"eval_runtime": 4.6767,
"eval_samples_per_second": 42.765,
"eval_steps_per_second": 10.691
}