teacher_math_qwq / all_results.json
sedrickkeh's picture
End of training
7e02779 verified
{
"epoch": 4.99492385786802,
"total_flos": 6.032180188216295e+18,
"train_loss": 0.8038945751461556,
"train_runtime": 33001.4558,
"train_samples_per_second": 4.774,
"train_steps_per_second": 0.037
}