a1_math_numina_math / all_results.json
sedrickkeh's picture
End of training
efdfa3a verified
raw
history blame
225 Bytes
{
"epoch": 4.982278481012658,
"total_flos": 2.4972863287905485e+18,
"train_loss": 0.29206294888039913,
"train_runtime": 132505.3394,
"train_samples_per_second": 1.192,
"train_steps_per_second": 0.009
}