d1_math_multiple_languages / all_results.json
sedrickkeh's picture
End of training
bc0b7be verified
{
"epoch": 5.0,
"total_flos": 4.90203692596514e+18,
"train_loss": 0.34299286624439335,
"train_runtime": 58019.4787,
"train_samples_per_second": 2.723,
"train_steps_per_second": 0.021
}