train_math_qa_1756735780 / train_results.json
rbelanec's picture
End of training
11aafe9 verified
{
"epoch": 10.0,
"num_input_tokens_seen": 35721744,
"total_flos": 1.6085339872348078e+18,
"train_loss": 0.8150753764391674,
"train_runtime": 12852.7266,
"train_samples_per_second": 20.893,
"train_steps_per_second": 10.447
}