train_wsc_456_1760356427 / train_results.json
rbelanec's picture
End of training
aa21961 verified
{
"epoch": 10.0,
"num_input_tokens_seen": 485152,
"total_flos": 2.1846175286820864e+16,
"train_loss": 0.39903659229278565,
"train_runtime": 185.1613,
"train_samples_per_second": 26.895,
"train_steps_per_second": 6.751
}