smolm-autoreg-bpe-seed_888 / train_results.json
kanishka's picture
End of training
92997d0
raw
history blame
197 Bytes
{
"epoch": 10.0,
"train_loss": 2.8881424867390284,
"train_runtime": 618.3653,
"train_samples": 46845,
"train_samples_per_second": 757.562,
"train_steps_per_second": 11.838
}