lua-mistral-1L-tiny / eval_results.json
nilq's picture
End of training
8577212 verified
raw
history blame contribute delete
227 Bytes
{
"epoch": 3.0,
"eval_loss": 2.1613609790802,
"eval_runtime": 368.5926,
"eval_samples": 570721,
"eval_samples_per_second": 1548.379,
"eval_steps_per_second": 193.55,
"perplexity": 8.682946929680876
}