simple_online_epoch_2_dpo_iter_5 / train_results.json
YYYYYYibo's picture
Model save
4264ef2 verified
raw
history blame contribute delete
194 Bytes
{
"epoch": 0.99,
"train_loss": 0.687320882978003,
"train_runtime": 21824.845,
"train_samples": 20000,
"train_samples_per_second": 0.916,
"train_steps_per_second": 0.007
}