zephyr-7b-hh-dpo / train_results.json
NicholasCorrado's picture
Model save
cea8f12 verified
{
"epoch": 0.9996020692399522,
"total_flos": 0.0,
"train_loss": 0.5721431234080321,
"train_runtime": 15529.5131,
"train_samples": 160800,
"train_samples_per_second": 10.354,
"train_steps_per_second": 0.04
}