zephyr-7b-dpo-qlora / train_results.json
RedaAlami's picture
Model save
efa96d7 verified
raw
history blame
232 Bytes
{
"epoch": 0.9924812030075187,
"total_flos": 0.0,
"train_loss": 0.6927223205566406,
"train_runtime": 34619.6666,
"train_samples": 4242,
"train_samples_per_second": 0.123,
"train_steps_per_second": 0.001
}