Files
llama-3-8b-base-new-dpo-hh-…/all_results.json

9 lines
231 B
JSON
Raw Normal View History

{
"epoch": 0.999244142101285,
"total_flos": 0.0,
"train_loss": 1.143015539862565,
"train_runtime": 1644.2601,
"train_samples": 42336,
"train_samples_per_second": 25.748,
"train_steps_per_second": 0.402
}