Files
Llama-3.1-8B-Instruct_SFT_m…/all_results.json

8 lines
219 B
JSON
Raw Normal View History

{
"total_flos": 5.628159003328302e+19,
"train_loss": 0.3662048727583553,
"train_runtime": 42474.8118,
"train_samples": 125770,
"train_samples_per_second": 8.883,
"train_steps_per_second": 0.555
}