Files
Llama-3.1-8B-Instruct_SFT_s…/train_results.json

8 lines
218 B
JSON
Raw Permalink Normal View History

{
"total_flos": 5.939640094684692e+17,
"train_loss": 0.34853814358897506,
"train_runtime": 8974.1904,
"train_samples": 28249,
"train_samples_per_second": 4.457,
"train_steps_per_second": 1.114
}