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

11 lines
325 B
JSON
Raw Permalink Normal View History

{
"ewc_loss": 0.041259765625,
"ewc_loss_parallel": 4.124641418457031e-05,
"ewc_loss_perp": 0.0,
"total_flos": 5.62815163329864e+19,
"train_loss": 0.43322460087045367,
"train_runtime": 46186.2128,
"train_samples": 125770,
"train_samples_per_second": 8.169,
"train_steps_per_second": 0.511
}