Files
armv8mac_to_x86_qwen25coder…/train_results.json

8 lines
227 B
JSON
Raw Permalink Normal View History

{
"epoch": 0.5000058835296471,
"total_flos": 1.1590600616501576e+19,
"train_loss": 0.024605854352245132,
"train_runtime": 113139.8787,
"train_samples_per_second": 0.751,
"train_steps_per_second": 0.094
}