Cronologia Commit

Autore SHA1 Messaggio Data
  Tobi Lütke 189916d6fb Move GRPO training out of default finetune pipeline 4 mesi fa
  Tobi Lutke 599935754b finetune: remove orphaned files and abandoned experiments 4 mesi fa
  Tobi Lütke 102ff861d3 fix: use Qwen3 recommended sampling params to prevent repetition loops 5 mesi fa
  Tobi Lütke bf1b8fc90a lots of training stuff 5 mesi fa