3 agent sessions
Runs keep going long after eval loss stops improving. Add early stopping.
claudeLleo · 1 commit · last 11 hours ago · f03e64a7
Eval numbers look too good. Check whether eval prompts leak into train and fix the split.
codexMmaya · 1 commit · last 3 days ago · 3b7ca10c
The 7B LoRA fine-tune keeps blowing up around step 400. Logs are in logs/run-17.log. Fix it.
claudeSsam · 2 commits · last 6 days ago · 285791a1