c59f028150
When the studio process is killed (SIGTERM/SIGKILL), atexit handlers may not run in the subprocess orchestrator, leaving llama-server processes orphaned and holding GPU memory. This caused OOM errors when trying to load a new model after a studio restart. On init, LlamaCppBackend now runs pgrep to find and SIGKILL any stale llama-server processes before starting fresh. |
||
|---|---|---|
| .. | ||
| data_recipe | ||
| export | ||
| inference | ||
| training | ||
| __init__.py | ||