Perf: test llm before RAPTOR. (#8897)

### What problem does this PR solve?


### Type of change

- [x] Performance Improvement
This commit is contained in:
Kevin Hu
2025-07-17 16:48:50 +08:00
committed by GitHub
parent 606bf20a3f
commit ecdb1701df
4 changed files with 22 additions and 12 deletions

View File

@ -670,3 +670,18 @@ def timeout(
return wrapper
return decorator
async def is_strong_enough(chat_model, embedding_model):
@timeout(30, 2)
async def _is_strong_enough():
nonlocal chat_model, embedding_model
_ = await trio.to_thread.run_sync(lambda: embedding_model.encode(["Are you strong enough!?"]))
res = await trio.to_thread.run_sync(lambda: chat_model.chat("Nothing special.", [{"role":"user", "content": "Are you strong enough!?"}], {}))
if res.find("**ERROR**") >= 0:
raise Exception(res)
# Pressure test for GraphRAG task
async with trio.open_nursery() as nursery:
for _ in range(12):
nursery.start_soon(_is_strong_enough, chat_model, embedding_model)