Sebastian Schoennenbeck

> I've read both this and the closed issue I haven't found any mention of the relationship between `kedro run` and `poetry run`. You might assume that the user always...

@simon-mo No worries about the delay, I very much appreciate the fact that you maintain this OSS-project in your free time. Thanks for the feedback, I'll have a look into...

@simon-mo New implementation: - `OpenAIServing`'s `__init__` now accepts an optional `await_post_init`, which can be any awaitable that will be awaited at the end of the `_post_init`-method. - Some corresponding changes...

@simon-mo Would it be possible to get some feedback on this PR. The code changes are pretty minimal and seeing as there was another PR/ bug report last week it...

@simon-mo Thanks a lot. Merge requests are resolved and remaining issues with tests are fixed.

Wouldn't it make more sense to check which layers the LoRA-adapter actually applies to? Most LoRA-frameworks don't even touch the embedding layer by default.

I added a test which is basically just the example code I put in the issue. Feel free to suggest a different method of testing this; I am not sure...

@mjpieters Could we get this merged?

Here is some sample code that demonstrates the issue: ```python import asyncio import time from aiolimiter import AsyncLimiter limiter = AsyncLimiter(3, 3) async def small_task(): await limiter.acquire(1) print(f"Small task completed...

Additional observation: If we are waiting for `1 / self._rate_per_sec * (amount - self.max_rate + self._level)` I don't think there is any real need for the `_waiters` since there cannot...