Liam van der Viven

Is there something significantly wrong in just using IndexIVF with a single list? With nlist=1=nprobe we perform an exhaustive search. Does this hinder parallelisation or anything else? ```py def create_index(dimension:...

Thanks for your help. That does work ;) I've now chose to use an IVF with nlist=1. They are memory-mappable ad well and keeps the remaining code the same. Any...

We're seeing the issue as well, forcing us to use fewer middlewares. Please fix.

Update?

this PR broke pydantic v1 support

``` worker-1 | File "/app/botbrains/core/llms.py", line 4, in worker-1 | import litellm worker-1 | File "/usr/local/lib/python3.10/site-packages/ddtrace/internal/module.py", line 220, in _exec_module worker-1 | self.loader.exec_module(module) worker-1 | File "/usr/local/lib/python3.10/site-packages/litellm/__init__.py", line 17, in...

Interesting, it does produce the generation trace, but does not correctly link between them:

@marcklingen The proxy is not an option for us, thus we cannot use `langfuse.openai`. You suggest we do this: ``` from langfuse.decorators import observe, langfuse_context import litellm import os #...