Add Portkey LLMOps integration #7877
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
dosubot
bot
added
Ɑ: embeddings
langchain/embeddings/openai.py
Outdated
| @@ -112,6 +113,7 @@ async def async_embed_with_retry(embeddings: OpenAIEmbeddings, **kwargs: Any) -> | |||
|
|
|||
| @_async_retry_decorator(embeddings) | |||
| async def _async_embed_with_retry(**kwargs: Any) -> Any: | |||
| kwargs.update(embeddings._invocation_params) # Add invocation params | |||
There was a problem hiding this comment.
invocation params should already be passed in as the kwargs, were you seeing issues here?
There was a problem hiding this comment.
Yes. For example, passing headers here: response = OpenAIEmbeddings(headers={"x-portkey-api-key": "API_KEY"}) isn't working.
This is unique only to OpenAIEmbeddings on Python. For other models + on JS, it is working.
There was a problem hiding this comment.
hm not able to recreate. if i mess with headers seems to propagate to top level commands (e.g. embed_query)
There was a problem hiding this comment.
Ok, dropping that change for now - will verify from my end later
Revert openai embeddings changes
|
@baskaryan fixed formartting erros. It should pass all checks now. |
|
looks amazing, thanks @vrushankportkey! |
baskaryan
added
the
lgtm
Integrating Portkey, which adds production features like caching, tracing, tagging, retries, etc. to langchain apps.
Also fixed a bug with OpenAIEmbeddings where headers weren't passing.
cc @baskaryan