Max Deichmann
Max Deichmann
@Doma1204, @MarkEdmondson1234, @kdcokenny i just released a fix for this. Please re-open if this persists for you: https://github.com/langfuse/langfuse-python/releases/tag/v2.29.1
Hi @rpgeddam thanks for reporting this. We will look into this. Which Langchain + Langfuse versions are you using?
@aunitt, @joshmcqueen, @rpgeddam: I just released a fix for this. Please re-open if this persists for you: https://github.com/langfuse/langfuse-python/releases/tag/v2.29.1
Hi @reza-mohideen thanks for raising this and providing all the details. We will look into this.
@Hokyjack thanks for the issue. Do you have a code example where your async implementation failed?
Hi, thanks for raising this! I got a bunch of questions: - Why did you flush within the worker? Usually, this should only happen at shutdown. - Is it possible...
> @maxdeichmann I am still running into this issue. > > > > > > Which parameter on the object is large in your case? Is it input, output, metadata...
@abhimasand are you self-hosting langfuse or are you using our cloud product?
Ok, are you somehow using nginx in kubernetes? I fear that nginx denied the request (https://docs.nginx.com/nginx-management-suite/acm/how-to/policies/request-body-size-limit/#about-the-policy) We allow bodies of up to 2.5 mb: https://github.com/langfuse/langfuse-python/blob/fbd04266a9de6d936af6f82d268cf0c5c31f226b/langfuse/task_manager.py#L34 Can you try to configure...
@fhrzn are you self-hosting as well?