Inference client to be added to the pipeline
#33
by
JulienGuy
- opened
Hello everyone,
I am using the inference endpoints to connect to the LLM. I have a RAG system and I want to add several components to a pipeline. When I do this, I get an error when adding the Inferenceclient() to the pipeline.
The error is '<class 'huggingface_hub.inference._client.InferenceClient'>' doesn't seem to be a component. Is this class decorated with @component?
I could not find any syntax to add the inference client to the pipeline. Your answer is much appreciated!
Thank you very much!
What do you mean by pipeline? Is this from another library like langchain? Any code you could share would be very useful.