Description
I’m trying to multiprocess my RAG with around 4 million queries. The queries are over one cluster but with different metadata filters. I have used ThreadPoolExecutor but there seems to be no improvement. I thought that each query can be paralleled on one cpu. Any idea how I should speed up the whole process? Thank you!
Server Setup Information
- Weaviate Server Version:
- Deployment Method:
- Multi Node? Number of Running Nodes:
- Client Language and Version:
- Multitenancy?: