fixstars_xinyu

AMD MI300 GPU (gfx942) is lower than expected

7

# Question MI300 (gfx942) supposed to be faster, but only receive 11.93 tokens per second Here is my inference command `./main -m ./models/llama-2-7b-chat.Q2_K.gguf -p "Building a website can be done...

[Feature Request]: Hybrid Search with BM25

19

### Describe the problem Can you provide keyword search combined with semantic search like other vector store? ### Describe the proposed solution keyword: BM25 ### Alternatives considered _No response_ ###...