Long Context Size Chunk
#27
by CVHvn - opened
With reranker model, I found all model only have 512 or 1024 context size. This is very small when compare with embedding model (like jina v3, openai v3). How we can re-ranking relevant chunks when we split chunk with long context size?