- Issue created by @andrewbelcher
- First commit to issue fork.
- 🇬🇧United Kingdom colinsherry
Hi,
When I use your explorer tool it s also returning general chatgpt seswrch data as weel as my pinecone data.
Can we restrict the search results to pinecone data only.
Thanks
- 🇹🇼Taiwan yelban
Hi,
I would like to kindly propose the possibility of incorporating additional vector database backends into this awesome project.I believe platforms like Weaviate, LlamaIndex, and Chromadb, in addition to our current use of Pinecone, might provide a wider range of features and benefits.
Thanks & Regards.,
- 🇬🇧United Kingdom yautja_cetanu
@yelban - We have looked into Weaviate and ran into a problem and submitted a patch to Weviate.
https://github.com/milvus-io/milvus/issues/25206 - We need this patch to be in a release
- 🇬🇧United Kingdom yautja_cetanu
Created an issue here: https://www.drupal.org/project/search_api_ai/issues/3375565 📌 Implement Weviate as a vector storage backend Active
- 🇬🇧United Kingdom yautja_cetanu
I've been told that bad chunking of cutting things off is fine. But the solution is to start the next chunk slightly before the previous one ended.
- Status changed to Closed: duplicate
about 1 year ago 4:30am 23 November 2023 - 🇬🇧United Kingdom scott_euser
There are:
https://www.drupal.org/project/search_api_ai/issues/3375565 📌 Implement Weviate as a vector storage backend Active now for Weviate
https://www.drupal.org/project/search_api_ai/issues/3361507 📌 Make the embeddings pluggable (so alternatives to OpenAI can be used) Active for OpenAISo suggest we close this in favour of those more specific issues.