Open 0xN1 opened 5 months ago
Make use of ollama embedding, since it is supported now. Perhaps we can work on this. There's also another model (command-r) which support 128k context. Very good for large context especially kisahari.
Proper embedding
-
ollama tweet
Description
Make use of ollama embedding, since it is supported now. Perhaps we can work on this. There's also another model (command-r) which support 128k context. Very good for large context especially kisahari.
Why?
Proper embedding
How to implement
-
Reference
ollama tweet