Content pfp
Content
@
https://opensea.io/collection/dev-21
0 reply
0 recast
2 reactions

megabyte pfp
megabyte
@megabyte
I hope gemini is better than openai at least for embeddings!
1 reply
0 recast
0 reaction

shoni.eth pfp
shoni.eth
@alexpaden
the embedding model or data being embedded?
1 reply
0 recast
0 reaction

megabyte pfp
megabyte
@megabyte
Embedding model
1 reply
0 recast
0 reaction

shoni.eth pfp
shoni.eth
@alexpaden
what do you feel is lacking in your current one? OpenAI small I assume? I think it’s an issue with what’s being embedded probably but am still working on a solution
2 replies
0 recast
0 reaction

Bhavya Gor pfp
Bhavya Gor
@bhavyagor
go on can you explain a bit more here?
1 reply
0 recast
0 reaction

shoni.eth pfp
shoni.eth
@alexpaden
what are you generating the embedding on? Raw text of a cast/reply?
1 reply
0 recast
0 reaction

Bhavya Gor pfp
Bhavya Gor
@bhavyagor
not precisely - we take the raw text and tell the AI agent to find intent from the cast and summarise that in keywords pair that is then used to generate embeddings on
1 reply
0 recast
0 reaction