3 ms·
Is there any specific reason for using LLAMA embedding models instead of the new models OpenAI just released? Are they better?
by cadence- 3y ago
Is there any specific reason for using LLAMA embedding models instead of the new models OpenAI just released? Are they better?
- boredemployee 3y agoNot an expert but I'd say prolly llama is preferred for privacy + open source reasons.
- cadence- 3y agoNot sure how much privacy is needed for a menu in a restaurant. Plus, it's all later sent to GPT which breaks privacy anyway.
- mvdtnz 3y agoThere's no reason for any of this bullshit. Like cryptocurrency in 2021 this is all a solution in search of a problem.
- rwhyan 3y agoWith this particular solution, I agree. But LLMs and RAG in general couldn't be further from crypto in terms of utility.