4 ms·

Why have the OpenAI dependency when there's local embeddings models that would be both faster and more accurate?

• yawnxyz 3y ago
  Which ones?
  • minimaxir 3y ago
    all-MiniLM-L6-v2 from SentenceTransformers is the most popular one as it balances speed and quality well: https://www.sbert.net/docs/pretrained_models.html https://www.sbert.net/docs/pretrained_models.html