4 ms·

The key feature here is the cost: compared to the same model size from the previous iteration (ada), it's 1/10th the cost, and I'm curious how they achieved tha

• gamegoblin 4y ago
  Their existing embedding offerings were ridiculously expensive and underpowered compared to free, open source models that you could run locally on your laptop. E.g. https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2 https://huggingface.co/sentence-transformers/all-MiniLM-L6-v... outperformed their best model, can run locally for free, and uses a smaller vector space. Strictly dominates. For a full write-up on how bad it was, see https://medium.com/@nils_reimers/openai-gpt-3-text-embeddings-really-a-new-state-of-the-art-in-dense-text-embeddings-6571fe3ec9d9 https://medium.com/@nils_reimers/openai-gpt-3-text-embedding... OpenAI is still lagging behind, but not as much, with this new offering.
  • minimaxir 4y ago
    There's a few caveats with SentenceTransformers that impact my ideas, so I am willing to give OpenAI a try.
  • yenniejun111 4y ago
    to be fair, the medium article you linked is from Jan 2022 -- before OpenAI released the new Ada 002 embeddings.