Find answers from the community

Updated 2 weeks ago

Experience with the vertexai multimodal embedding

Anyone have any experience with the vertexai multimodal embedding here ? Not sure what the maturity of the implementation is since I don't really see a lot of documentation regarding it. If anyone has any thoughts on how to use it for txt2img retrieval that would be great! Thanks.
L
A
5 comments
Even though the google colab is more code, I prefer it, its clearer whats actually going on
Thank you! I'll take a look.
Also, just wondering if it's possible to use ImageDocument to contain both text and image content that will be embedded together. That is, I'm wondering if it's possible to get some sort of cross-modality (mix of text and image) embedding if I pass ImageDocument(text=text, image_url=....). @Logan M
eh, not really. Lots of reasons for this, but the main reason is image embedding models are good at retrieving images, usually not great at retrieving text. So you end up with two embedding models, and thus two vector store collections
Got it. Thanks!
Add a reply
Sign up and join the conversation on Discord