Does the encode method in sentence_transformers support base64-encoded image input?
#71
by
zhangxiangchn
- opened
When using the sentence_transformers library to load the jina-embeddings-v4 model locally and encode images, the official examples require uploading the image's URL link. Is it possible, like the official API, to support uploading image data in base64 encoding? If so, could you recommend some sample code? Thank you.