Batch inference

#25
by gordonyeh - opened

Hi

I am trying to do the batch inference.

Is there a plan to support it on this model (microsoft/Phi-3.5-vision-instruct)

sample format like: [prompt1, prompt2], [[img0, img1], [img_a, img_b, img_c]]

Thanks

@gordonyeh I implemented batching for multiple (prompt, img) pairs. I think it can be extended to (prompt, multiple_imgs) pairs.

https://huggingface.co./microsoft/Phi-3.5-vision-instruct/discussions/5#66c6fa5859ff3e4811b4349b

Sign up or log in to comment