question about "Batch inference code for captioner"

[zZer0o]

Thanks for your work!

Could you explain which part of the inference stage in ShareCaptioner-Video can be modified for batch inference mode? Or maybe I have misunderstood the meaning of "batch inference". The code to be released is actually for batch inference of multiple videos, rather than performing batch inference in a single inference process.

[xiaoachen98]

Thanks for your work!

Could you explain which part of the inference stage in ShareCaptioner-Video can be modified for batch inference mode? Or maybe I have misunderstood the meaning of "batch inference". The code to be released is actually for batch inference of multiple videos, rather than performing batch inference in a single inference process.

It's for annotating a batch of videos at once under the sliding and fast captioning mode.