Evaluation on large-scale dataset

Hello,
Thank you for your amazing work. The demo runs fine for a single video.

I'm curious if there are any provisions for generating inference on a larger dataset of videos, each accompanied by corresponding text questions. Additionally, I'm interested to know if there's an API available for this purpose.

Looking forward to your insights on this matter.

The issue does not identify a file, test, entry point, or existing batch-inference interface to inspect. Clarify the intended dataset format, API behavior, and evaluation criteria before implementation can be scoped.

Written by the indexing model from the issue text.