Finetuning with Multiple jsons

[ZhangGongjie]

Thanks for the awesome work!

May I know whether this repo can support training with multiple json files?

I would be really nice if it could support something like https://github.com/LLaVA-VL/LLaVA-NeXT/blob/main/scripts/train/single_image.yaml

[2U1]

Well it dosen't support the featrue for that, but you could just merge the json file for doing it.
If there is some reason that you can't merge the json files, I'll give it a try.

[ZhangGongjie]

When doing large-scale experiments, it would be very inconvenient to merge everything into a single json for each experiment. So I think such a feature would be necessary to support large-scale training.

[2U1]

Okay I'll give it a try. Thanks for the comment.