Some Other MLLMs Finetuning

[superdabuniu]

Hi, Can you write a tutorial about PaliGemma or LLaVA-NeXT or OneVision Finetuning?

[2U1]

I made the codes where there were no fine-tuning code or need to use some other frameworks.
Using the official repo for LLaVA-Next would be just okay I think.
For the Paligemma, I'll give it a try.

[superdabuniu]

hello, Can you improve the logger information in each module during training? Meanwhile the developer can open the log file to debug and analysis the training loss in differnet checkpoints.