Required prerequisites

• I have read the documentation https://align-anything.readthedocs.io.
• I have searched the Issue Tracker and Discussions that this hasn't already been reported. (+1 or comment there if it has.)
• Consider asking first in a Discussion.

What version of align-anything are you using?

Newest (cutoff 4.27)

System information

System version: 3.11.11 (main, Dec 11 2024, 16:28:39) [GCC 11.2.0] linux
Align Anything version: newest, cutoff date 0427, after commit eea5af6

Problem description

The existing Janus implementation was merged in February 2025. Since then, align-anything has undergone several major refactorings, which have resulted in the following issues with the Janus implementation:

• The image output finetuning contains an outdated version of the chat template, which prevents it from simultaneously supporting both Janus and Janus Pro models
• The image output DPO has a naming inconsistency with DPOTextTrainer, causing a TypeError: DPOTrainer.loss() got an unexpected keyword argument 'batch' error
• The image input finetuning has issues with dataset loading and inconsistent function naming in the backend, causing a data loading error.

Reproducible example code

/

Traceback

Expected behavior

No response

Additional context

I'm currently fixing these bugs. To be specific:

• All the bugs in the image output fine-tuning/DPO are fixed.
• The dataset loading error in the image input fine-tuning has been successfully located. I will fix this error in the next few days.

After all the bugs are fixed and tested, I will open a PR and merge these modifications. All the updates during this process will be reported in this thread.