train code

[dadawdadD]

@zytx121
Thank you for your excellent work! When will the code for training and merging the model be made public?

[zytx121]

We acknowledge the support received. This work utilizes the LLaVA 1.5 framework for training. Further details regarding the text-mask implementation are available at: https://github.com/mc-lan/Text4Seg

The comprehensive source code will be open-sourced after the paper's acceptance.