[Feature Request] Need to support Any2Text model

[XuyaoWang]

Required prerequisites

• [X] 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.

Motivation

As fields such as V2T (Video to Text) and I2T (Image to Text) mature, we need a any 2 text model to SFT and XPO to align

Solution

See our PR

Alternatives

No response

Additional context

No response