FireRed-OCR fine-tuning

[llery2021]

Kalorda now supports fine-tuning for FireRed-OCR model/vllm v0.16.0/ms-swift v4.0.0
https://github.com/vlmOCR/Kalorda

[johnlockejrr]

It does but doesn't support HF datasets so you have to throw files in the UI, that's not very helpful.