prithivMLmods/Qwen3-VL-4B-Instruct-Unredacted-MAX-FP8
Image-Text-to-Text
•
4B
•
Updated
•
43
•
1
Unredacted MAX - VL [FP8]
[FP8_DYNAMIC] multi-stage trained continual "abliteration" approach for the qwen3-vl series models