An abliterated text encoder for Krea 2 or any other model that uses Qwen 3 VL Instruct 4b.
Note that this is neither necessary nor sufficient to make NSFW generations, but it may be better at expanding NSFW prompts, and I've preferred results from abliterated models before. (After experimenting, it seems like it may make your results slightly more nsfw if you're otherwise not doing any other jailbreaks, but not reliably.)
Not guaranteed to be better than the normal model. I'm just uploading it because I did the conversion and thought I'd share in case anyone else needs it.
HF link to original model in diffusers format: https://huggingface.co/svjack/Qwen3-VL-4B-Instruct-heretic-7refusal
Description
FAQ
Comments (4)
Please tell what about a speed? Is it faster than default TE, how much on what hardware?
I haven't timed this one in particular, but my experience with int8 convrot so far is that it's ~50% faster on a 3090 and 4090, and I'm told there's a speed up across most nvidia cards. You need to make sure your cuda and pytorch are recent (cuda 13.0 is working for me) or it'll actually be slower.
what node loads this?
Any chances for GGUF Q4 / bnb-NF4 ? Unfortunately Neo doesn't recognize versions presented on HF.