This is a custom loader i created to use lora's with overly quantified models on LTX2.3 like my FP4ME models.
This has presets to fix issue with cloudy or distorted outputs when using lora's on my LTX2.3 FP4ME model (lora's were over powering certain quantified blocks)
just place comfy_bathroom.py in cumfyui/custom_nodes/
Description
Comfy Bathroom V1
FAQ
Comments (3)
i noticed some gguf outputs are having that weird spot issue as well, this lora loader should fix that if some one wants to try it on a gguf.
JellaiMar 29, 2026
Does this have a github too?
@Jellai haven't put it on github yet, but do have a hf repo for it.
https://huggingface.co/MrReclusive/LTX-2.3-FP4