Oriel — AI & technology worth noticing, every day.

chatrlabs/Qwen-Image-2.1-Uncensored-Omni-Quantized

AI model · https://huggingface.co/chatrlabs/Qwen-Image-2.1-Uncensored-Omni-Quantized

First seen by Oriel on .

AI summary

ChatR Labs provides quantized transformer-only builds of a rank-128 Omni adapter merged into Qwen/Qwen-Image-2.1 for text-to-image generation and image editing.

  • Available formats include GGUF, FP8, INT8 ConvRot, NVFP4, NF4 and BF16, with differing ComfyUI, Diffusers, fidelity and loader requirements.
  • Recommended settings are 35 steps, true CFG 5.0, er_sde with the simple scheduler, KV caching, and a single-space negative prompt.
  • At step 90,000, adult prompt adherence reached 89.7 versus 85.1 for the base, while SFW performance remained comparable and anatomy did not improve.
  • On an RTX 5090, BF16 used approximately 14 GB, Q5_K_M 8 GB and NVFP4 4 GB, with generation taking roughly 28–30 seconds.
  • The repository excludes the text encoder, processor and VAE, requires base components from Qwen/Qwen-Image-2.1, and is limited to non-commercial research and evaluation.

#qwen-image-2.1 #image generation #quantization #lora

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Likes

Total likes reported by the source.

  • Hugging Face Models: 402, latest observation ; +221 over observed period since

Downloads (30d)

Rolling 30-day downloads. A change in this value is not new downloads during the observed period.

  • Hugging Face Models: 142,901, latest observation ; +74,805 over observed period since

Aliases

SourceExternal IDURL
Hugging Face Modelschatrlabs/Qwen-Image-2.1-Uncensored-Omni-Quantizedhttps://huggingface.co/chatrlabs/Qwen-Image-2.1-Uncensored-Omni-Quantized Read discussion