━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🎯 Qwen-Image 2.1 Glamour Realism LoRA v1Natural-looking skin and photographic colour for glamour portraits on Qwen-Image 2.1.━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━A small style LoRA (rank 16, about 80 MB) trained locally on 45 AI-generated portraits of adult women. It aims to reduce the plastic-skin and washed-out look in glamour-style portraits while keeping the outfit, pose and framing you prompt for. Use it as a gentle polish: recommended strength is 0.6.✨ Features• Aims for more natural skin and richer, more photographic colour in glamour portraits• Gentle by design: at 0.5 to 0.7 it keeps outfit, pose and framing close to your prompt• Small file (about 80 MB), no trigger word• Tested on a 16 GB GPU📦 Required Models• Base model: Qwen-Image 2.1 (this LoRA is for Qwen-Image 2.1 only)• LoRA file: qwen21_glamour_realism_v1.safetensors (about 80 MB)• Not compatible with the original Qwen-Image (20B): LoRAs are not interchangeable between the two• Use the same Qwen-Image 2.1 files you already use for your workflow🧩 Required Custom Nodes• None specific. Any standard LoRA loader node that works with Qwen-Image 2.1 is fine🚀 How to Use• Step 1: load Qwen-Image 2.1 as usual• Step 2: add a LoRA loader after the model loader and pick qwen21_glamour_realism_v1.safetensors• Step 3: set strength to 0.6• Step 4: prompt in plain sentences describing the woman, outfit, setting and light (like the training captions). No trigger word needed⚙️ Settings &amp; Parameters• Strength: 0.6 recommended, 0.5 to 0.7 usable• At 0.8 and above, outfit and pose may start to drift from your prompt• The strength and prompt of each example image are in its generation data💡 Tips• If skin looks too polished or an outfit changes, lower the strength to 0.5• It keeps a soft, polished glamour finish. It is not a raw or gritty skin-texture LoRA• Not tested stacked with other style LoRAs🔗 More from me• My Qwen-Image 2.1 Glamour workflows (SFW and After Dark) are on my profile📝 Notes &amp; AI Disclosure• All example images are AI-generated with this LoRA on Qwen-Image 2.1• Training images were AI-generated too. All subjects are adult women and none are real people━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⭐ Found this useful?• Like if it saved you time• Comment your results, I read every one• Follow for new LoRAs and ComfyUI workflows, all tested on 16 GB VRAM━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⚖️ Model Attribution &amp; Licensing• Built with Qwen• Base model: Qwen-Image 2.1 by the Qwen team. Please follow its license when you use it• This LoRA is an adapter I trained myself and contains no Qwen model weights

Qwen-Image 2.1 Glamour Realism LoRA (16 GB tested)

v1.0: first release. Recommended strength 0.6 (0.5 to 0.7 usable). No trigger word.

Files with SHA256: 4592b5317f710addd8d959897648eb7ac4f2898c988fd02ab9c563b43fe59863