--- title: Qwen-Image-2.1-LoRAs canonical_url: "https://www.modelscope.cn/models/WarmBloodAban/Qwen-Image-2.1-LoRAs" md_url: "https://www.modelscope.cn/models/WarmBloodAban/Qwen-Image-2.1-LoRAs.md" repository: WarmBloodAban/Qwen-Image-2.1-LoRAs last_updated: 2026-09-28 license: apache-2.0 pipeline_tag: text-to-image-synthesis tasks: - text-to-image-synthesis base_model: - Qwen/Qwen-Image-2.1 base_model_relation: adapter parameters: 243.3M tensor_type: - BF16 library_name: - lora - diffusers - safetensors supports_inference: txt2img downloads: 120 stars: 4 tags: - text-to-image - lora - diffusers - "template:diffusion-lora" --- # Qwen-Image-2.1-LoRAs > Qwen-Image-2.1-LoRAs - WarmBloodAban 在 ModelScope 开源的模型。Qwen-Image-2.1-LoRAs WarmBloodAban/Qwen-Image-2.1-LoRAs 是 ModelScope 魔搭社区上的 243.3M 参数text-to-image-synthesis模型，采用 apache-2.0 许可，基于 Qwen/Qwen-Image-2.1 构建，并支持在线推理（txt2img）。 - **Repository**: WarmBloodAban/Qwen-Image-2.1-LoRAs - **License**: apache-2.0 - **Tasks**: text-to-image-synthesis - **Parameters**: 243.3M - **Base model**: Qwen/Qwen-Image-2.1 - **Online inference**: txt2img - **Tags**: text-to-image, lora, diffusers, template:diffusion-lora - **Downloads**: 120 - **Stars**: 4 - **Last updated**: 2026-09-28 Source: https://www.modelscope.cn/models/WarmBloodAban/Qwen-Image-2.1-LoRAs --- # Qwen-Image-2.1-LoRAs ## Model description --- license: apache-2.0 base_model: Qwen/Qwen-Image-2.1 tags: - qwen - qwen-image - image-editing - lora - anime --- # 🎨 Qwen-Image-2.1-Edit-LoRAs Welcome to the **Qwen-Image-2.1-Edit-LoRAs** repository! This repository is dedicated to LoRA models specifically fine-tuned for the image editing capabilities of `Qwen-Image-2.1`. I will be continuously updating and publishing various functional and stylistic LoRAs here. --- ## 📦 Available Models ### 1. Qwen2.1_Anime_consistency * **Overview**: A LoRA designed to enhance character consistency during anime-style image editing. * **Training Data**: Primarily fine-tuned on character model sheets (4-view references) and a variety of facial expression edits. * **Status**: ⚠️ **Experimental** > *Note: This model is currently in an experimental phase. It was primarily developed to test and benchmark the optimal training parameters for Qwen Image 2.1. As a result, specific editing outcomes and output stability are not strictly guaranteed.* --- ### 2. Qwen2.1_Anything2RealCharacters * **Overview**: A LoRA designed to transform images from any artistic style into realistic human character images. * **Training Data**: Primarily fine-tuned on a large dataset of real human images paired with facial expression control reference groups. * **Recommended Settings**: * **Sampler**: `Euler` * **Scheduler**: `FlowMatchEulerDiscreteScheduler` > *Tip: Using these recommended settings helps achieve softer, more authentic, and high-fidelity realistic character rendering.* * **Preview / Showcase**:
Preview
--- ## ⚙️ Recommended Parameters To ensure optimal performance, please adhere to the official base model settings: - **Sampling Steps**: Follow the official recommended settings for Qwen-Image 2.1. - **CFG Scale**: Follow the official recommended settings for Qwen-Image 2.1. - **LoRA Weight**: It is recommended to start testing between `0.6` and `0.8`, adjusting based on your specific editing needs. --- ## 🤝 Community & Commercial Inquiries Feel free to connect for tutorials, community discussions, workflow sharing, or commercial collaborations: - 📺 **YouTube Channel**: [AIGC-Singularity](https://www.youtube.com/@AIGC-Singularity) - 📺 **Bilibili Channel**: [AIGC-Singularity](https://space.bilibili.com/49766729) - 💬 **QQ Group 2**: `1072010342` *(Please specify your intent when requesting to join)* - 💼 **Business Inquiries (WeChat)**: `aigctyd` - 📧 **Email**: a592991299@gmail.com ## Download model [Download](/WarmBloodAban/Qwen-Image-2.1-LoRAs/tree/main) them in the Files & versions tab.