--- title: Qwen-Image-i2L canonical_url: "https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-i2L" md_url: "https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-i2L.md" repository: DiffSynth-Studio/Qwen-Image-i2L chinese_name: "Qwen-Image-i2L (Image to LoRA)" last_updated: 2025-12-17 license: "Apache License 2.0" pipeline_tag: text-to-image-synthesis tasks: - text-to-image-synthesis parameters: 17.9B tensor_type: - BF16 library_name: - pytorch - safetensors frameworks: - Pytorch downloads: 32523 stars: 114 --- # Qwen-Image-i2L > Qwen-Image-i2L - DiffSynth-Studio 在 ModelScope 开源的模型。Qwen-Image-i2L (Image to LoRA) DiffSynth-Studio/Qwen-Image-i2L 是 ModelScope 魔搭社区上的 17.9B 参数text-to-image-synthesis模型，采用 Apache License 2.0 许可。 - **Repository**: DiffSynth-Studio/Qwen-Image-i2L - **License**: Apache License 2.0 - **Tasks**: text-to-image-synthesis - **Parameters**: 17.9B - **Downloads**: 32523 - **Stars**: 114 - **Last updated**: 2025-12-17 Source: https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-i2L --- # Qwen-Image-i2L (Image to LoRA) ## 模型介绍 i2L (Image to LoRA) 模型是我们以疯狂的思路设计的模型结构。模型的输入为一张图片，输出为这张图片训练出的 LoRA 模型。 更多关于训练策略和实现细节，欢迎查看我们的[技术博客](https://modelscope.cn/learn/3343)。 本次开源的模型包括四个： * Qwen-Image-i2L-Style * 简介：这是我们第一个称得上训练成功的模型，它的细节保持能力非常弱，但这反而让它能够有效地提取出图片中的风格信息，因此这个模型可用于风格迁移。 * 图像编码器：SigLIP2、DINOv3 * 参数量：2.4B * Qwen-Image-i2L-Coarse * 简介：这个模型是 Qwen-Image-i2L-Style 的 scaling up 版本。它产生的 LoRA 已经能够保留图片中的内容信息，但细节并不完善。如果将这个模型用于风格迁移，则需要输入更多图片，否则模型将倾向于生成输入图片中的内容。我们不建议单独使用这个模型。 * 图像编码器：SigLIP2、DINOv3、Qwen-VL（分辨率为 224 x 224） * 参数量：7.9B * Qwen-Image-i2L-Fine * 简介：这个模型是 Qwen-Image-i2L-Coarse 的增量更新版本，必须与 Qwen-Image-i2L-Coarse 同时使用。它将 Qwen-VL 的图像编码分辨率提高到了 1024 x 1024，从而获得更多细节信息。 * 图像编码器：SigLIP2、DINOv3、Qwen-VL（分辨率为 1024 x 1024） * 参数量：7.6B * Qwen-Image-i2L-Bias * 简介：这个模型是静态的补充 LoRA。由于 Qwen-Image-i2L-Coarse 与 Qwen-Image-i2L-Fine 的训练数据与 Qwen-Image 基础模型的训练数据分布不同，因此其生成的 LoRA 生成的图片与 Qwen-Image 的偏好对齐不一致。使用此 LoRA 模型后，生成的图片会更接近 Qwen-Image 的风格。 * 图像编码器：无 * 参数量：30M **这些模型仍有诸多局限，在泛化能力、细节保持能力等方面还有很大改进空间，我们将这些模型开源，以启发更多创新性的研究工作。** ## 效果展示 ### Style Qwen-Image-i2L-Style 模型可用于快速生成风格 LoRA，只需输入几张风格统一的图像。以下是我们生成的结果，随机种子都是 0。 #### 风格1：抽象矢量 输入图像： |![](./assets/style/2/0.jpg)|![](./assets/style/2/1.jpg)|![](./assets/style/2/2.jpg)|![](./assets/style/2/3.jpg)|![](./assets/style/2/4.jpg)|![](./assets/style/2/5.jpg)| |-|-|-|-|-|-| 生成图像： |a cat|a dog|a girl| |-|-|-| |![](./assets/style/2/image_0.jpg)|![](./assets/style/2/image_1.jpg)|![](./assets/style/2/image_2.jpg)| #### 风格2：黑白素描 输入图像： |![](./assets/style/3/0.jpg)|![](./assets/style/3/1.jpg)|![](./assets/style/3/2.jpg)|![](./assets/style/3/3.jpg)| |-|-|-|-| 生成图像： |a cat|a dog|a girl| |-|-|-| |![](./assets/style/3/image_0.jpg)|![](./assets/style/3/image_1.jpg)|![](./assets/style/3/image_2.jpg)| #### 风格3：潦草手绘 输入图像： |![](./assets/style/1/0.jpg)|![](./assets/style/1/1.jpg)|![](./assets/style/1/2.jpg)|![](./assets/style/1/3.jpg)|![](./assets/style/1/4.jpg)| |-|-|-|-|-| 生成图像： |a cat|a dog|a girl| |-|-|-| |![](./assets/style/1/image_0.jpg)|![](./assets/style/1/image_1.jpg)|![](./assets/style/1/image_2.jpg)| #### 风格4：蓝底扁平 输入图像： |![](./assets/style/4/0.jpg)|![](./assets/style/4/1.jpg)|![](./assets/style/4/2.jpg)|![](./assets/style/4/3.jpg)| |-|-|-|-| 生成图像： |a cat|a dog|a girl| |-|-|-| |![](./assets/style/4/image_0.jpg)|![](./assets/style/4/image_1.jpg)|![](./assets/style/4/image_2.jpg)| ### Coarse + Fine + Bias Qwen-Image-i2L-Coarse、Qwen-Image-i2L-Fine、Qwen-Image-i2L-Bias 三者组合后可以生成保留图像内容和细节信息的 LoRA 权重，这组权重作为 LoRA 训练的初始化权重能够加速收敛速度。 #### LoRA 数据集：小狗背包 训练数据： |![](assets/lora/1/0.jpg)|![](assets/lora/1/1.jpg)|![](assets/lora/1/2.jpg)|![](assets/lora/1/3.jpg)|![](assets/lora/1/4.jpg)| |-|-|-|-|-| 训练中的样图生成： ||训练步数：100|训练步数：200|训练步数：300|训练步数：400|训练步数：500| |-|-|-|-|-|-| |随机初始化|![](assets/lora/1/image_original_100.jpg)|![](assets/lora/1/image_original_200.jpg)|![](assets/lora/1/image_original_300.jpg)|![](assets/lora/1/image_original_400.jpg)|![](assets/lora/1/image_original_500.jpg)| |Image to LoRA 初始化|![](assets/lora/1/image_i2l_100.jpg)|![](assets/lora/1/image_i2l_200.jpg)|![](assets/lora/1/image_i2l_300.jpg)|![](assets/lora/1/image_i2l_400.jpg)|![](assets/lora/1/image_i2l_500.jpg)| #### LoRA 数据集：小熊玩偶 训练数据： |![](assets/lora/2/0.jpg)|![](assets/lora/2/1.jpg)|![](assets/lora/2/2.jpg)|![](assets/lora/2/3.jpg)|![](assets/lora/2/4.jpg)| |-|-|-|-|-| 训练中的样图生成： ||训练步数：100|训练步数：200|训练步数：300|训练步数：400|训练步数：500| |-|-|-|-|-|-| |随机初始化|![](assets/lora/2/image_original_100.jpg)|![](assets/lora/2/image_original_200.jpg)|![](assets/lora/2/image_original_300.jpg)|![](assets/lora/2/image_original_400.jpg)|![](assets/lora/2/image_original_500.jpg)| |Image to LoRA 初始化|![](assets/lora/2/image_i2l_100.jpg)|![](assets/lora/2/image_i2l_200.jpg)|![](assets/lora/2/image_i2l_300.jpg)|![](assets/lora/2/image_i2l_400.jpg)|![](assets/lora/2/image_i2l_500.jpg)| #### LoRA 数据集：蓝莓小碗 训练数据： |![](assets/lora/3/0.jpg)|![](assets/lora/3/1.jpg)|![](assets/lora/3/2.jpg)|![](assets/lora/3/3.jpg)|![](assets/lora/3/4.jpg)|![](assets/lora/3/5.jpg)| |-|-|-|-|-|-| 训练中的样图生成： ||训练步数：100|训练步数：200|训练步数：300|训练步数：400|训练步数：500| |-|-|-|-|-|-| |随机初始化|![](assets/lora/3/image_original_100.jpg)|![](assets/lora/3/image_original_200.jpg)|![](assets/lora/3/image_original_300.jpg)|![](assets/lora/3/image_original_400.jpg)|![](assets/lora/3/image_original_500.jpg)| |Image to LoRA 初始化|![](assets/lora/3/image_i2l_100.jpg)|![](assets/lora/3/image_i2l_200.jpg)|![](assets/lora/3/image_i2l_300.jpg)|![](assets/lora/3/image_i2l_400.jpg)|![](assets/lora/3/image_i2l_500.jpg)| ## 推理代码 安装 [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)： ```shell git clone https://github.com/modelscope/DiffSynth-Studio.git cd DiffSynth-Studio pip install -e . ``` ### Qwen-Image-i2L-Style ```python from diffsynth.pipelines.qwen_image import ( QwenImagePipeline, ModelConfig, QwenImageUnit_Image2LoRAEncode, QwenImageUnit_Image2LoRADecode ) from modelscope import snapshot_download from safetensors.torch import save_file import torch from PIL import Image vram_config_disk_offload = { "offload_dtype": "disk", "offload_device": "disk", "onload_dtype": "disk", "onload_device": "disk", "preparing_dtype": torch.bfloat16, "preparing_device": "cuda", "computation_dtype": torch.bfloat16, "computation_device": "cuda", } # Load models pipe = QwenImagePipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="DiffSynth-Studio/General-Image-Encoders", origin_file_pattern="SigLIP2-G384/model.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/General-Image-Encoders", origin_file_pattern="DINOv3-7B/model.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/Qwen-Image-i2L", origin_file_pattern="Qwen-Image-i2L-Style.safetensors", **vram_config_disk_offload), ], processor_config=ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="processor/"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) # Load images snapshot_download( model_id="DiffSynth-Studio/Qwen-Image-i2L", allow_file_pattern="assets/style/1/*", local_dir="data/examples" ) images = [ Image.open("data/examples/assets/style/1/0.jpg"), Image.open("data/examples/assets/style/1/1.jpg"), Image.open("data/examples/assets/style/1/2.jpg"), Image.open("data/examples/assets/style/1/3.jpg"), Image.open("data/examples/assets/style/1/4.jpg"), ] # Model inference with torch.no_grad(): embs = QwenImageUnit_Image2LoRAEncode().process(pipe, image2lora_images=images) lora = QwenImageUnit_Image2LoRADecode().process(pipe, **embs)["lora"] save_file(lora, "model_style.safetensors") ``` ### Qwen-Image-i2L-Coarse、Qwen-Image-i2L-Fine、Qwen-Image-i2L-Bias ```python from diffsynth.pipelines.qwen_image import ( QwenImagePipeline, ModelConfig, QwenImageUnit_Image2LoRAEncode, QwenImageUnit_Image2LoRADecode ) from diffsynth.utils.lora import merge_lora from diffsynth import load_state_dict from modelscope import snapshot_download from safetensors.torch import save_file import torch from PIL import Image vram_config_disk_offload = { "offload_dtype": "disk", "offload_device": "disk", "onload_dtype": "disk", "onload_device": "disk", "preparing_dtype": torch.bfloat16, "preparing_device": "cuda", "computation_dtype": torch.bfloat16, "computation_device": "cuda", } # Load models pipe = QwenImagePipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="text_encoder/model*.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/General-Image-Encoders", origin_file_pattern="SigLIP2-G384/model.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/General-Image-Encoders", origin_file_pattern="DINOv3-7B/model.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/Qwen-Image-i2L", origin_file_pattern="Qwen-Image-i2L-Coarse.safetensors", **vram_config_disk_offload), ModelConfig(model_id="DiffSynth-Studio/Qwen-Image-i2L", origin_file_pattern="Qwen-Image-i2L-Fine.safetensors", **vram_config_disk_offload), ], processor_config=ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="processor/"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) # Load images snapshot_download( model_id="DiffSynth-Studio/Qwen-Image-i2L", allow_file_pattern="assets/lora/3/*", local_dir="data/examples" ) images = [ Image.open("data/examples/assets/lora/3/0.jpg"), Image.open("data/examples/assets/lora/3/1.jpg"), Image.open("data/examples/assets/lora/3/2.jpg"), Image.open("data/examples/assets/lora/3/3.jpg"), Image.open("data/examples/assets/lora/3/4.jpg"), Image.open("data/examples/assets/lora/3/5.jpg"), ] # Model inference with torch.no_grad(): embs = QwenImageUnit_Image2LoRAEncode().process(pipe, image2lora_images=images) lora = QwenImageUnit_Image2LoRADecode().process(pipe, **embs)["lora"] lora_bias = ModelConfig(model_id="DiffSynth-Studio/Qwen-Image-i2L", origin_file_pattern="Qwen-Image-i2L-Bias.safetensors") lora_bias.download_if_necessary() lora_bias = load_state_dict(lora_bias.path, torch_dtype=torch.bfloat16, device="cuda") lora = merge_lora([lora, lora_bias]) save_file(lora, "model_coarse_fine_bias.safetensors") ``` ### 使用生成的 LoRA 生成图像 ```python from diffsynth.pipelines.qwen_image import QwenImagePipeline, ModelConfig import torch vram_config = { "offload_dtype": "disk", "offload_device": "disk", "onload_dtype": torch.bfloat16, "onload_device": "cpu", "preparing_dtype": torch.bfloat16, "preparing_device": "cuda", "computation_dtype": torch.bfloat16, "computation_device": "cuda", } pipe = QwenImagePipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors", **vram_config), ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="text_encoder/model*.safetensors", **vram_config), ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="tokenizer/"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) pipe.load_lora(pipe.dit, "model_style.safetensors") image = pipe("a cat", seed=0, height=1024, width=1024, num_inference_steps=50) image.save("image.jpg") ```