Merge pull request #824 from mi804/low_res_fix

support qwen-image-edit lowres fix
2026-03-18 22:08:13 +00:00 · 2025-08-20 10:24:11 +08:00
parent 3a9621f6da 838b8109b1
commit 9d0683df25
8 changed files with 133 additions and 6 deletions
--- a/README.md
+++ b/README.md
@@ -97,6 +97,7 @@ image.save("image.jpg")
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Canny.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Depth.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|
+|[DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix)|[code](./examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py)|-|-|-|-|

 </details>

@@ -369,8 +370,10 @@ https://github.com/Artiprocher/DiffSynth-Studio/assets/35051019/59fb2f7b-8de0-44


 ## Update History
+- **August 20, 2025** We open-sourced [DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix), which improves the editing performance of Qwen-Image-Edit on low-resolution image inputs. Please refer to [our example code](./examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py).
+
 - **August 19, 2025** 🔥 Qwen-Image-Edit is now open source. Welcome the new member to the image editing model family!
- 
+
 - **August 18, 2025** We trained and open-sourced the Inpaint ControlNet model for Qwen-Image, [DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint), which adopts a lightweight architectural design. Please refer to [our sample code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py).

 - **August 15, 2025** We open-sourced the [Qwen-Image-Self-Generated-Dataset](https://www.modelscope.cn/datasets/DiffSynth-Studio/Qwen-Image-Self-Generated-Dataset). This is an image dataset generated using the Qwen-Image model, with a total of 160,000 `1024 x 1024` images. It includes the general, English text rendering, and Chinese text rendering subsets. We provide caption, entity and control images annotations for each image. Developers can use this dataset to train models such as ControlNet and EliGen for the Qwen-Image model. We aim to promote technological development through open-source contributions!
--- a/README_zh.md
+++ b/README_zh.md
@@ -99,6 +99,7 @@ image.save("image.jpg")
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Canny.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Depth.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint)|[code](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_training/full/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./examples/qwen_image/model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./examples/qwen_image/model_training/lora/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./examples/qwen_image/model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|
+|[DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix)|[code](./examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py)|[code](./examples/qwen_image/model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py)|-|-|-|-|

 </details>

@@ -385,6 +386,8 @@ https://github.com/Artiprocher/DiffSynth-Studio/assets/35051019/59fb2f7b-8de0-44


 ## 更新历史
+- **2025年8月20日** 我们开源了 [DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix) 模型，提升了 Qwen-Image-Edit 对低分辨率图像输入的编辑效果。请参考[我们的示例代码](./examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py)
+
 - **2025年8月19日** 🔥 Qwen-Image-Edit 开源，欢迎图像编辑模型新成员！

 - **2025年8月18日** 我们训练并开源了 Qwen-Image 的图像重绘 ControlNet 模型 [DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint)，模型结构采用了轻量化的设计，请参考[我们的示例代码](./examples/qwen_image/model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py)。
--- a/diffsynth/models/qwen_image_dit.py
+++ b/diffsynth/models/qwen_image_dit.py
@@ -166,6 +166,66 @@ class QwenEmbedRope(nn.Module):
        return vid_freqs, txt_freqs


+    def forward_sampling(self, video_fhw, txt_seq_lens, device):
+        self._expand_pos_freqs_if_needed(video_fhw, txt_seq_lens)
+        if self.pos_freqs.device != device:
+            self.pos_freqs = self.pos_freqs.to(device)
+            self.neg_freqs = self.neg_freqs.to(device)
+
+        vid_freqs = []
+        max_vid_index = 0
+        for idx, fhw in enumerate(video_fhw):
+            frame, height, width = fhw
+            rope_key = f"{idx}_{height}_{width}"
+            if idx > 0 and f"{0}_{height}_{width}" not in self.rope_cache:
+                frame_0, height_0, width_0 = video_fhw[0]
+
+                rope_key_0 = f"0_{height_0}_{width_0}"
+                spatial_freqs_0 = self.rope_cache[rope_key_0].reshape(frame_0, height_0, width_0, -1)
+                h_indices = torch.linspace(0, height_0 - 1, height).long()
+                w_indices = torch.linspace(0, width_0 - 1, width).long()
+                h_grid, w_grid = torch.meshgrid(h_indices, w_indices, indexing='ij')
+                sampled_rope = spatial_freqs_0[:, h_grid, w_grid, :]
+
+                freqs_pos = self.pos_freqs.split([x // 2 for x in self.axes_dim], dim=1)
+                freqs_frame = freqs_pos[0][idx : idx + frame].view(frame, 1, 1, -1).expand(frame, height, width, -1)
+                sampled_rope[:, :, :, :freqs_frame.shape[-1]] = freqs_frame
+
+                seq_lens = frame * height * width
+                self.rope_cache[rope_key] = sampled_rope.reshape(seq_lens, -1).clone()
+            if rope_key not in self.rope_cache:
+                seq_lens = frame * height * width
+                freqs_pos = self.pos_freqs.split([x // 2 for x in self.axes_dim], dim=1)
+                freqs_neg = self.neg_freqs.split([x // 2 for x in self.axes_dim], dim=1)
+                freqs_frame = freqs_pos[0][idx : idx + frame].view(frame, 1, 1, -1).expand(frame, height, width, -1)
+                if self.scale_rope:
+                    freqs_height = torch.cat(
+                        [freqs_neg[1][-(height - height // 2) :], freqs_pos[1][: height // 2]], dim=0
+                    )
+                    freqs_height = freqs_height.view(1, height, 1, -1).expand(frame, height, width, -1)
+                    freqs_width = torch.cat([freqs_neg[2][-(width - width // 2) :], freqs_pos[2][: width // 2]], dim=0)
+                    freqs_width = freqs_width.view(1, 1, width, -1).expand(frame, height, width, -1)
+
+                else:
+                    freqs_height = freqs_pos[1][:height].view(1, height, 1, -1).expand(frame, height, width, -1)
+                    freqs_width = freqs_pos[2][:width].view(1, 1, width, -1).expand(frame, height, width, -1)
+
+                freqs = torch.cat([freqs_frame, freqs_height, freqs_width], dim=-1).reshape(seq_lens, -1)
+                self.rope_cache[rope_key] = freqs.clone()
+            vid_freqs.append(self.rope_cache[rope_key].contiguous())
+
+            if self.scale_rope:
+                max_vid_index = max(height // 2, width // 2, max_vid_index)
+            else:
+                max_vid_index = max(height, width, max_vid_index)
+
+        max_len = max(txt_seq_lens)
+        txt_freqs = self.pos_freqs[max_vid_index : max_vid_index + max_len, ...]
+        vid_freqs = torch.cat(vid_freqs, dim=0)
+
+        return vid_freqs, txt_freqs
+
+
 class QwenFeedForward(nn.Module):
    def __init__(
        self,
--- a/diffsynth/pipelines/qwen_image.py
+++ b/diffsynth/pipelines/qwen_image.py
@@ -68,10 +68,10 @@ class QwenImagePipeline(BasePipeline):
            QwenImageUnit_NoiseInitializer(),
            QwenImageUnit_InputImageEmbedder(),
            QwenImageUnit_Inpaint(),
+            QwenImageUnit_EditImageEmbedder(),
            QwenImageUnit_PromptEmbedder(),
            QwenImageUnit_EntityControl(),
            QwenImageUnit_BlockwiseControlNet(),
-            QwenImageUnit_EditImageEmbedder(),
        ]
        self.model_fn = model_fn_qwen_image
        
@@ -280,6 +280,7 @@ class QwenImagePipeline(BasePipeline):
        eligen_enable_on_negative: bool = False,
        # Edit Image
        edit_image: Image.Image = None,
+        edit_rope_interpolation: bool = False,
        # FP8
        enable_fp8_attention: bool = False,
        # Tile
@@ -310,7 +311,7 @@ class QwenImagePipeline(BasePipeline):
            "blockwise_controlnet_inputs": blockwise_controlnet_inputs,
            "tiled": tiled, "tile_size": tile_size, "tile_stride": tile_stride,
            "eligen_entity_prompts": eligen_entity_prompts, "eligen_entity_masks": eligen_entity_masks, "eligen_enable_on_negative": eligen_enable_on_negative,
-            "edit_image": edit_image,
+            "edit_image": edit_image, "edit_rope_interpolation": edit_rope_interpolation,
        }
        for unit in self.units:
            inputs_shared, inputs_posi, inputs_nega = self.unit_runner(unit, self, inputs_shared, inputs_posi, inputs_nega)
@@ -583,11 +584,11 @@ class QwenImageUnit_BlockwiseControlNet(PipelineUnit):
 class QwenImageUnit_EditImageEmbedder(PipelineUnit):
    def __init__(self):
        super().__init__(
-            input_params=("edit_image", "height", "width", "tiled", "tile_size", "tile_stride"),
+            input_params=("edit_image", "tiled", "tile_size", "tile_stride"),
            onload_model_names=("vae",)
        )

-    def process(self, pipe: QwenImagePipeline, edit_image, height, width, tiled, tile_size, tile_stride):
+    def process(self, pipe: QwenImagePipeline, edit_image, tiled, tile_size, tile_stride):
        if edit_image is None:
            return {}
        pipe.load_models_to_device(['vae'])
@@ -616,6 +617,7 @@ def model_fn_qwen_image(
    enable_fp8_attention=False,
    use_gradient_checkpointing=False,
    use_gradient_checkpointing_offload=False,
+    edit_rope_interpolation=False,
    **kwargs
 ):
    img_shapes = [(latents.shape[0], latents.shape[2]//2, latents.shape[3]//2)]
@@ -640,7 +642,10 @@ def model_fn_qwen_image(
        )
    else:
        text = dit.txt_in(dit.txt_norm(prompt_emb))
-        image_rotary_emb = dit.pos_embed(img_shapes, txt_seq_lens, device=latents.device)
+        if edit_rope_interpolation:
+            image_rotary_emb = dit.pos_embed.forward_sampling(img_shapes, txt_seq_lens, device=latents.device)
+        else:
+            image_rotary_emb = dit.pos_embed(img_shapes, txt_seq_lens, device=latents.device)
        attention_mask = None
        
    if blockwise_controlnet_conditioning is not None:
--- a/examples/qwen_image/README.md
+++ b/examples/qwen_image/README.md
@@ -50,6 +50,7 @@ image.save("image.jpg")
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Canny.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Depth.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|
+|[DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix)|[code](./model_inference/Qwen-Image-Edit-Lowres-Fix.py)|[code](./model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py)|-|-|-|-|

 ## Model Inference

--- a/examples/qwen_image/README_zh.md
+++ b/examples/qwen_image/README_zh.md
@@ -50,6 +50,7 @@ image.save("image.jpg")
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Canny)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Canny.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Canny.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Canny.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Depth)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Depth.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Depth.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Depth.py)|
 |[DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint](https://modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Blockwise-ControlNet-Inpaint)|[code](./model_inference/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_inference_low_vram/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_training/full/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./model_training/validate_full/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|[code](./model_training/lora/Qwen-Image-Blockwise-ControlNet-Inpaint.sh)|[code](./model_training/validate_lora/Qwen-Image-Blockwise-ControlNet-Inpaint.py)|
+|[DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix](https://www.modelscope.cn/models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix)|[code](./model_inference/Qwen-Image-Edit-Lowres-Fix.py)|[code](./model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py)|-|-|-|-|

 ## 模型推理

--- a/examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py
+++ b/examples/qwen_image/model_inference/Qwen-Image-Edit-Lowres-Fix.py
@@ -0,0 +1,26 @@
+from diffsynth.pipelines.qwen_image import QwenImagePipeline, ModelConfig
+import torch
+from modelscope import snapshot_download
+
+pipe = QwenImagePipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors"),
+        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="text_encoder/model*.safetensors"),
+        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
+    ],
+    tokenizer_config=None,
+    processor_config=ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="processor/"),
+)
+snapshot_download("DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix", local_dir="models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix", allow_file_pattern="model.safetensors")
+pipe.load_lora(pipe.dit, "models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix/model.safetensors")
+
+prompt = "精致肖像，水下少女，蓝裙飘逸，发丝轻扬，光影透澈，气泡环绕，面容恬静，细节精致，梦幻唯美。"
+image = pipe(prompt=prompt, seed=0, num_inference_steps=40, height=1024, width=768)
+image.save("image.jpg")
+
+prompt = "将裙子变成粉色"
+image = image.resize((512, 384))
+image = pipe(prompt, edit_image=image, seed=1, num_inference_steps=40, height=1024, width=768, edit_rope_interpolation=True)
+image.save(f"image2.jpg")
--- a/examples/qwen_image/model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py
+++ b/examples/qwen_image/model_inference_low_vram/Qwen-Image-Edit-Lowres-Fix.py
@@ -0,0 +1,28 @@
+from diffsynth.pipelines.qwen_image import QwenImagePipeline, ModelConfig
+import torch
+from modelscope import snapshot_download
+
+pipe = QwenImagePipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors", offload_device="cpu", offload_dtype=torch.float8_e4m3fn),
+        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="text_encoder/model*.safetensors", offload_device="cpu", offload_dtype=torch.float8_e4m3fn),
+        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors", offload_device="cpu", offload_dtype=torch.float8_e4m3fn),
+    ],
+    tokenizer_config=None,
+    processor_config=ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="processor/"),
+)
+pipe.enable_vram_management()
+
+snapshot_download("DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix", local_dir="models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix", allow_file_pattern="model.safetensors")
+pipe.load_lora(pipe.dit, "models/DiffSynth-Studio/Qwen-Image-Edit-Lowres-Fix/model.safetensors")
+
+prompt = "精致肖像，水下少女，蓝裙飘逸，发丝轻扬，光影透澈，气泡环绕，面容恬静，细节精致，梦幻唯美。"
+image = pipe(prompt=prompt, seed=0, num_inference_steps=40, height=1024, width=768)
+image.save("image.jpg")
+
+prompt = "将裙子变成粉色"
+image = image.resize((512, 384))
+image = pipe(prompt, edit_image=image, seed=1, num_inference_steps=40, height=1024, width=768, edit_rope_interpolation=True)
+image.save(f"image2.jpg")