Update app.py

app.py

@@ -1,42 +1,37 @@
+import os
 import gradio as gr
 import numpy as np
-import random
-import torch
 import spaces
-import os
+import torch
+import random
 from PIL import Image
 from typing import Iterable
 
-from diffusers import FlowMatchEulerDiscreteScheduler
-from optimization import optimize_pipeline_
-from qwenimage.pipeline_qwenimage_edit_plus import QwenImageEditPlusPipeline
-from qwenimage.transformer_qwenimage import QwenImageTransformer2DModel
-from qwenimage.qwen_fa3_processor import QwenDoubleStreamAttnProcessorFA3
-
+# --- Gradio Theme ---
 from gradio.themes import Soft
 from gradio.themes.utils import colors, fonts, sizes
 
-colors.steel_blue = colors.Color(
-    name="steel_blue",
-    c50="#EBF3F8",
-    c100="#D3E5F0",
-    c200="#A8CCE1",
-    c300="#7DB3D2",
-    c400="#529AC3",
-    c500="#4682B4",
-    c600="#3E72A0",
-    c700="#36638C",
-    c800="#2E5378",
-    c900="#264364",
-    c950="#1E3450",
+colors.blue_ish = colors.Color(
+    name="blue_ish",
+    c50="#F0F5FF",
+    c100="#E0EBFF",
+    c200="#C2D7FF",
+    c300="#A3C2FF",
+    c400="#85AFFF",
+    c500="#4A8DFF",
+    c600="#3374E6",
+    c700="#1A5CCC",
+    c800="#0043B3",
+    c900="#002B80",
+    c950="#00144D",
 )
 
-class SteelBlueTheme(Soft):
+class QwenTheme(Soft):
     def __init__(
         self,
         *,
         primary_hue: colors.Color | str = colors.gray,
-        secondary_hue: colors.Color | str = colors.steel_blue,
+        secondary_hue: colors.Color | str = colors.blue_ish,
         neutral_hue: colors.Color | str = colors.slate,
         text_size: sizes.Size | str = sizes.text_lg,
         font: fonts.Font | str | Iterable[fonts.Font | str] = (
@@ -55,46 +50,35 @@ class SteelBlueTheme(Soft):
             font_mono=font_mono,
         )
         super().set(
-            background_fill_primary="*primary_50",
-            background_fill_primary_dark="*primary_900",
-            body_background_fill="linear-gradient(135deg, *primary_200, *primary_100)",
+            body_background_fill="linear-gradient(135deg, *primary_100, *primary_50)",
             body_background_fill_dark="linear-gradient(135deg, *primary_900, *primary_800)",
-            button_primary_text_color="white",
-            button_primary_text_color_hover="white",
             button_primary_background_fill="linear-gradient(90deg, *secondary_500, *secondary_600)",
             button_primary_background_fill_hover="linear-gradient(90deg, *secondary_600, *secondary_700)",
-            button_primary_background_fill_dark="linear-gradient(90deg, *secondary_600, *secondary_800)",
-            button_primary_background_fill_hover_dark="linear-gradient(90deg, *secondary_500, *secondary_500)",
-            button_secondary_text_color="black",
-            button_secondary_text_color_hover="white",
-            button_secondary_background_fill="linear-gradient(90deg, *primary_300, *primary_300)",
-            button_secondary_background_fill_hover="linear-gradient(90deg, *primary_400, *primary_400)",
-            button_secondary_background_fill_dark="linear-gradient(90deg, *primary_500, *primary_600)",
-            button_secondary_background_fill_hover_dark="linear-gradient(90deg, *primary_500, *primary_500)",
+            button_primary_text_color="white",
             slider_color="*secondary_500",
             slider_color_dark="*secondary_600",
             block_title_text_weight="600",
-            block_border_width="3px",
+            block_border_width="2px",
             block_shadow="*shadow_drop_lg",
-            button_primary_shadow="*shadow_drop_lg",
-            button_large_padding="11px",
-            color_accent_soft="*primary_100",
-            block_label_background_fill="*primary_200",
         )
 
-steel_blue_theme = SteelBlueTheme()
+qwen_theme = QwenTheme()
+
+
+# --- Model Loading ---
+from diffusers import FlowMatchEulerDiscreteScheduler
+from optimization import optimize_pipeline_
+from qwenimage.pipeline_qwenimage_edit_plus import QwenImageEditPlusPipeline
+from qwenimage.transformer_qwenimage import QwenImageTransformer2DModel
+from qwenimage.qwen_fa3_processor import QwenDoubleStreamAttnProcessorFA3
 
-# --- Constants and Setup ---
-MAX_SEED = np.iinfo(np.int32).max
 dtype = torch.bfloat16
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-# --- Model Loading ---
-# Load the base pipeline and the optimized transformer
 pipe = QwenImageEditPlusPipeline.from_pretrained(
-    "Qwen/Qwen-Image-Edit-2509", 
+    "Qwen/Qwen-Image-Edit-2509",
     transformer=QwenImageTransformer2DModel.from_pretrained(
-        "linoyts/Qwen-Image-Edit-Rapid-AIO", 
+        "linoyts/Qwen-Image-Edit-Rapid-AIO",
         subfolder='transformer',
         torch_dtype=dtype,
         device_map='cuda'
@@ -102,218 +86,176 @@ pipe = QwenImageEditPlusPipeline.from_pretrained(
     torch_dtype=dtype
 ).to(device)
 
-# Load all LoRA adapters with unique names
-pipe.load_lora_weights(
-    "dx8152/Qwen-Image-Edit-2509-Light_restoration", 
-    weight_name="移除光影.safetensors", 
-    adapter_name="light_restoration"
-)
-pipe.load_lora_weights(
-    "dx8152/Qwen-Edit-2509-Multiple-angles",
-    weight_name="镜头转换.safetensors",
-    adapter_name="multiple_angles"
-)
-pipe.load_lora_weights(
-    "autoweeb/Qwen-Image-Edit-2509-Photo-to-Anime",
-    weight_name="Qwen-Image-Edit-2509-Photo-to-Anime_000001000.safetensors",
-    adapter_name="photo_to_anime"
-)
-pipe.load_lora_weights(
-    "dx8152/Qwen-Image-Edit-2509-Relight",
-    weight_name="Qwen-Edit-Relight.safetensors",
-    adapter_name="relight"
-)
+# Load all LoRA adapters
+pipe.load_lora_weights("autoweeb/Qwen-Image-Edit-2509-Photo-to-Anime",
+                       weight_name="Qwen-Image-Edit-2509-Photo-to-Anime_000001000.safetensors",
+                       adapter_name="anime")
+pipe.load_lora_weights("dx8152/Qwen-Edit-2509-Multiple-angles",
+                       weight_name="镜头转换.safetensors",
+                       adapter_name="multiple-angles")
+pipe.load_lora_weights("dx8152/Qwen-Image-Edit-2509-Light_restoration",
+                       weight_name="移除光影.safetensors",
+                       adapter_name="light-restoration")
+pipe.load_lora_weights("dx8152/Qwen-Image-Edit-2509-Relight",
+                       weight_name="Qwen-Edit-Relight.safetensors",
+                       adapter_name="relight")
 
-# Apply optimizations
-pipe.transformer.__class__ = QwenImageTransformer2DModel
 pipe.transformer.set_attn_processor(QwenDoubleStreamAttnProcessorFA3())
 optimize_pipeline_(pipe, image=[Image.new("RGB", (1024, 1024)), Image.new("RGB", (1024, 1024))], prompt="prompt")
 
+MAX_SEED = np.iinfo(np.int32).max
+
+# --- Helper Functions ---
+def update_dimensions_on_upload(image):
+    if image is None:
+        return 1024, 1024
+    
+    original_width, original_height = image.size
+    
+    if original_width > original_height:
+        new_width = 1024
+        aspect_ratio = original_height / original_width
+        new_height = int(new_width * aspect_ratio)
+    else:
+        new_height = 1024
+        aspect_ratio = original_width / original_height
+        new_width = int(new_height * aspect_ratio)
+        
+    # Ensure dimensions are multiples of 8
+    new_width = (new_width // 8) * 8
+    new_height = (new_height // 8) * 8
+    
+    return new_width, new_height
 
-# --- Inference Logic ---
+# --- Main Inference Function ---
 @spaces.GPU
 def infer(
-    image,
+    input_image,
     prompt,
     lora_adapter,
     seed,
     randomize_seed,
-    true_guidance_scale,
-    num_inference_steps,
-    height,
+    guidance_scale,
+    steps,
     width,
+    height,
     progress=gr.Progress(track_tqdm=True)
 ):
-    if image is None:
-        raise gr.Error("Please upload an image to get started.")
-    
-    # Set the active LoRA adapter based on user selection
-    if lora_adapter == "Shadow/Light Restoration":
-        pipe.set_adapters(["light_restoration"], adapter_weights=[1.0])
-    elif lora_adapter == "Multiple Angles":
-        pipe.set_adapters(["multiple_angles"], adapter_weights=[1.0])
-    elif lora_adapter == "Photo to Anime":
-        pipe.set_adapters(["photo_to_anime"], adapter_weights=[1.0])
-    elif lora_adapter == "Advanced Relighting":
-        pipe.set_adapters(["relight"], adapter_weights=[1.0])
+    if input_image is None:
+        raise gr.Error("Please upload an image to edit.")
 
+    # Dynamically set the adapter
+    if lora_adapter == "Photo-to-Anime":
+        pipe.set_adapters(["anime"], adapter_weights=[1.0])
+    elif lora_adapter == "Multiple-Angles":
+        pipe.set_adapters(["multiple-angles"], adapter_weights=[1.0])
+    elif lora_adapter == "Light-Restoration":
+        pipe.set_adapters(["light-restoration"], adapter_weights=[1.0])
+    elif lora_adapter == "Relight":
+        pipe.set_adapters(["relight"], adapter_weights=[1.0])
+        
     if randomize_seed:
         seed = random.randint(0, MAX_SEED)
+    
     generator = torch.Generator(device=device).manual_seed(seed)
-
+    
     result = pipe(
-        image=image.convert("RGB"),
+        image=input_image.convert("RGB"),
         prompt=prompt,
         height=height,
         width=width,
-        num_inference_steps=num_inference_steps,
+        num_inference_steps=steps,
         generator=generator,
-        true_cfg_scale=true_guidance_scale,
+        true_cfg_scale=guidance_scale,
         num_images_per_prompt=1,
     ).images[0]
 
-    return result, seed
+    return result, seed, gr.Button(visible=True)
 
-# --- UI Helper Functions ---
-def update_dimensions_on_upload(image):
-    """Adjusts the height and width sliders to match the uploaded image's aspect ratio."""
-    if image is None:
-        return 1024, 1024
-    
-    original_width, original_height = image.size
-    
-    if original_width > original_height:
-        new_width = 1024
-        new_height = int(new_width * (original_height / original_width))
-    else:
-        new_height = 1024
-        new_width = int(new_height * (original_width / original_height))
-        
-    # Ensure dimensions are multiples of 8 for model compatibility
-    new_width = (new_width // 8) * 8
-    new_height = (new_height // 8) * 8
-    
-    return new_width, new_height
-
-def update_prompt_on_adapter_change(adapter_name):
-    """Provides a suggested prompt when a new adapter is selected."""
-    prompts = {
-        "Shadow/Light Restoration": "Remove shadows and relight the image using soft lighting.",
-        "Multiple Angles": "A photo of the scene from a top-down view.",
-        "Photo to Anime": "Transform into anime, masterpiece, best quality.",
-        "Advanced Relighting": "Relight the image using soft, diffused lighting that simulates sunlight filtering through curtains."
-    }
-    return prompts.get(adapter_name, "")
+# Wrapper for examples
+@spaces.GPU
+def infer_example(input_image, prompt, lora_adapter):
+    input_pil = Image.open(input_image).convert("RGB")
+    width, height = update_dimensions_on_upload(input_pil)
+    result, seed, _ = infer(input_pil, prompt, lora_adapter, 0, True, 1.0, 4, width, height)
+    return result, seed
 
-# --- Gradio UI ---
-css = '''
-#col-container { 
-    max-width: 960px; 
-    margin: 0 auto; 
-    font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif;
-}
-.dark .progress-text { color: white !important }
-#examples { max-width: 960px; margin: 0 auto; }
-.gradio-container {
-    background: linear-gradient(135deg, #f5f7fa 0%, #c3cfe2 100%);
-}
-.gr-button-primary {
-    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
-    border: none !important;
-    border-radius: 12px !important;
-    padding: 12px 24px !important;
-    font-weight: 600 !important;
+# --- UI Layout ---
+css="""
+#col-container {
+    margin: 0 auto;
+    max-width: 960px;
 }
-.gr-box {
-    border-radius: 16px !important;
-    box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1) !important;
-}
-'''
+#main-title h1 {font-size: 2.1em !important;}
+"""
 
-with gr.Blocks(theme=steel_blue_theme, css=css) as demo:
+with gr.Blocks(css=css, theme=qwen_theme) as demo:
     with gr.Column(elem_id="col-container"):
-        gr.Markdown("# Qwen Image Edit - Fast LoRA")
+        gr.Markdown("# **Qwen-Image-Edit-2509-LoRAs-Fast**", elem_id="main-title")
+        gr.Markdown("Perform diverse image edits using specialized LoRA adapters for the Qwen-Image-Edit model.")
+        
         with gr.Row():
-            with gr.Column(scale=1):
-                image = gr.Image(label="Upload Image", type="pil", height=450)
+            with gr.Column():
+                input_image = gr.Image(label="Upload Image", type="pil", height=320)
                 
                 lora_adapter = gr.Dropdown(
-                    label="Choose an Editing Tool",
-                    choices=[
-                        "Shadow/Light Restoration", 
-                        "Multiple Angles", 
-                        "Photo to Anime", 
-                        "Advanced Relighting"
-                    ],
-                    value="Shadow/Light Restoration"
+                    label="Choose Editing Style",
+                    choices=["Photo-to-Anime", "Multiple-Angles", "Light-Restoration", "Relight"],
+                    value="Photo-to-Anime"
                 )
-
-                prompt = gr.Textbox(
-                    label="Prompt", 
-                    value="Remove shadows and relight the image using soft lighting.",
-                    lines=2
+                
+                prompt = gr.Text(
+                    label="Edit Prompt",
+                    show_label=True,
+                    placeholder="e.g., transform into anime",
                 )
                 
-                run_btn = gr.Button("Generate", variant="primary", size="lg")
-
+                run_button = gr.Button("Run", variant="primary")
+                
                 with gr.Accordion("⚙️ Advanced Settings", open=False):
                     seed = gr.Slider(label="Seed", minimum=0, maximum=MAX_SEED, step=1, value=0)
                     randomize_seed = gr.Checkbox(label="Randomize Seed", value=True)
-                    true_guidance_scale = gr.Slider(label="Guidance Scale", minimum=1.0, maximum=10.0, step=0.1, value=1.0)
-                    num_inference_steps = gr.Slider(label="Inference Steps", minimum=1, maximum=40, step=1, value=4)
-                    height = gr.Slider(label="Height", minimum=256, maximum=2048, step=8, value=1024)
-                    width = gr.Slider(label="Width", minimum=256, maximum=2048, step=8, value=1024)
+                    guidance_scale = gr.Slider(label="Guidance Scale", minimum=1.0, maximum=10.0, step=0.1, value=1.0)
+                    steps = gr.Slider(label="Inference Steps", minimum=1, maximum=40, step=1, value=4)
+                    # Hidden sliders to hold image dimensions
+                    height = gr.Slider(label="Height", minimum=256, maximum=2048, step=8, value=1024, visible=False)
+                    width = gr.Slider(label="Width", minimum=256, maximum=2048, step=8, value=1024, visible=False)
 
-            with gr.Column(scale=1):
-                result = gr.Image(label="Output Image", interactive=False, height=500, format="png")
+            with gr.Column():
+                output_image = gr.Image(label="Output Image", show_label=True, interactive=False, format="png", height=480)
+                reuse_button = gr.Button("Reuse this image", visible=False)
 
         gr.Examples(
-            elem_id="examples",
             examples=[
-                [
-                    "examples/example1.png",
-                    "A photo of the scene from a low angle shot.",
-                    "Multiple Angles",
-                ],
-                [
-                    "examples/example2.png",
-                    "Remove shadows and relight the image using soft lighting.",
-                    "Shadow/Light Restoration",
-                ],
-                [
-                    "examples/example3.png",
-                    "Transform into anime, masterpiece, best quality, girl with cherry blossoms.",
-                    "Photo to Anime",
-                ],
-                [
-                    "examples/example4.png",
-                    "Relight the image using soft, diffused lighting that simulates sunlight filtering through curtains.",
-                    "Advanced Relighting",
-                ],
+                ["examples/anime_example.jpg", "transform into anime", "Photo-to-Anime"],
+                ["examples/car_example.jpg", "view from the side", "Multiple-Angles"],
+                ["examples/shadow_example.jpg", "Remove shadows and relight the image using soft lighting.", "Light-Restoration"],
+                ["examples/relight_example.jpg", "Relight the image using soft, diffused lighting that simulates sunlight filtering through curtains.", "Relight"],
             ],
-            inputs=[image, prompt, lora_adapter],
-            outputs=[result, seed],
-            fn=infer,
-            cache_examples=False
+            inputs=[input_image, prompt, lora_adapter],
+            outputs=[output_image, seed],
+            fn=infer_example,
+            cache_examples="lazy",
+            label="Examples"
         )
             
     # --- Event Handlers ---
-    run_btn.click(
-        fn=infer, 
-        inputs=[image, prompt, lora_adapter, seed, randomize_seed, true_guidance_scale, num_inference_steps, height, width], 
-        outputs=[result, seed]
+    run_button.click(
+        fn=infer,
+        inputs=[input_image, prompt, lora_adapter, seed, randomize_seed, guidance_scale, steps, width, height],
+        outputs=[output_image, seed, reuse_button]
     )
     
-    image.upload(
-        fn=update_dimensions_on_upload,
-        inputs=[image],
-        outputs=[width, height]
+    reuse_button.click(
+        fn=lambda img: img,
+        inputs=[output_image],
+        outputs=[input_image]
     )
     
-    lora_adapter.change(
-        fn=update_prompt_on_adapter_change,
-        inputs=[lora_adapter],
-        outputs=[prompt]
+    input_image.upload(
+        fn=update_dimensions_on_upload,
+        inputs=[input_image],
+        outputs=[width, height]
     )
 
-demo.launch(mcp_server=True, ssr_mode=False, show_error=True)
+demo.launch(debug=True)