Update app.py

app.py

@@ -172,6 +172,13 @@ tag_model = ram(pretrained='preset/models/ram_swin_large_14m.pth',
 tag_model.eval()
 tag_model.to(device, dtype=weight_dtype)
 
+def preprocess_image(
+    input_image: Image.Image
+    )
+    input_image = input_image.resize((256, 256), Image.Resampling.BILINEAR)
+
+    return input_image
+
 @spaces.GPU()
 def process(
     input_image: Image.Image,
@@ -187,7 +194,7 @@ def process(
     sample_times = 1,
     ) -> List[np.ndarray]:
 
-    input_image = input_image.resize((256, 256), Image.Resampling.BILINEAR)
+    
     
     process_size = 512
     resize_preproc = transforms.Compose([
@@ -251,62 +258,74 @@ def process(
     return input_image, images[0]
 
 
-#
-MARKDOWN = \
-"""
-## SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution
-
-[GitHub](https://github.com/cswry/SeeSR) | [Paper](https://arxiv.org/abs/2311.16518)
-
-If SeeSR is helpful for you, please help star the GitHub Repo. Thanks!
-"""
-
-block = gr.Blocks().queue()
-with block:
-    with gr.Row():
-        gr.Markdown(MARKDOWN)
-    with gr.Row():
-        with gr.Column():
-            input_image = gr.Image(type="pil")
-            run_button = gr.Button("Magnify 4x")
-            with gr.Accordion("Options", visible=False):
-                user_prompt = gr.Textbox(label="User Prompt", value="")
-                positive_prompt = gr.Textbox(label="Positive Prompt", value="clean, high-resolution, 8k, best quality, masterpiece")
-                negative_prompt = gr.Textbox(
-                    label="Negative Prompt",
-                    value="dotted, noise, blur, lowres, oversmooth, longbody, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality"
-                )
-                cfg_scale = gr.Slider(label="Classifier Free Guidance Scale (Set to 1.0 in sd-turbo)", minimum=1, maximum=10, value=7.5, step=0)
-                num_inference_steps = gr.Slider(label="Inference Steps", minimum=2, maximum=100, value=50, step=1)
-                seed = gr.Slider(label="Seed", minimum=-1, maximum=2147483647, step=1, value=231)
-                sample_times = gr.Slider(label="Sample Times", minimum=1, maximum=10, step=1, value=1)
-                latent_tiled_size = gr.Slider(label="Diffusion Tile Size", minimum=128, maximum=480, value=320, step=1)
-                latent_tiled_overlap = gr.Slider(label="Diffusion Tile Overlap", minimum=4, maximum=16, value=4, step=1)
-                scale_factor = gr.Number(label="SR Scale", value=4)
-        with gr.Column():
-            result_gallery = ImageSlider(
-                    interactive=False,
-                    label="Generated Image",
-                )
-            examples = gr.Examples(
-                examples=[
-                    [
-                        "preset/datasets/test_datasets/179.png", 
-                    ],         
-                    [
-                        "preset/datasets/test_datasets/apologise.png", 
-                    ], 
-                ],
-                inputs=[
-                    input_image,
-                ],
-                outputs=[result_gallery],
-                fn=process,
-                cache_examples=True,
+css = """
+    #col-container {
+        margin: 0 auto;
+        max-width: 1024px;
+    }
+    """
+
+with gr.Blocks(css=css) as demo:
+
+    with gr.Column(elem_id="col-container"):
+            
+        with gr.Row():
+            gr.HTML(
+                """
+                <div style="text-align: center;">
+                    <p style="font-size:16px; display: inline; margin: 0;">
+                        <strong>SeeSR</strong> – Towards Semantics-Aware Real-World Image Super-Resolution
+                    </p>
+                    <a href="https://github.com/cswry/SeeSR" style="display: inline-block; vertical-align: middle; margin-left: 0.5em;">
+                        <img src="https://img.shields.io/badge/GitHub-Repo-blue" alt="GitHub Repo">
+                    </a>
+                </div>
+                """
             )
-    inputs = [
-        input_image,
-    ]
-    run_button.click(fn=process, inputs=inputs, outputs=[result_gallery])
+        with gr.Row():
+            with gr.Column():
+                input_image = gr.Image(type="pil")
+                run_button = gr.Button("Magnify 4x")
+                preprocessed_image = gr.Image(label="preprocess image", type="pil")
+                with gr.Accordion("Options", visible=False):
+                    user_prompt = gr.Textbox(label="User Prompt", value="")
+                    positive_prompt = gr.Textbox(label="Positive Prompt", value="clean, high-resolution, 8k, best quality, masterpiece")
+                    negative_prompt = gr.Textbox(
+                        label="Negative Prompt",
+                        value="dotted, noise, blur, lowres, oversmooth, longbody, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality"
+                    )
+                    cfg_scale = gr.Slider(label="Classifier Free Guidance Scale (Set to 1.0 in sd-turbo)", minimum=1, maximum=10, value=7.5, step=0)
+                    num_inference_steps = gr.Slider(label="Inference Steps", minimum=2, maximum=100, value=50, step=1)
+                    seed = gr.Slider(label="Seed", minimum=-1, maximum=2147483647, step=1, value=231)
+                    sample_times = gr.Slider(label="Sample Times", minimum=1, maximum=10, step=1, value=1)
+                    latent_tiled_size = gr.Slider(label="Diffusion Tile Size", minimum=128, maximum=480, value=320, step=1)
+                    latent_tiled_overlap = gr.Slider(label="Diffusion Tile Overlap", minimum=4, maximum=16, value=4, step=1)
+                    scale_factor = gr.Number(label="SR Scale", value=4)
+            with gr.Column():
+                result_gallery = ImageSlider(
+                        interactive=False,
+                        label="Generated Image",
+                    )
+                examples = gr.Examples(
+                    examples=[
+                        [
+                            "preset/datasets/test_datasets/179.png", 
+                        ],         
+                        [
+                            "preset/datasets/test_datasets/apologise.png", 
+                        ], 
+                    ],
+                    inputs=[
+                        input_image,
+                    ],
+                    outputs=[result_gallery],
+                    fn=process,
+                    cache_examples=True,
+                )
+        inputs = [
+            input_image,
+        ]
+        run_button.click(fn=process, inputs=preprocessed_image, outputs=[result_gallery])
+        input_image.upload(fn=preprocess_image,inputs=input_image, outputs=[preprocessed_image])
 
-block.launch(share=True)
+demo.launch(share=True)