Update app.py

app.py

@@ -266,143 +266,126 @@ examples = [
     ],
 ]
 
-def create_interface():
-    with gr.Blocks(css=CUSTOM_CSS) as demo:
-        gr.Markdown(DESCRIPTION)
-        
-        with gr.Group(elem_classes="container"):
-            with gr.Row():
-                with gr.Column(scale=1):
-                    image = gr.Image(
-                        type="pil",
-                        label="Upload Image",
-                        elem_classes="input-box"
-                    )
-                
-                with gr.Column(scale=2):
-                    with gr.Tabs(elem_classes="tab-nav"):
-                        with gr.Tab(label="✨ Image Captioning"):
-                            caption_button = gr.Button(
-                                "Generate Caption",
-                                elem_classes="button-primary"
-                            )
-                            caption_output = gr.Textbox(
-                                label="Generated Caption",
-                                elem_classes="output-box"
-                            )
-                            
-                        with gr.Tab(label="💭 Visual Q&A"):
-                            chatbot = gr.Chatbot(
-                                elem_classes="chatbot-message"
-                            )
-                            history_orig = gr.State(value=[])
-                            history_qa = gr.State(value=[])
-                            vqa_input = gr.Textbox(
-                                placeholder="Ask me anything about the image...",
-                                elem_classes="input-box"
-                            )
-                            
-                            with gr.Row():
-                                clear_button = gr.Button(
-                                    "Clear Chat",
-                                    elem_classes="button-secondary"
-                                )
-                                submit_button = gr.Button(
-                                    "Send Message",
-                                    elem_classes="button-primary"
-                                )
-
-            with gr.Accordion("🛠️ Advanced Settings", open=False, elem_classes="advanced-settings"):
-                with gr.Row():
-                    with gr.Column():
-                        text_decoding_method = gr.Radio(
-                            choices=["Beam search", "Nucleus sampling"],
-                            value="Nucleus sampling",
-                            label="Decoding Method"
-                        )
-                        temperature = gr.Slider(
-                            minimum=0.5,
-                            maximum=1.0,
-                            value=1.0,
-                            label="Temperature",
-                            info="Used with nucleus sampling",
-                            elem_classes="slider-container"
-                        )
-                        length_penalty = gr.Slider(
-                            minimum=-1.0,
-                            maximum=2.0,
-                            value=1.0,
-                            label="Length Penalty",
-                            info="Set to larger for longer sequence",
-                            elem_classes="slider-container"
-                        )
-                    with gr.Column():
-                        repetition_penalty = gr.Slider(
-                            minimum=1.0,
-                            maximum=5.0,
-                            value=1.5,
-                            label="Repetition Penalty",
-                            info="Larger value prevents repetition",
-                            elem_classes="slider-container"
-                        )
-                        max_length = gr.Slider(
-                            minimum=20,
-                            maximum=512,
-                            value=50,
-                            label="Max Length",
-                            elem_classes="slider-container"
+with gr.Blocks(css=CUSTOM_CSS) as demo:
+    gr.Markdown(DESCRIPTION)
+    
+    with gr.Group(elem_classes="container"):
+        with gr.Row():
+            with gr.Column(scale=1):
+                image = gr.Image(
+                    type="pil",
+                    label="Upload Image",
+                    elem_classes="input-box"
+                )
+            
+            with gr.Column(scale=2):
+                with gr.Tabs(elem_classes="tab-nav"):
+                    with gr.Tab(label="✨ Image Captioning"):
+                        caption_button = gr.Button(
+                            "Generate Caption",
+                            elem_classes="button-primary"
                         )
-                        min_length = gr.Slider(
-                            minimum=1,
-                            maximum=100,
-                            value=1,
-                            label="Min Length",
-                            elem_classes="slider-container"
+                        caption_output = gr.Textbox(
+                            label="Generated Caption",
+                            elem_classes="output-box"
                         )
-                        num_beams = gr.Slider(
-                            minimum=1,
-                            maximum=10,
-                            value=5,
-                            label="Number of Beams",
-                            elem_classes="slider-container"
+                        
+                    with gr.Tab(label="💭 Visual Q&A"):
+                        chatbot = gr.Chatbot(
+                            elem_classes="chatbot-message"
                         )
-                        top_p = gr.Slider(
-                            minimum=0.5,
-                            maximum=1.0,
-                            value=0.9,
-                            label="Top P",
-                            info="Used with nucleus sampling",
-                            elem_classes="slider-container"
+                        history_orig = gr.State(value=[])
+                        history_qa = gr.State(value=[])
+                        vqa_input = gr.Textbox(
+                            placeholder="Ask me anything about the image...",
+                            elem_classes="input-box"
                         )
+                        
+                        with gr.Row():
+                            clear_button = gr.Button(
+                                "Clear Chat",
+                                elem_classes="button-secondary"
+                            )
+                            submit_button = gr.Button(
+                                "Send Message",
+                                elem_classes="button-primary"
+                            )
+
+        with gr.Accordion("🛠️ Advanced Settings", open=False, elem_classes="advanced-settings"):
+            with gr.Row():
+                with gr.Column():
+                    text_decoding_method = gr.Radio(
+                        choices=["Beam search", "Nucleus sampling"],
+                        value="Nucleus sampling",
+                        label="Decoding Method"
+                    )
+                    temperature = gr.Slider(
+                        minimum=0.5,
+                        maximum=1.0,
+                        value=1.0,
+                        label="Temperature",
+                        info="Used with nucleus sampling",
+                        elem_classes="slider-container"
+                    )
+                    length_penalty = gr.Slider(
+                        minimum=-1.0,
+                        maximum=2.0,
+                        value=1.0,
+                        label="Length Penalty",
+                        info="Set to larger for longer sequence",
+                        elem_classes="slider-container"
+                    )
+                with gr.Column():
+                    repetition_penalty = gr.Slider(
+                        minimum=1.0,
+                        maximum=5.0,
+                        value=1.5,
+                        label="Repetition Penalty",
+                        info="Larger value prevents repetition",
+                        elem_classes="slider-container"
+                    )
+                    max_length = gr.Slider(
+                        minimum=20,
+                        maximum=512,
+                        value=50,
+                        label="Max Length",
+                        elem_classes="slider-container"
+                    )
+                    min_length = gr.Slider(
+                        minimum=1,
+                        maximum=100,
+                        value=1,
+                        label="Min Length",
+                        elem_classes="slider-container"
+                    )
+                    num_beams = gr.Slider(
+                        minimum=1,
+                        maximum=10,
+                        value=5,
+                        label="Number of Beams",
+                        elem_classes="slider-container"
+                    )
+                    top_p = gr.Slider(
+                        minimum=0.5,
+                        maximum=1.0,
+                        value=0.9,
+                        label="Top P",
+                        info="Used with nucleus sampling",
+                        elem_classes="slider-container"
+                    )
 
-        with gr.Group(elem_classes="examples-container"):
-            gr.Examples(
-                examples=examples,
-                inputs=[image, vqa_input],
-                label="Try these examples"
-            )
-
-        # Event handlers
-        caption_button.click(
-            fn=generate_caption,
-            inputs=[
-                image,
-                text_decoding_method,
-                temperature,
-                length_penalty,
-                repetition_penalty,
-                max_length,
-                min_length,
-                num_beams,
-                top_p,
-            ],
-            outputs=caption_output,
-            api_name="caption",
+    with gr.Group(elem_classes="examples-container"):
+        gr.Examples(
+            examples=examples,
+            inputs=[image, vqa_input],
+            label="Try these examples"
         )
 
-        chat_inputs = [
+    # Event handlers
+    caption_button.click(
+        fn=generate_caption,
+        inputs=[
             image,
-            vqa_input,
             text_decoding_method,
             temperature,
             length_penalty,
@@ -411,72 +394,70 @@ def create_interface():
             min_length,
             num_beams,
             top_p,
+        ],
+        outputs=caption_output,
+        api_name="caption",
+    )
+
+    chat_inputs = [
+        image,
+        vqa_input,
+        text_decoding_method,
+        temperature,
+        length_penalty,
+        repetition_penalty,
+        max_length,
+        min_length,
+        num_beams,
+        top_p,
+        history_orig,
+        history_qa,
+    ]
+    chat_outputs = [
+        chatbot,
+        history_orig,
+        history_qa,
+    ]
+
+    vqa_input.submit(
+        fn=chat,
+        inputs=chat_inputs,
+        outputs=chat_outputs
+    ).success(
+        fn=lambda: "",
+        outputs=vqa_input,
+        queue=False,
+        api_name=False
+    )
+
+    clear_button.click(
+        fn=lambda: ("", [], [], []),
+        inputs=None,
+        outputs=[
+            vqa_input,
+            chatbot,
             history_orig,
             history_qa,
-        ]
-        chat_outputs = [
+        ],
+        queue=False,
+        api_name="clear"
+    )
+
+    image.change(
+        fn=lambda: ("", [], [], []),
+        inputs=None,
+        outputs=[
+            caption_output,
             chatbot,
             history_orig,
             history_qa,
-        ]
-
-        vqa_input.submit(
-            fn=chat,
-            inputs=chat_inputs,
-            outputs=chat_outputs,
-            api_name="chat",
-        ).success(
-            fn=lambda: "",
-            outputs=vqa_input,
-            queue=False,
-            api_name=False,
-        )
-
-        clear_button.click(
-            fn=lambda: ("", [], [], []),
-            inputs=None,
-            outputs=[
-                vqa_input,
-                chatbot,
-                history_orig,
-                history_qa,
-            ],
-            queue=False,
-            api_name="clear",
-        )
-
-        image.change(
-            fn=lambda: ("", [], [], []),
-            inputs=None,
-            outputs=[
-                caption_output,
-                chatbot,
-                history_orig,
-                history_qa,
-            ],
-            queue=False,
-        )
-
-    return demo
+        ],
+        queue=False
+    )
 
 if __name__ == "__main__":
-    demo = create_interface()
-    demo.queue(max_size=10).launch(),
-        ).success(
-            fn=lambda: "",
-            outputs=vqa_input,
-            queue=False,
-            api_name=False,
-        )
-
-        submit_button.click(
-            fn=chat,
-            inputs=chat_inputs,
-            outputs=chat_outputs,
-            api_name="chat"
-        ).success(
-            fn=lambda: "",
-            outputs=vqa_input,
-            queue=False,
-            api_name=False
-        )
+    demo.queue(max_size=10).launch()
+        outputs=vqa_input,
+        queue=False,
+        api_name=False
+    )