Update app.py

app.py

@@ -63,14 +63,6 @@
 # if __name__ == "__main__":
 #     demo.launch()
 
-import gradio as gr
-from huggingface_hub import InferenceClient
-from PIL import Image
-import io
-import base64
-
-client = InferenceClient("meta-llama/Llama-3.2-11B-Vision-Instruct")
-
 def respond(
     message,
     history: list[tuple[str, str]],
@@ -78,9 +70,8 @@ def respond(
     max_tokens,
     temperature,
     top_p,
-    image: Image,  # Add image input to the function
+    image: Image,  # Image input
 ):
-    # Prepare the system message and history for the conversation
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -89,48 +80,31 @@ def respond(
         if val[1]:
             messages.append({"role": "assistant", "content": val[1]})
 
-    # Add the current user message
+    # Add current user message
     messages.append({"role": "user", "content": message})
 
-    # Convert the image to a base64-encoded string
+    # Convert image to base64
     image_bytes = io.BytesIO()
     image.save(image_bytes, format='PNG')
     image_bytes.seek(0)
     image_base64 = base64.b64encode(image_bytes.getvalue()).decode('utf-8')
 
-    # Use InferenceClient to handle the image and text input to the model
-    # Pass the base64-encoded image as the input
-    response_data = client.text_to_image(images=image_base64, prompt=message)  # Pass the base64 string as 'images'
-
-    # Debug: print the response data to check its content
-    print("Response Data: ", response_data)
-    
     try:
-        # Assuming the response is in the correct format, check for image content
+        # Pass the base64 image and prompt
+        response_data = client.text_to_image(images=image_base64, prompt=message)
+        
+        # Print out the response data for debugging
+        print("Response Data:", response_data)
+
+        # Check if the response contains an image
         if isinstance(response_data, dict) and 'image' in response_data:
+            # Decode the base64 image response
             image_response = response_data['image']
-            # Decode the base64 image back into an image object
             image_bytes = base64.b64decode(image_response)
             image = Image.open(io.BytesIO(image_bytes))
-            image.show()  # Or return the image in Gradio
+            image.show()  # Display the image or return it
             return "Image processed successfully"
         else:
-            return f"Error: Unexpected response format or no image data found. Response: {response_data}"
+            return f"Error: No valid image found in response. Response: {response_data}"
     except Exception as e:
         return f"Error processing image: {e}"
-
-# Create the Gradio interface with an image input
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
-        gr.Image(type="pil", label="Upload an Image"),  # Image input for vision tasks
-    ],
-)
-
-if __name__ == "__main__":
-    demo.launch(share=True)
-  # Set share=True to create a public link