Update app.py

app.py

@@ -68,9 +68,6 @@ from huggingface_hub import InferenceClient
 from PIL import Image
 import io
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
 client = InferenceClient("meta-llama/Llama-3.2-11B-Vision-Instruct")
 
 def respond(
@@ -82,7 +79,6 @@ def respond(
     top_p,
     image: Image
 ):
-    # Prepare messages
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -93,27 +89,23 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    # Convert the image to bytes for HuggingFace API processing
+    # Convert image to the format expected by the model
     image_bytes = io.BytesIO()
     image.save(image_bytes, format='PNG')
     image_bytes.seek(0)
 
-    # Make the API call with the image and messages
+    # Use a different method for image inputs if chat_completion does not support it
     response = ""
 
-    for result in client.chat_completion(
-        messages,
-        image=image_bytes,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = result.choices[0].delta.content
-        response += token
-        yield response
-
-# Gradio demo for ChatInterface with image support
+    # Modify here to use the correct API for image + text queries (check model docs)
+    response_data = client.text2image(images=image_bytes, text=message)  # example method; adjust according to actual API
+
+    # Process the response (if any) from the image model
+    for result in response_data:
+        response += result['text']  # or adjust based on response format
+
+    return response
+
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
@@ -134,3 +126,4 @@ demo = gr.ChatInterface(
 if __name__ == "__main__":
     demo.launch()
 
+