Update app.py

app.py

@@ -1,13 +1,13 @@
 import gradio as gr
 import google.generativeai as genai
 from PIL import Image
-import os
 import io
 
 # Initialize Gemini
-GOOGLE_API_KEY = "AIzaSyBdz-qcLFRDsR-mm37AlRf2w6RZws2lDL0"  # Your API key
+GOOGLE_API_KEY = "AIzaSyBdz-qcLFRDsR-mm37AlRf2w6RZws2lDL0"  # Replace with your actual API key
 genai.configure(api_key=GOOGLE_API_KEY)
 
+# Function to process the image with Gemini
 def process_image(image):
     try:
         prompt = """
@@ -30,23 +30,16 @@ def process_image(image):
             "max_output_tokens": 1024,
         }
         
-        # Convert image to bytes if it's not already
-        if isinstance(image, str):
-            with open(image, "rb") as f:
-                image_data = f.read()
-        else:
-            # Convert PIL Image to bytes
-            image_bytes = io.BytesIO()
-            image.save(image_bytes, format='PNG')
-            image_data = image_bytes.getvalue()
+        # Convert PIL Image to bytes
+        image_bytes = io.BytesIO()
+        image.save(image_bytes, format='PNG')
+        image_data = image_bytes.getvalue()
         
-        # Create content parts
         content_parts = [
             {"text": prompt},
             {"inline_data": {"mime_type": "image/jpeg", "data": image_data}}
         ]
         
-        # Generate response with specific parameters
         response = model.generate_content(
             content_parts,
             generation_config=generation_config,
@@ -58,42 +51,48 @@ def process_image(image):
             ]
         )
         
-        description = response.text
+        return response.text
         
-        # Pass the description to FLUX model
-        flux_interface = gr.load(
-            "models/black-forest-labs/FLUX.1-dev",
-            provider="together"
-        )
-        
-        # Generate image based on the description
-        generated_image = flux_interface(description)
-        
-        return description, generated_image
-    
     except Exception as e:
-        print(f"Error: {str(e)}")
-        return f"Error processing image: {str(e)}", None
+        return f"Error processing image: {str(e)}"
+
+# Function to generate image from text using Hugging Face model
+def generate_image_from_text(text):
+    try:
+        # Load the text-to-image model from Hugging Face Spaces
+        hf_model = gr.load("models/black-forest-labs/FLUX.1-dev", provider="together")
+        generated_image = hf_model(text)  # Pass the text to the model
+        return generated_image
+    except Exception as e:
+        return f"Error generating image: {str(e)}"
 
-# Create a simpler Gradio interface to avoid compatibility issues
-with gr.Blocks() as demo:
-    gr.Markdown("# Image Analysis and Generation")
-    gr.Markdown("Upload an image to get a detailed description from Gemini, then generate a new image based on that description using FLUX.1-dev.")
+# Combined function to process the workflow
+def full_workflow(image):
+    # Step 1: Get detailed analysis from Gemini
+    analysis_text = process_image(image)
+    
+    # Step 2: Pass the analysis text to the text-to-image model
+    generated_image = generate_image_from_text(analysis_text)
     
+    # Return both the text analysis and the generated image
+    return analysis_text, generated_image
+
+# Create Gradio interface
+with gr.Blocks(title="Image Analysis and Regeneration") as demo:
+    gr.Markdown("### Upload an Image to Analyze and Regenerate")
+    with gr.Row():
+        image_input = gr.Image(type="pil", label="Upload Image")
+        submit_btn = gr.Button("Analyze and Generate")
     with gr.Row():
-        with gr.Column():
-            input_image = gr.Image(type="pil")
-            submit_btn = gr.Button("Analyze and Generate")
-            
-        with gr.Column():
-            output_text = gr.Textbox(lines=15, label="Image Description")
-            output_image = gr.Image(label="Generated Image")
+        text_output = gr.Textbox(label="Gemini Analysis", lines=15)
+        image_output = gr.Image(label="Generated Image")
     
+    # Connect the button to the full workflow
     submit_btn.click(
-        fn=process_image,
-        inputs=input_image,
-        outputs=[output_text, output_image]
+        fn=full_workflow,
+        inputs=image_input,
+        outputs=[text_output, image_output]
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(share=True)