Create app.py

app.py

@@ -0,0 +1,208 @@
+import gradio as gr
+import requests
+import os
+import base64
+from io import BytesIO
+from PIL import Image
+import json
+
+def generate_image(prompt, cfg_scale, width, height, seed, steps, input_image=None):
+    """
+    Generate image using NVIDIA's Flux.1-dev API
+    """
+    try:
+        # Get API key from environment
+        api_key = os.getenv("NVIDIA_API_KEY")
+        if not api_key:
+            return None, "Error: NVIDIA_API_KEY environment variable not set"
+        
+        # API endpoint
+        invoke_url = "https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-dev"
+        
+        # Headers
+        headers = {
+            "Authorization": f"Bearer {api_key}",
+            "Accept": "application/json",
+        }
+        
+        # Prepare payload
+        payload = {
+            "prompt": prompt,
+            "cfg_scale": cfg_scale,
+            "width": int(width),
+            "height": int(height),
+            "seed": int(seed) if seed != -1 else 0,
+            "steps": int(steps)
+        }
+        
+        # Handle input image for img2img or control modes
+        if input_image is not None:
+            # Convert PIL image to base64
+            buffered = BytesIO()
+            input_image.save(buffered, format="PNG")
+            img_str = base64.b64encode(buffered.getvalue()).decode()
+            payload["image"] = f"data:image/png;base64,{img_str}"
+            payload["mode"] = "canny"  # You can make this configurable
+        
+        # Make API request
+        response = requests.post(invoke_url, headers=headers, json=payload, timeout=120)
+        response.raise_for_status()
+        
+        response_body = response.json()
+        
+        # Check if response contains image data
+        if "image" in response_body:
+            # Decode base64 image
+            image_data = response_body["image"]
+            if image_data.startswith("data:image"):
+                # Remove data URL prefix
+                image_data = image_data.split(",")[1]
+            
+            # Decode and create PIL Image
+            image_bytes = base64.b64decode(image_data)
+            image = Image.open(BytesIO(image_bytes))
+            
+            return image, f"✅ Image generated successfully!\nSeed used: {response_body.get('seed', 'unknown')}"
+        else:
+            return None, f"❌ Error: No image in response\n{json.dumps(response_body, indent=2)}"
+            
+    except requests.exceptions.RequestException as e:
+        return None, f"❌ API Request Error: {str(e)}"
+    except Exception as e:
+        return None, f"❌ Error: {str(e)}"
+
+def create_gradio_interface():
+    """
+    Create and configure the Gradio interface
+    """
+    
+    with gr.Blocks(title="NVIDIA Flux.1-dev Image Generator", theme=gr.themes.Soft()) as demo:
+        gr.Markdown(
+            """
+            # 🎨 NVIDIA Flux.1-dev Image Generator
+            Generate high-quality images using NVIDIA's Flux.1-dev model via their API.
+            
+            **Requirements:** Set your `NVIDIA_API_KEY` environment variable before running.
+            """
+        )
+        
+        with gr.Row():
+            with gr.Column(scale=1):
+                # Input controls
+                prompt = gr.Textbox(
+                    label="Prompt",
+                    placeholder="Describe the image you want to generate...",
+                    lines=3,
+                    value="a simple coffee shop interior"
+                )
+                
+                input_image = gr.Image(
+                    label="Input Image (optional - for img2img/control)",
+                    type="pil",
+                    height=300
+                )
+                
+                with gr.Row():
+                    cfg_scale = gr.Slider(
+                        minimum=1.0,
+                        maximum=20.0,
+                        value=3.5,
+                        step=0.5,
+                        label="CFG Scale",
+                        info="Higher values follow prompt more closely"
+                    )
+                    
+                    steps = gr.Slider(
+                        minimum=10,
+                        maximum=100,
+                        value=50,
+                        step=5,
+                        label="Steps",
+                        info="More steps = higher quality but slower"
+                    )
+                
+                with gr.Row():
+                    width = gr.Dropdown(
+                        choices=[512, 768, 1024, 1280],
+                        value=1024,
+                        label="Width"
+                    )
+                    
+                    height = gr.Dropdown(
+                        choices=[512, 768, 1024, 1280],
+                        value=1024,
+                        label="Height"
+                    )
+                
+                seed = gr.Number(
+                    label="Seed",
+                    value=-1,
+                    info="Use -1 for random seed"
+                )
+                
+                generate_btn = gr.Button(
+                    "🎨 Generate Image",
+                    variant="primary",
+                    size="lg"
+                )
+            
+            with gr.Column(scale=1):
+                # Output
+                output_image = gr.Image(
+                    label="Generated Image",
+                    type="pil",
+                    height=500
+                )
+                
+                status_text = gr.Textbox(
+                    label="Status",
+                    lines=3,
+                    interactive=False
+                )
+        
+        # Example prompts
+        gr.Markdown("### 💡 Example Prompts")
+        examples = gr.Examples(
+            examples=[
+                ["a cozy coffee shop with warm lighting and vintage furniture"],
+                ["a futuristic cityscape at sunset with flying cars"],
+                ["a magical forest with glowing mushrooms and fairy lights"],
+                ["a minimalist modern kitchen with marble countertops"],
+                ["a steampunk laboratory with brass machinery and glowing tubes"],
+                ["a serene mountain lake reflecting snow-capped peaks"],
+            ],
+            inputs=[prompt]
+        )
+        
+        # Event handlers
+        generate_btn.click(
+            fn=generate_image,
+            inputs=[prompt, cfg_scale, width, height, seed, steps, input_image],
+            outputs=[output_image, status_text]
+        )
+        
+        # Keyboard shortcut
+        prompt.submit(
+            fn=generate_image,
+            inputs=[prompt, cfg_scale, width, height, seed, steps, input_image],
+            outputs=[output_image, status_text]
+        )
+    
+    return demo
+
+if __name__ == "__main__":
+    # Check if API key is set
+    if not os.getenv("NVIDIA_API_KEY"):
+        print("⚠️  Warning: NVIDIA_API_KEY environment variable not set!")
+        print("Please set it before running the app:")
+        print("export NVIDIA_API_KEY=your_api_key_here")
+        print()
+    
+    # Create and launch the app
+    demo = create_gradio_interface()
+    demo.launch(
+        share=False,  # Set to True if you want a public link
+        server_name="0.0.0.0",  # Allow external connections
+        server_port=7860,
+        show_error=True
+    )