Create app.py

app.py

@@ -0,0 +1,79 @@
+import gradio as gr
+import torch
+from diffusers import AutoPipelineForText2Image
+
+# 2. Load the model
+# We use AutoPipelineForText2Image to automatically load the correct pipeline for SDXL.
+# torch_dtype=torch.float16 is a memory optimization.
+# variant="fp16" tells the model to use a smaller, faster version of the weights.
+# .to("cuda") moves the model to the GPU for fast inference.
+print("Loading model...")
+pipe = AutoPipelineForText2Image.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0", 
+    torch_dtype=torch.float16, 
+    variant="fp16", 
+    use_safetensors=True
+).to("cuda")
+print("Model loaded.")
+
+# 3. Define the image generation function
+# This function will take a text prompt and return a generated image.
+def generate_image(prompt, negative_prompt, steps, guidance):
+    print(f"Generating image for prompt: {prompt}")
+    # The 'pipe' object does all the work. We pass it the prompt and other parameters.
+    # num_inference_steps controls how many steps the model takes to generate the image.
+    # guidance_scale controls how much the model follows the prompt.
+    image = pipe(
+        prompt=prompt, 
+        negative_prompt=negative_prompt,
+        num_inference_steps=steps,
+        guidance_scale=guidance
+    ).images[0]
+    
+    # The output is a PIL Image object, which Gradio can display directly.
+    return image
+
+# 4. Create the Gradio interface
+# This is where we design the web UI.
+with gr.Blocks(theme=gr.themes.Default(primary_hue="blue", secondary_hue="sky")) as demo:
+    gr.Markdown("# 🖼️ Stable Diffusion XL Text-to-Image")
+    gr.Markdown("Enter a text prompt and see the magic of AI-powered image generation!")
+    
+    with gr.Row():
+        with gr.Column(scale=4):
+            # Textbox for the main prompt
+            prompt_input = gr.Textbox(label="Your Prompt", placeholder="An astronaut riding a horse on Mars, cinematic, 4k")
+            # Textbox for the negative prompt
+            negative_prompt_input = gr.Textbox(label="Negative Prompt", placeholder="low quality, blurry, watermark, text")
+            # Submit button
+            submit_btn = gr.Button("Generate Image", variant="primary")
+        
+        with gr.Column(scale=1):
+            # Sliders for advanced options
+            steps_slider = gr.Slider(minimum=10, maximum=100, value=25, step=1, label="Inference Steps")
+            guidance_slider = gr.Slider(minimum=0, maximum=20, value=7.5, step=0.1, label="Guidance Scale")
+    
+    # Image component to display the output
+    output_image = gr.Image(label="Generated Image", width=768, height=768)
+
+    # Define some example prompts to make it easy for users to start
+    gr.Examples(
+        examples=[
+            ["A majestic lion wearing a crown, photorealistic", "cartoon, drawing", 30, 8],
+            ["A beautiful cityscape at sunset, painted by Van Gogh", "blurry, modern", 25, 7.5],
+            ["A cute robot serving coffee in a futuristic cafe, 3D render", "text, humans", 35, 9],
+        ],
+        inputs=[prompt_input, negative_prompt_input, steps_slider, guidance_slider]
+    )
+
+    # 5. Connect the button to the function
+    # When the submit button is clicked, it will call the `generate_image` function
+    # with the values from the input components. The result will be displayed in `output_image`.
+    submit_btn.click(
+        fn=generate_image,
+        inputs=[prompt_input, negative_prompt_input, steps_slider, guidance_slider],
+        outputs=output_image
+    )
+
+# 6. Launch the application
+demo.launch()