init

app.py

@@ -0,0 +1,106 @@
+import gradio as gr
+import torch
+from diffusers import StableDiffusionPipeline
+from PIL import Image
+import numpy as np
+import spaces
+
+MODEL_ID = "runwayml/stable-diffusion-v1-5"
+
+# Global pipeline variable
+pipe = None
+
+
+def initialize_pipeline():
+    """Initialize the pipeline if not already loaded."""
+    global pipe
+    if pipe is None:
+        device = "cuda" if torch.cuda.is_available() else "cpu"
+        print(f"Initializing pipeline on device: {device}")
+
+        pipe = StableDiffusionPipeline.from_pretrained(
+            MODEL_ID,
+            torch_dtype=torch.float16 if device == "cuda" else torch.float32,
+        )
+        pipe = pipe.to(device)
+    return pipe
+
+
+@spaces.GPU
+def generate_image(prompt, seed, num_inference_steps):
+    """
+    Generate an image using Stable Diffusion.
+    This function runs on Zero GPU when deployed on Hugging Face Spaces.
+
+    Args:
+        prompt: Text description of the image to generate
+        seed: Random seed for reproducibility
+        num_inference_steps: Number of denoising steps
+
+    Returns:
+        PIL Image
+    """
+    # Initialize pipeline
+    pipeline = initialize_pipeline()
+    device = pipeline.device
+
+    # Set the random seed for reproducibility
+    generator = torch.Generator(device=device).manual_seed(int(seed))
+
+    # Generate the image
+    with torch.no_grad():
+        result = pipeline(
+            prompt=prompt,
+            num_inference_steps=int(num_inference_steps),
+            generator=generator,
+        )
+
+    return result.images[0]
+
+
+def create_interface():
+    """Create and configure the Gradio interface."""
+    # Create the Gradio interface
+    interface = gr.Interface(
+        fn=generate_image,
+        inputs=[
+            gr.Textbox(
+                label="Prompt",
+                placeholder="Enter a text description of the image you want to generate...",
+                lines=3,
+            ),
+            gr.Slider(
+                minimum=0,
+                maximum=2147483647,
+                value=42,
+                step=1,
+                label="Seed",
+                info="Random seed for reproducibility",
+            ),
+            gr.Slider(
+                minimum=1,
+                maximum=150,
+                value=50,
+                step=1,
+                label="Diffusion Steps",
+                info="Number of denoising steps (more steps = higher quality but slower)",
+            ),
+        ],
+        outputs=gr.Image(label="Generated Image", type="pil"),
+        title="Stable Diffusion Image Generator",
+        description="Generate images from text using Stable Diffusion. Enter a prompt, set the seed for reproducibility, and adjust the number of diffusion steps.",
+        examples=[
+            ["A beautiful sunset over mountains", 42, 50],
+            ["A cat wearing a space suit, digital art", 123, 50],
+            ["Cyberpunk city at night, neon lights", 456, 75],
+        ],
+        cache_examples=False,
+    )
+
+    return interface
+
+
+if __name__ == "__main__":
+    # Create and launch the interface
+    demo = create_interface()
+    demo.launch(share=False, server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -0,0 +1,11 @@
+torch>=2.0.0
+torchvision>=0.15.0
+numpy>=1.24.0
+Pillow>=9.5.0
+tqdm>=4.65.0
+transformers>=4.30.0
+diffusers>=0.27.0
+accelerate>=0.20.0
+safetensors>=0.3.1
+gradio>=4.0.0
+spaces>=0.19.0