Upload 3 files

app.py

@@ -0,0 +1,74 @@
+import gradio as gr
+from diffusers import StableDiffusionXLPipeline
+import torch
+
+# Load model (cache locally first)
+def load_model():
+    model_id = "stabilityai/stable-diffusion-xl-base-1.0"
+    pipe = StableDiffusionXLPipeline.from_pretrained(
+        model_id,
+        torch_dtype=torch.float16,
+        variant="fp16",
+        use_safetensors=True
+    )
+    
+    if torch.cuda.is_available():
+        pipe.to("cuda")
+        pipe.enable_xformers_memory_efficient_attention()
+    elif torch.backends.mps.is_available():
+        pipe.to("mps")
+    else:
+        pipe.to("cpu")
+    
+    return pipe
+
+# Image generation function
+def generate_image(prompt, negative_prompt, steps, guidance_scale, width=1024, height=1024):
+    pipe = load_model()
+    
+    # Generate image
+    image = pipe(
+        prompt=prompt,
+        negative_prompt=negative_prompt,
+        num_inference_steps=int(steps),
+        guidance_scale=guidance_scale,
+        width=width,
+        height=height,
+    ).images[0]
+    
+    return image
+
+# Gradio interface
+with gr.Blocks(title="Text to Image Generator") as demo:
+    gr.Markdown("# 🎨 Text-to-Image Generator with SDXL")
+    
+    with gr.Row():
+        with gr.Column():
+            prompt = gr.Textbox(label="Prompt", placeholder="A futuristic cityscape at sunset...")
+            negative_prompt = gr.Textbox(
+                label="Negative Prompt",
+                value="low quality, blurry, distorted, watermark",
+            )
+            steps = gr.Slider(10, 100, value=30, label="Inference Steps")
+            guidance_scale = gr.Slider(1.0, 20.0, value=7.5, label="Guidance Scale")
+            submit = gr.Button("Generate Image")
+        
+        with gr.Column():
+            output_image = gr.Image(label="Generated Image", type="pil")
+
+    submit.click(
+        generate_image,
+        inputs=[prompt, negative_prompt, steps, guidance_scale],
+        outputs=output_image,
+    )
+
+    gr.Examples(
+        examples=[
+            ["A cyberpunk cat wearing neon sunglasses, digital art", "", 30, 7.5],
+            ["A realistic photo of a dinosaur working in a modern office", "cartoonish, drawing", 40, 8.5],
+        ],
+        inputs=[prompt, negative_prompt, steps, guidance_scale],
+    )
+
+if __name__ == "__main__":
+    demo.launch(debug=True)

requirements.txt

@@ -0,0 +1,6 @@
+torch>=2.0.0
+transformers>=4.30.0
+diffusers>=0.20.0
+accelerate>=0.21.0
+gradio>=3.40.0
+xformers==0.0.22