import gradio as gr import numpy as np import random import spaces from diffusers import DiffusionPipeline import torch device = "cuda" if torch.cuda.is_available() else "cpu" model_repo_id = "stabilityai/stable-diffusion-3.5-large" if torch.cuda.is_available(): torch_dtype = torch.bfloat16 else: torch_dtype = torch.float32 pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype) pipe = pipe.to(device) MAX_SEED = np.iinfo(np.int32).max MAX_IMAGE_SIZE = 1024 @spaces.GPU(duration=65) def infer( prompt, negative_prompt="", seed=42, randomize_seed=False, width=1024, height=1024, guidance_scale=4.5, num_inference_steps=40, progress=gr.Progress(track_tqdm=True), ): if randomize_seed: seed = random.randint(0, MAX_SEED) generator = torch.Generator().manual_seed(seed) image = pipe( prompt=prompt, negative_prompt=negative_prompt, guidance_scale=guidance_scale, num_inference_steps=num_inference_steps, width=width, height=height, generator=generator, ).images[0] return image, seed # Enhanced examples with creative prompts examples = [ "A capybara wearing a suit holding a sign that reads Hello World", "A steampunk-style flying ship made of brass and wood, floating through cotton candy clouds", "A magical library where books are flying and glowing, with a wise owl librarian", "A cyberpunk street food vendor selling neon-colored dumplings in the rain", "A group of penguins having a formal tea party in the Antarctic", "A treehouse city at sunset with bioluminescent plants and floating lanterns" ] # Custom CSS with modern styling css = """ :root { --primary-color: #7B2CBF; --secondary-color: #9D4EDD; --background-color: #10002B; --text-color: #E0AAFF; --card-bg: #240046; } #col-container { max-width: 850px !important; margin: 0 auto; padding: 20px; background: var(--background-color); border-radius: 15px; box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1); } .main-title { color: var(--text-color) !important; text-align: center; font-size: 2.5em !important; margin-bottom: 1em !important; text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.3); } .gradio-container { background: var(--background-color) !important; color: var(--text-color) !important; } .gr-button { background: var(--primary-color) !important; border: none !important; color: white !important; transition: transform 0.2s !important; } .gr-button:hover { transform: translateY(-2px) !important; background: var(--secondary-color) !important; } .gr-input, .gr-box { background: var(--card-bg) !important; border: 1px solid var(--primary-color) !important; color: var(--text-color) !important; } .footer-custom a { color: var(--text-color); text-decoration: none; margin: 0 10px; transition: color 0.3s; } .footer-custom a:hover { color: var(--secondary-color); text-decoration: underline; } """ # Footer HTML footer = """ """ with gr.Blocks(css=css) as demo: with gr.Column(elem_id="col-container"): gr.HTML( '

Stable Diffusion 3.5 Large (8B)

' '
' 'Visit Stability.ai' '
' ) with gr.Row(): prompt = gr.Text( label="Prompt", show_label=False, max_lines=1, placeholder="Enter your prompt", container=False, ) run_button = gr.Button("Generate", scale=0, variant="primary") result = gr.Image(label="Result", show_label=False) with gr.Accordion("Advanced Settings", open=False): negative_prompt = gr.Text( label="Negative prompt", max_lines=1, placeholder="Enter a negative prompt", visible=False, ) seed = gr.Slider( label="Seed", minimum=0, maximum=MAX_SEED, step=1, value=0, ) randomize_seed = gr.Checkbox(label="Randomize seed", value=True) with gr.Row(): width = gr.Slider( label="Width", minimum=512, maximum=MAX_IMAGE_SIZE, step=32, value=1024, ) height = gr.Slider( label="Height", minimum=512, maximum=MAX_IMAGE_SIZE, step=32, value=1024, ) with gr.Row(): guidance_scale = gr.Slider( label="Guidance scale", minimum=0.0, maximum=7.5, step=0.1, value=4.5, ) num_inference_steps = gr.Slider( label="Number of inference steps", minimum=1, maximum=50, step=1, value=40, ) gr.Examples( examples=examples, inputs=[prompt], outputs=[result, seed], fn=infer, cache_examples=True, cache_mode="lazy" ) gr.HTML(footer) gr.on( triggers=[run_button.click, prompt.submit], fn=infer, inputs=[ prompt, negative_prompt, seed, randomize_seed, width, height, guidance_scale, num_inference_steps, ], outputs=[result, seed], ) if __name__ == "__main__": demo.launch()