app.py

import gradio as gr
import numpy as np
import random
import spaces
import torch
from diffusers import DiffusionPipeline
from transformers import CLIPTokenizer

dtype = torch.bfloat16
device = "cuda" if torch.cuda.is_available() else "cpu"

# Initialize CLIP tokenizer for prompt length checking
tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-base-patch32")

pipe = DiffusionPipeline.from_pretrained(
    "UnfilteredAI/NSFW-Flux-v1",
    torch_dtype=dtype
).to(device)

MAX_SEED = np.iinfo(np.int32).max
MAX_IMAGE_SIZE = 2048
MAX_TOKENS = 77  # CLIP's maximum token length

def truncate_prompt(prompt):
    """Truncate the prompt to fit within CLIP's token limit"""
    tokens = tokenizer.encode(prompt, truncation=True, max_length=MAX_TOKENS)
    return tokenizer.decode(tokens)

@spaces.GPU()
def infer(
    prompt,
    seed=42,
    randomize_seed=False,
    width=1024,
    height=1024,
    num_inference_steps=4,
    progress=gr.Progress(track_tqdm=True)
):
    # Truncate prompt if necessary
    truncated_prompt = truncate_prompt(prompt)
    
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    
    generator = torch.Generator().manual_seed(seed)
    
    try:
        image = pipe(
            prompt=truncated_prompt,
            width=width,
            height=height,
            num_inference_steps=num_inference_steps,
            generator=generator,
            guidance_scale=0.0
        ).images[0]
        
        return image, seed
    except Exception as e:
        raise gr.Error(f"Error generating image: {str(e)}")

examples = [
    "a tiny astronaut hatching from an egg on the moon",
    "a cat holding a sign that says hello world",
    "an anime illustration of a wiener schnitzel",
]

css = """
#col-container {
    margin: 0 auto;
    max-width: 520px;
}
"""

with gr.Blocks(css=css) as demo:
    with gr.Column(elem_id="col-container"):
        gr.Markdown("""
            NSFW-Flux-v1 is a 12 billion parameter rectified flow transformer 
            capable of generating images from text descriptions. 
            Finetuned by UnfilteredAI, this model is designed to produce 
            a wide range of images, including explicit and NSFW 
            (Not Safe For Work) images from textual inputs.
            
            Note: Long prompts will be automatically truncated to fit the model's requirements.
        """)
        
        with gr.Row():
            prompt = gr.Text(
                label="Prompt",
                show_label=False,
                max_lines=1,
                placeholder="Enter your prompt",
                container=False,
            )
            run_button = gr.Button("Run", scale=0)
        
        result = gr.Image(label="Result", show_label=False)
        
        with gr.Accordion("Advanced Settings", open=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=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )
                height = gr.Slider(
                    label="Height",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )
            
            with gr.Row():
                num_inference_steps = gr.Slider(
                    label="Number of inference steps",
                    minimum=1,
                    maximum=50,
                    step=1,
                    value=4,
                )
        
        gr.Examples(
            examples=examples,
            fn=infer,
            inputs=[prompt],
            outputs=[result, seed],
            cache_examples="lazy"
        )
        
        gr.on(
            triggers=[run_button.click, prompt.submit],
            fn=infer,
            inputs=[
                prompt,
                seed,
                randomize_seed,
                width,
                height,
                num_inference_steps
            ],
            outputs=[result, seed]
        )

demo.launch()