app.py

import gradio as gr
import numpy as np
import random
import spaces  # Uncomment if you're using ZeroGPU
from diffusers import DiffusionPipeline
import torch
from tags import participant_tags, tribe_tags, role_tags, skin_tone_tags, body_type_tags, tattoo_tags, piercing_tags, expression_tags, eye_tags, hair_style_tags, position_tags, fetish_tags, location_tags, camera_tags, atmosphere_tags

device = "cuda" if torch.cuda.is_available() else "cpu"
model_repo_id = "John6666/wai-ani-nsfw-ponyxl-v8-sdxl"  # Replace with your desired model

if torch.cuda.is_available():
    torch_dtype = torch.float16
else:
    torch_dtype = torch.float32

pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
pipe.to(device)

MAX_SEED = np.iinfo(np.int32).max
MAX_IMAGE_SIZE = 1024


@spaces.GPU  # Uncomment if using ZeroGPU
def infer(
    prompt, negative_prompt, seed, randomize_seed, width, height, guidance_scale, num_inference_steps,
    selected_participant_tags, selected_tribe_tags, selected_role_tags, selected_skin_tone_tags, selected_body_type_tags,
    selected_tattoo_tags, selected_piercing_tags, selected_expression_tags, selected_eye_tags, selected_hair_style_tags,
    selected_position_tags, selected_fetish_tags, selected_location_tags, selected_camera_tags, selected_atmosphere_tags,
    active_tab, progress=gr.Progress(track_tqdm=True)
):
    # Handle the active tab and generate the prompt accordingly
    if active_tab == "Prompt Input":
        final_prompt = f"score_9, score_8_up, score_7_up, source_anime, {prompt}"
    else:
        tag_list = (
            [participant_tags[tag] for tag in selected_participant_tags] +
            [tribe_tags[tag] for tag in selected_tribe_tags] +
            [role_tags[tag] for tag in selected_role_tags] +
            [skin_tone_tags[tag] for tag in selected_skin_tone_tags] +
            [body_type_tags[tag] for tag in selected_body_type_tags] +
            [tattoo_tags[tag] for tag in selected_tattoo_tags] +
            [piercing_tags[tag] for tag in selected_piercing_tags] +
            [expression_tags[tag] for tag in selected_expression_tags] +
            [eye_tags[tag] for tag in selected_eye_tags] +
            [hair_style_tags[tag] for tag in selected_hair_style_tags] +
            [position_tags[tag] for tag in selected_position_tags] +
            [fetish_tags[tag] for tag in selected_fetish_tags] +
            [location_tags[tag] for tag in selected_location_tags] +
            [camera_tags[tag] for tag in selected_camera_tags] +
            [atmosphere_tags[tag] for tag in selected_atmosphere_tags]
        )
        final_prompt = f"score_9, score_8_up, score_7_up, source_anime, {', '.join(tag_list)}"

    # Concatenate additional negative prompts
    additional_negatives = "worst quality, bad quality, jpeg artifacts, source_cartoon, 3d, (censor), monochrome, blurry, lowres, watermark"
    full_negative_prompt = f"{additional_negatives}, {negative_prompt}"

    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    generator = torch.Generator(device=device).manual_seed(seed)

    image = pipe(
        prompt=final_prompt,
        negative_prompt=full_negative_prompt,
        guidance_scale=guidance_scale,
        num_inference_steps=num_inference_steps,
        width=width,
        height=height,
        generator=generator
    ).images[0]

    return image, seed, f"Prompt: {final_prompt}\nNegative Prompt: {full_negative_prompt}"


css = """
#col-container {
    margin: 0 auto;
    max-width: 640px;
}
"""

with gr.Blocks(css=css) as demo:
    with gr.Column(elem_id="col-container"):
        gr.Markdown("# Image Generator with Tags and Prompts")

        result = gr.Image(label="Result", show_label=False)
        prompt_info = gr.Textbox(label="Prompts Used", lines=3, interactive=False)
        active_tab = gr.State("Prompt Input")

        with gr.Tabs() as tabs:
            # Prompt Input Tab
            with gr.TabItem("Prompt Input"):
                prompt = gr.Textbox(label="Prompt", placeholder="Enter your custom prompt")
                tabs.select(lambda: "Prompt Input", inputs=None, outputs=active_tab)

            # Straight Tab
            with gr.TabItem("Straight"):
                selected_participant_tags = gr.CheckboxGroup(choices=list(participant_tags.keys()), label="Participant Tags")
                selected_tribe_tags = gr.CheckboxGroup(choices=list(tribe_tags.keys()), label="Tribe Tags")
                selected_role_tags = gr.CheckboxGroup(choices=list(role_tags.keys()), label="Role Tags")
                selected_skin_tone_tags = gr.CheckboxGroup(choices=list(skin_tone_tags.keys()), label="Skin Tone Tags")
                selected_body_type_tags = gr.CheckboxGroup(choices=list(body_type_tags.keys()), label="Body Type Tags")
                selected_tattoo_tags = gr.CheckboxGroup(choices=list(tattoo_tags.keys()), label="Tattoo Tags")
                selected_piercing_tags = gr.CheckboxGroup(choices=list(piercing_tags.keys()), label="Piercing Tags")
                selected_expression_tags = gr.CheckboxGroup(choices=list(expression_tags.keys()), label="Expression Tags")
                selected_eye_tags = gr.CheckboxGroup(choices=list(eye_tags.keys()), label="Eye Tags")
                selected_hair_style_tags = gr.CheckboxGroup(choices=list(hair_style_tags.keys()), label="Hair Style Tags")
                selected_position_tags = gr.CheckboxGroup(choices=list(position_tags.keys()), label="Position Tags")
                selected_fetish_tags = gr.CheckboxGroup(choices=list(fetish_tags.keys()), label="Fetish Tags")
                selected_location_tags = gr.CheckboxGroup(choices=list(location_tags.keys()), label="Location Tags")
                selected_camera_tags = gr.CheckboxGroup(choices=list(camera_tags.keys()), label="Camera Tags")
                selected_atmosphere_tags = gr.CheckboxGroup(choices=list(atmosphere_tags.keys()), label="Atmosphere Tags")
                tabs.select(lambda: "Straight", inputs=None, outputs=active_tab)

            # Gay Tab
            with gr.TabItem("Gay"):
                selected_participant_tags = gr.CheckboxGroup(choices=list(participant_tags.keys()), label="Participant Tags")
                selected_tribe_tags = gr.CheckboxGroup(choices=list(tribe_tags.keys()), label="Tribe Tags")
                selected_role_tags = gr.CheckboxGroup(choices=list(role_tags.keys()), label="Role Tags")
                selected_skin_tone_tags = gr.CheckboxGroup(choices=list(skin_tone_tags.keys()), label="Skin Tone Tags")
                selected_body_type_tags = gr.CheckboxGroup(choices=list(body_type_tags.keys()), label="Body Type Tags")
                selected_tattoo_tags = gr.CheckboxGroup(choices=list(tattoo_tags.keys()), label="Tattoo Tags")
                selected_piercing_tags = gr.CheckboxGroup(choices=list(piercing_tags.keys()), label="Piercing Tags")
                selected_expression_tags = gr.CheckboxGroup(choices=list(expression_tags.keys()), label="Expression Tags")
                selected_eye_tags = gr.CheckboxGroup(choices=list(eye_tags.keys()), label="Eye Tags")
                selected_hair_style_tags = gr.CheckboxGroup(choices=list(hair_style_tags.keys()), label="Hair Style Tags")
                selected_position_tags = gr.CheckboxGroup(choices=list(position_tags.keys()), label="Position Tags")
                selected_fetish_tags = gr.CheckboxGroup(choices=list(fetish_tags.keys()), label="Fetish Tags")
                selected_location_tags = gr.CheckboxGroup(choices=list(location_tags.keys()), label="Location Tags")
                selected_camera_tags = gr.CheckboxGroup(choices=list(camera_tags.keys()), label="Camera Tags")
                selected_atmosphere_tags = gr.CheckboxGroup(choices=list(atmosphere_tags.keys()), label="Atmosphere Tags")
                tabs.select(lambda: "Gay", inputs=None, outputs=active_tab)

            # Lesbian Tab
            with gr.TabItem("Lesbian"):
                selected_participant_tags = gr.CheckboxGroup(choices=list(participant_tags.keys()), label="Participant Tags")
                selected_tribe_tags = gr.CheckboxGroup(choices=list(tribe_tags.keys()), label="Tribe Tags")
                selected_role_tags = gr.CheckboxGroup(choices=list(role_tags.keys()), label="Role Tags")
                selected_skin_tone_tags = gr.CheckboxGroup(choices=list(skin_tone_tags.keys()), label="Skin Tone Tags")
                selected_body_type_tags = gr.CheckboxGroup(choices=list(body_type_tags.keys()), label="Body Type Tags")
                selected_tattoo_tags = gr.CheckboxGroup(choices=list(tattoo_tags.keys()), label="Tattoo Tags")
                selected_piercing_tags = gr.CheckboxGroup(choices=list(piercing_tags.keys()), label="Piercing Tags")
                selected_expression_tags = gr.CheckboxGroup(choices=list(expression_tags.keys()), label="Expression Tags")
                selected_eye_tags = gr.CheckboxGroup(choices=list(eye_tags.keys()), label="Eye Tags")
                selected_hair_style_tags = gr.CheckboxGroup(choices=list(hair_style_tags.keys()), label="Hair Style Tags")
                selected_position_tags = gr.CheckboxGroup(choices=list(position_tags.keys()), label="Position Tags")
                selected_fetish_tags = gr.CheckboxGroup(choices=list(fetish_tags.keys()), label="Fetish Tags")
                selected_location_tags = gr.CheckboxGroup(choices=list(location_tags.keys()), label="Location Tags")
                selected_camera_tags = gr.CheckboxGroup(choices=list(camera_tags.keys()), label="Camera Tags")
                selected_atmosphere_tags = gr.CheckboxGroup(choices=list(atmosphere_tags.keys()), label="Atmosphere Tags")
                tabs.select(lambda: "Lesbian", inputs=None, outputs=active_tab)

        # Advanced Settings
        with gr.Accordion("Advanced Settings", open=False):
            negative_prompt = gr.Textbox(label="Negative Prompt", placeholder="Enter negative prompt")
            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():
                guidance_scale = gr.Slider(label="Guidance Scale", minimum=0.0, maximum=10.0, step=0.1, value=7)
                num_inference_steps = gr.Slider(label="Number of Inference Steps", minimum=1, maximum=50, step=1, value=35)

        run_button = gr.Button("Run")
        run_button.click(
            infer,
            inputs=[prompt, negative_prompt, seed, randomize_seed, width, height, guidance_scale, num_inference_steps,
                    selected_participant_tags, selected_tribe_tags, selected_role_tags, selected_skin_tone_tags, selected_body_type_tags,
                    selected_tattoo_tags, selected_piercing_tags, selected_expression_tags, selected_eye_tags,
                    selected_hair_style_tags, selected_position_tags, selected_fetish_tags, selected_location_tags,
                    selected_camera_tags, selected_atmosphere_tags, active_tab],
            outputs=[result, seed, prompt_info]
        )

demo.queue().launch()