app.py

import os
import subprocess
import sys
import io
import gradio as gr
import numpy as np
import random
import spaces
import torch
from diffusers import Flux2Pipeline, Flux2Transformer2DModel
from diffusers import BitsAndBytesConfig as DiffBitsAndBytesConfig
from optimization import optimize_pipeline_
import requests
from PIL import Image
import json

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

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


def remote_text_encoder(prompts):
    response = requests.post(
        "https://remote-text-encoder-flux-2.huggingface.co/predict",
        json={"prompt": prompts},
        headers={
            "Authorization": f"Bearer {os.environ['HF_TOKEN']}",
            "Content-Type": "application/json",
        },
    )

    assert response.status_code == 200, f"{response.status_code=}"

    prompt_embeds = torch.load(io.BytesIO(response.content))

    return prompt_embeds


# Load model
repo_id = "black-forest-labs/FLUX.2-dev"

dit = Flux2Transformer2DModel.from_pretrained(
    repo_id, subfolder="transformer", torch_dtype=torch.bfloat16
)

pipe = Flux2Pipeline.from_pretrained(
    repo_id, text_encoder=None, transformer=dit, torch_dtype=torch.bfloat16
)
pipe.to("cuda")

pipe.transformer.set_attention_backend("_flash_3_hub")

try:
    optimize_pipeline_(
        pipe,
        image=[Image.new("RGB", (1024, 1024))],
        prompt_embeds=remote_text_encoder("prompt").to("cuda"),
        guidance_scale=2.5,
        width=1024,
        height=1024,
        num_inference_steps=1,
    )
except Exception as e:
    print(f"Optimization failed: {e}")


def get_duration(
    prompt,
    input_images=None,
    seed=42,
    randomize_seed=False,
    width=1024,
    height=1024,
    num_inference_steps=50,
    guidance_scale=2.5,
    progress=gr.Progress(track_tqdm=True),
):
    num_images = 0 if input_images is None else len(input_images)
    step_duration = 1 + 0.7 * num_images
    return num_inference_steps * step_duration + 10


@spaces.GPU(duration=get_duration)
def infer(
    prompt,
    input_images=None,
    seed=42,
    randomize_seed=False,
    width=1024,
    height=1024,
    num_inference_steps=50,
    guidance_scale=2.5,
    progress=gr.Progress(track_tqdm=True),
):

    if randomize_seed:
        seed = random.randint(0, MAX_SEED)

    # Get prompt embeddings from remote text encoder
    progress(0.1, desc="Encoding prompt...")
    try:
        prompt_embeds = remote_text_encoder(prompt).to("cuda")
    except Exception as e:
        raise gr.Error(f"Remote text encoder failed: {e}")

    # Prepare image list (convert None or empty gallery to None)
    image_list = None
    if input_images is not None and len(input_images) > 0:
        image_list = []
        for item in input_images:
            image_list.append(item[0])

    # Generate image
    progress(0.3, desc="Generating image...")
    generator = torch.Generator(device=device).manual_seed(seed)
    image = pipe(
        prompt_embeds=prompt_embeds,
        image=image_list,
        width=width,
        height=height,
        num_inference_steps=num_inference_steps,
        guidance_scale=guidance_scale,
        generator=generator,
    ).images[0]

    return image, seed


# --- UI Configuration ---

css = """
@import url('https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600&display=swap');

html, body, .gradio-container {
    background-color: #000000 !important;
    color: #ffffff !important;
    font-family: 'Inter', sans-serif !important;
    margin: 0;
    padding: 0 !important;
    overflow: hidden !important;
    height: 100vh !important;
    max-height: 100vh !important;
    width: 100vw !important;
    max-width: 100vw !important;
    --color-background-primary: #000000;
    --color-background-secondary: #050505;
    --color-border-primary: #171717;
    --color-text-primary: #ffffff;
    --color-text-secondary: #a3a3a3;
}

footer {
    display: none !important;
}

/* Layout */
#main-container {
    position: fixed !important;
    top: 0;
    left: 0;
    width: 100vw !important;
    height: 100vh !important;
    gap: 0 !important;
    display: flex;
    flex-wrap: nowrap;
    overflow: hidden;
    z-index: 10;
}

#right-sidebar {
    background-color: #050505;
    border-left: 1px solid #171717;
    width: 320px !important;
    max-width: 320px !important;
    flex: none !important;
    padding: 0 !important;
    height: 100%;
    overflow-y: auto;
}

#center-canvas {
    background-color: #090909;
    flex-grow: 1 !important;
    display: flex;
    flex-direction: column;
    justify-content: center;
    align-items: center;
    padding: 20px;
    background-image: radial-gradient(#151515 1px, transparent 1px);
    background-size: 20px 20px;
    height: 100%;
    position: relative;
}

/* Components */
#generate-btn {
    background: #ffffff !important;
    color: #000000 !important;
    border-radius: 6px !important;
    font-weight: 600 !important;
    text-transform: uppercase;
    font-size: 11px !important;
    border: none !important;
}

#prompt-input textarea {
    background-color: #000000 !important;
    border: 1px solid #262626 !important;
    color: white !important;
    border-radius: 8px !important;
}

#prompt-input span {
    display: none; /* Hide default label if needed, or style it */
}

/* Accordions */
.accordion {
    background: transparent !important;
    border: none !important;
    border-bottom: 1px solid #171717 !important;
}

.accordion-label {
    font-size: 11px !important;
    font-weight: 600 !important;
    text-transform: uppercase;
    color: #a3a3a3 !important;
}

/* Sliders */
input[type=range] {
    accent-color: white !important;
}

/* Gallery in Sidebar */
#history-gallery {
    flex-grow: 1;
    overflow-y: auto;
    padding: 10px;
}

#history-gallery .grid-wrap {
    grid-template-columns: 1fr !important; /* Force list view */
}

/* Main Image */
#main-image {
    background: transparent !important;
    border: 1px solid #171717;
    border-radius: 8px;
    overflow: hidden;
    box-shadow: 0 20px 25px -5px rgba(0, 0, 0, 0.1), 0 10px 10px -5px rgba(0, 0, 0, 0.04);
}

/* Scrollbars */
::-webkit-scrollbar {
    width: 6px;
    height: 6px;
}
::-webkit-scrollbar-track {
    background: #000000; 
}
::-webkit-scrollbar-thumb {
    background: #333; 
    border-radius: 3px;
}
"""

controls_header_html = """
<div style="padding: 20px 20px 10px 20px;">
    <h2 style="font-size: 11px; font-weight: 600; text-transform: uppercase; color: #666; margin: 0;">Configuration</h2>
</div>
"""

with gr.Blocks(title="FLUX.2 [dev]") as demo:
    with gr.Row(elem_id="main-container", variant="compact"):
        # --- Center Canvas ---
        with gr.Column(elem_id="center-canvas"):
            with gr.Row(elem_id="canvas-toolbar"):
                gr.Markdown("Canvas", elem_id="canvas-info")

            result_image = gr.Image(
                elem_id="main-image", interactive=False, show_label=False
            )

        # --- Right Sidebar ---
        with gr.Column(elem_id="right-sidebar", min_width=320):
            gr.HTML(controls_header_html)

            # Prompt Section
            prompt = gr.Textbox(
                elem_id="prompt-input",
                lines=4,
                placeholder="Describe your imagination...",
                label="Prompt",
                show_label=True,
            )
            run_button = gr.Button("Generate Image", elem_id="generate-btn")

            # Settings
            input_images = gr.Gallery(
                label="Input Image(s)", type="pil", columns=3, rows=1
            )

            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,
            )

            guidance_scale = gr.Slider(
                label="Guidance Scale", minimum=0.0, maximum=10.0, step=0.1, value=4
            )
            num_inference_steps = gr.Slider(
                label="Inference Steps", minimum=1, maximum=100, step=1, value=30
            )
            seed = gr.Slider(label="Seed", minimum=0, maximum=MAX_SEED, step=1, value=0)
            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)

    # Wiring
    gr.on(
        triggers=[run_button.click, prompt.submit],
        fn=infer,
        inputs=[
            prompt,
            input_images,
            seed,
            randomize_seed,
            width,
            height,
            num_inference_steps,
            guidance_scale,
        ],
        outputs=[result_image, seed],
    )

demo.launch(css=css)