app.py

# PyTorch 2.8 (temporary hack)
import os
os.system('pip install --upgrade --pre --extra-index-url https://download.pytorch.org/whl/nightly/cu126 "torch<2.9" spaces')

import gradio as gr
import numpy as np
import random
import torch
import spaces

from PIL import Image
from diffusers import QwenImageEditPipeline
from diffusers import DiffusionPipeline, FlowMatchEulerDiscreteScheduler
import torch 
import math

from optimization import optimize_pipeline_


# --- Model Loading ---
dtype = torch.bfloat16
device = "cuda" if torch.cuda.is_available() else "cpu"

# Load the model pipeline

# scheduler config needed for the LoRA
# From https://github.com/ModelTC/Qwen-Image-Lightning/blob/342260e8f5468d2f24d084ce04f55e101007118b/generate_with_diffusers.py#L82C9-L97C10
scheduler_config = {
    "base_image_seq_len": 256,
    "base_shift": math.log(3),  # We use shift=3 in distillation
    "invert_sigmas": False,
    "max_image_seq_len": 8192,
    "max_shift": math.log(3),  # We use shift=3 in distillation
    "num_train_timesteps": 1000,
    "shift": 1.0,
    "shift_terminal": None,  # set shift_terminal to None
    "stochastic_sampling": False,
    "time_shift_type": "exponential",
    "use_beta_sigmas": False,
    "use_dynamic_shifting": True,
    "use_exponential_sigmas": False,
    "use_karras_sigmas": False,
}
scheduler = FlowMatchEulerDiscreteScheduler.from_config(scheduler_config)
pipe = QwenImageEditPipeline.from_pretrained("Qwen/Qwen-Image-Edit", scheduler=scheduler, torch_dtype=dtype).to(device)

# lora loading
pipe.load_lora_weights(
    "lightx2v/Qwen-Image-Lightning", weight_name="Qwen-Image-Lightning-8steps-V1.0.safetensors", adapter_name="lightx2v"
)
pipe.set_adapters(["lightx2v"], adapter_weights=[1.])
pipe.fuse_lora(adapter_names=["lightx2v"], lora_scale=1., components=["transformer"])
pipe.unload_lora_weights()

optimize_pipeline_(pipe, image=Image.new("RGB", (1024, 1024)), prompt='prompt')

# --- UI Constants and Helpers ---
MAX_SEED = np.iinfo(np.int32).max

# --- Main Inference Function (with hardcoded negative prompt) ---
@spaces.GPU(duration=120)
def infer(
    image,
    prompt,
    seed=42,
    randomize_seed=False,
    guidance_scale=4.0,
    true_guidance_scale=1.0,
    num_inference_steps=8,
    progress=gr.Progress(track_tqdm=True),
):
    """
    Generates an image using the local Qwen-Image diffusers pipeline.
    """
    # Hardcode the negative prompt as requested
    negative_prompt = "text, watermark, copyright, blurry, low resolution"
    
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)

    # Set up the generator for reproducibility
    generator = torch.Generator(device=device).manual_seed(seed)
    
    print(f"Calling pipeline with prompt: '{prompt}'")
    print(f"Negative Prompt: '{negative_prompt}'")
    print(f"Seed: {seed}, Steps: {num_inference_steps}, Guidance: {guidance_scale}")

    # Generate the image
    image = pipe(
        image,
        prompt=prompt,
        negative_prompt=negative_prompt,
        num_inference_steps=num_inference_steps,
        generator=generator,
        true_cfg_scale=true_guidance_scale,
        guidance_scale=guidance_scale  
    ).images[0]

    return image, seed

# --- Examples and UI Layout ---
examples = []

css = """
#col-container {
    margin: 0 auto;
    max-width: 1024px;
}
#edit_text{margin-top: -62px !important}
"""

with gr.Blocks(css=css) as demo:
    with gr.Column(elem_id="col-container"):
        gr.HTML('<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/qwen_image_logo.png" alt="Qwen-Image Logo" width="400" style="display: block; margin: 0 auto;">')
        gr.HTML('<h1 style="text-align: center;margin-left: 80px;color: #5b47d1;font-style: italic;">Edit</h1>', elem_id="edit_text")
        gr.Markdown("[Learn more](https://github.com/QwenLM/Qwen-Image) about the Qwen-Image series. Try on [Qwen Chat](https://chat.qwen.ai/), or [download model](https://huggingface.co/Qwen/Qwen-Image-Edit) to run locally with ComfyUI or diffusers.")
        with gr.Row():
            with gr.Column():
                input_image = gr.Image(label="Input Image", show_label=False, type="pil")
                prompt = gr.Text(
                    label="Prompt",
                    show_label=False,
                    placeholder="describe the edit instruction",
                    container=False,
                )
                run_button = gr.Button("Edit!", variant="primary")

            result = gr.Image(label="Result", show_label=False, type="pil")

        with gr.Accordion("Advanced Settings", open=False):
            # Negative prompt UI element is removed here

            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():
                guidance_scale = gr.Slider(
                    label="Distilled guidance scale",
                    minimum=0.0,
                    maximum=10.0,
                    step=0.1,
                    value=1.0,
                )

                true_guidance_scale = gr.Slider(
                    label="True guidance scale",
                    minimum=1.0,
                    maximum=10.0,
                    step=0.1,
                    value=1.0
                )

                num_inference_steps = gr.Slider(
                    label="Number of inference steps",
                    minimum=1,
                    maximum=50,
                    step=1,
                    value=8,
                )

        # gr.Examples(examples=examples, inputs=[prompt], outputs=[result, seed], fn=infer, cache_examples=False)

    gr.on(
        triggers=[run_button.click, prompt.submit],
        fn=infer,
        inputs=[
            input_image,
            prompt,
            # negative_prompt is no longer an input from the UI
            seed,
            randomize_seed,
            guidance_scale,
            true_guidance_scale,
            num_inference_steps,
        ],
        outputs=[result, seed],
    )

if __name__ == "__main__":
    demo.launch()