app_v2.py

import gradio as gr
import numpy as np
import random
import spaces
import torch
from diffusers import DiffusionPipeline, AutoencoderTiny, AutoencoderKL
from transformers import CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5TokenizerFast
from live_preview_helpers import calculate_shift, retrieve_timesteps, flux_pipe_call_that_returns_an_iterable_of_images
import requests
import base64
import os
from PIL import Image
from io import BytesIO
from gradio_imageslider import ImageSlider  # Assicurati di avere questa libreria installata
from loadimg import load_img  # Assicurati che questa funzione sia disponibile
from dotenv import load_dotenv

# Carica le variabili di ambiente dal file .env
load_dotenv()

dtype = torch.bfloat16
device = "cuda" if torch.cuda.is_available() else "cpu"
taef1 = AutoencoderTiny.from_pretrained("madebyollin/taef1", torch_dtype=dtype).to(device)
good_vae = AutoencoderKL.from_pretrained("black-forest-labs/FLUX.1-dev", subfolder="vae", torch_dtype=dtype).to(device)
pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=dtype, vae=taef1).to(device)
torch.cuda.empty_cache()

MAX_SEED = np.iinfo(np.int32).max
MAX_IMAGE_SIZE = 2048

pipe.flux_pipe_call_that_returns_an_iterable_of_images = flux_pipe_call_that_returns_an_iterable_of_images.__get__(pipe)

output_folder = 'output_images'
if not os.path.exists(output_folder):
    os.makedirs(output_folder)


def numpy_to_pil(image):
    """Convert a numpy array to a PIL Image."""
    if image.dtype == np.uint8:  # Most common case
        mode = "RGB"
    else:
        mode = "F"  # Floating point
    return Image.fromarray(image.astype('uint8'), mode)


def process_image(image):
    image = numpy_to_pil(image)  # Convert numpy array to PIL Image
    buffered = BytesIO()
    image.save(buffered, format="PNG")
    img_str = base64.b64encode(buffered.getvalue()).decode('utf-8')
    response = requests.post(
        os.getenv('BACKEND_URL'),
        files={"file": ("image.png", base64.b64decode(img_str), "image/png")}
    )
    result = response.json()
    processed_image_b64 = result["processed_image"]
    processed_image = Image.open(BytesIO(base64.b64decode(processed_image_b64)))
    image_path = os.path.join(output_folder, "no_bg_image.png")
    processed_image.save(image_path)
    return (processed_image, image), image_path


@spaces.GPU(duration=75)
def infer(prompt, seed=42, randomize_seed=False, width=1024, height=1024, guidance_scale=3.5, num_inference_steps=28,
          progress=gr.Progress(track_tqdm=True)):
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    generator = torch.Generator().manual_seed(seed)
    for img in pipe.flux_pipe_call_that_returns_an_iterable_of_images(
            prompt=prompt,
            guidance_scale=guidance_scale,
            num_inference_steps=num_inference_steps,
            width=width,
            height=height,
            generator=generator,
            output_type="pil",
            good_vae=good_vae,
    ):
        img_np = np.array(img)
        processed_images, image_path = process_image(img_np)
        yield processed_images[0], seed, processed_images[1], image_path


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(f"""# FLUX.1 [dev]  
12B param rectified flow transformer guidance-distilled from [FLUX.1 [pro]](https://blackforestlabs.ai/)  [[non-commercial license](https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md)] [[blog](https://blackforestlabs.ai/announcing-black-forest-labs/)] [[model](https://huggingface.co/black-forest-labs/FLUX.1-dev)]  
        """)
        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="Generated Image", show_label=False)
        output_slider = ImageSlider(label="Processed Photo", type="pil")
        output_file = gr.File(label="Output PNG file")
        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():
                guidance_scale = gr.Slider(
                    label="Guidance Scale",
                    minimum=1,
                    maximum=15,
                    step=0.1,
                    value=3.5,
                )
                num_inference_steps = gr.Slider(
                    label="Number of inference steps",
                    minimum=1,
                    maximum=50,
                    step=1,
                    value=28,
                )
        gr.Examples(
            examples=examples,
            fn=infer,
            inputs=[prompt],
            outputs=[result, seed, output_slider, output_file],
            cache_examples="lazy"
        )
        gr.on(
            triggers=[run_button.click, prompt.submit],
            fn=infer,
            inputs=[prompt, seed, randomize_seed, width, height, guidance_scale, num_inference_steps],
            outputs=[result, seed, output_slider, output_file]
        )
demo.launch()