app.py

# # import logging
# # import random
# # import warnings
# # import os
# # import gradio as gr
# # import numpy as np
# # import spaces
# # import torch
# # from diffusers import FluxControlNetModel
# # from diffusers.pipelines import FluxControlNetPipeline
# # from gradio_imageslider import ImageSlider
# # from PIL import Image
# # from huggingface_hub import snapshot_download

# # css = """
# # #col-container {
# #     margin: 0 auto;
# #     max-width: 512px;
# # }
# # """

# # if torch.cuda.is_available():
# #     power_device = "GPU"
# #     device = "cuda"
# # else:
# #     power_device = "CPU"
# #     device = "cpu"


# # huggingface_token = os.getenv("HUGGINFACE_TOKEN")

# # model_path = snapshot_download(
# #     repo_id="black-forest-labs/FLUX.1-dev", 
# #     repo_type="model", 
# #     ignore_patterns=["*.md", "*..gitattributes"],
# #     local_dir="FLUX.1-dev",
# #     token=huggingface_token, # type a new token-id.
# # )


# # # Load pipeline
# # controlnet = FluxControlNetModel.from_pretrained(
# #     "jasperai/Flux.1-dev-Controlnet-Upscaler", torch_dtype=torch.bfloat16
# # ).to(device)
# # pipe = FluxControlNetPipeline.from_pretrained(
# #     model_path, controlnet=controlnet, torch_dtype=torch.bfloat16
# # )
# # pipe.to(device)

# # MAX_SEED = 1000000
# # MAX_PIXEL_BUDGET = 1024 * 1024


# # def process_input(input_image, upscale_factor, **kwargs):
# #     w, h = input_image.size
# #     w_original, h_original = w, h
# #     aspect_ratio = w / h

# #     was_resized = False

# #     if w * h * upscale_factor**2 > MAX_PIXEL_BUDGET:
# #         warnings.warn(
# #             f"Requested output image is too large ({w * upscale_factor}x{h * upscale_factor}). Resizing to ({int(aspect_ratio * MAX_PIXEL_BUDGET ** 0.5 // upscale_factor), int(MAX_PIXEL_BUDGET ** 0.5 // aspect_ratio // upscale_factor)}) pixels."
# #         )
# #         gr.Info(
# #             f"Requested output image is too large ({w * upscale_factor}x{h * upscale_factor}). Resizing input to ({int(aspect_ratio * MAX_PIXEL_BUDGET ** 0.5 // upscale_factor), int(MAX_PIXEL_BUDGET ** 0.5 // aspect_ratio // upscale_factor)}) pixels budget."
# #         )
# #         input_image = input_image.resize(
# #             (
# #                 int(aspect_ratio * MAX_PIXEL_BUDGET**0.5 // upscale_factor),
# #                 int(MAX_PIXEL_BUDGET**0.5 // aspect_ratio // upscale_factor),
# #             )
# #         )
# #         was_resized = True

# #     # resize to multiple of 8
# #     w, h = input_image.size
# #     w = w - w % 8
# #     h = h - h % 8

# #     return input_image.resize((w, h)), w_original, h_original, was_resized


# # @spaces.GPU#(duration=42)
# # def infer(
# #     seed,
# #     randomize_seed,
# #     input_image,
# #     num_inference_steps,
# #     upscale_factor,
# #     controlnet_conditioning_scale,
# #     progress=gr.Progress(track_tqdm=True),
# # ):
# #     if randomize_seed:
# #         seed = random.randint(0, MAX_SEED)
# #     true_input_image = input_image
# #     input_image, w_original, h_original, was_resized = process_input(
# #         input_image, upscale_factor
# #     )

# #     # rescale with upscale factor
# #     w, h = input_image.size
# #     control_image = input_image.resize((w * upscale_factor, h * upscale_factor))

# #     generator = torch.Generator().manual_seed(seed)

# #     gr.Info("Upscaling image...")
# #     image = pipe(
# #         prompt="",
# #         control_image=control_image,
# #         controlnet_conditioning_scale=controlnet_conditioning_scale,
# #         num_inference_steps=num_inference_steps,
# #         guidance_scale=3.5,
# #         height=control_image.size[1],
# #         width=control_image.size[0],
# #         generator=generator,
# #     ).images[0]

# #     if was_resized:
# #         gr.Info(
# #             f"Resizing output image to targeted {w_original * upscale_factor}x{h_original * upscale_factor} size."
# #         )

# #     # resize to target desired size
# #     image = image.resize((w_original * upscale_factor, h_original * upscale_factor))
# #     image.save("output.jpg")
# #     # convert to numpy
# #     return [true_input_image, image, seed]


# # with gr.Blocks(css=css) as demo:
# #     # with gr.Column(elem_id="col-container"):
# #     gr.Markdown(
# #         f"""
# #     # ⚡ Flux.1-dev Upscaler ControlNet ⚡
# #     This is an interactive demo of [Flux.1-dev Upscaler ControlNet](https://huggingface.co/jasperai/Flux.1-dev-Controlnet-Upscaler) taking as input a low resolution image to generate a high resolution image.
# #     Currently running on {power_device}.

# #     *Note*: Even though the model can handle higher resolution images, due to GPU memory constraints, this demo was limited to a generated output not exceeding a pixel budget of 1024x1024. If the requested size exceeds that limit, the input will be first resized keeping the aspect ratio such that the output of the controlNet model does not exceed the allocated pixel budget. The output is then resized to the targeted shape using a simple resizing. This may explain some artifacts for high resolution input. To adress this, run the demo locally or consider implementing a tiling strategy. Happy upscaling! 🚀
# #     """
# #     )

# #     with gr.Row():
# #         run_button = gr.Button(value="Run")

# #     with gr.Row():
# #         with gr.Column(scale=4):
# #             input_im = gr.Image(label="Input Image", type="pil")
# #         with gr.Column(scale=1):
# #             num_inference_steps = gr.Slider(
# #                 label="Number of Inference Steps",
# #                 minimum=8,
# #                 maximum=50,
# #                 step=1,
# #                 value=28,
# #             )
# #             upscale_factor = gr.Slider(
# #                 label="Upscale Factor",
# #                 minimum=1,
# #                 maximum=4,
# #                 step=1,
# #                 value=4,
# #             )
# #             controlnet_conditioning_scale = gr.Slider(
# #                 label="Controlnet Conditioning Scale",
# #                 minimum=0.1,
# #                 maximum=1.5,
# #                 step=0.1,
# #                 value=0.6,
# #             )
# #             seed = gr.Slider(
# #                 label="Seed",
# #                 minimum=0,
# #                 maximum=MAX_SEED,
# #                 step=1,
# #                 value=42,
# #             )

# #             randomize_seed = gr.Checkbox(label="Randomize seed", value=True)

# #     with gr.Row():
# #         result = ImageSlider(label="Input / Output", type="pil", interactive=True)

# #     examples = gr.Examples(
# #         examples=[
# #         #    [42, False, "examples/image_1.jpg", 28, 4, 0.6],
# #             [42, False, "examples/image_2.jpg", 28, 4, 0.6],
# #         #    [42, False, "examples/image_3.jpg", 28, 4, 0.6],
# #             [42, False, "examples/image_4.jpg", 28, 4, 0.6],
# #         #    [42, False, "examples/image_5.jpg", 28, 4, 0.6],
# #         #    [42, False, "examples/image_6.jpg", 28, 4, 0.6],
# #         ],
# #         inputs=[
# #             seed,
# #             randomize_seed,
# #             input_im,
# #             num_inference_steps,
# #             upscale_factor,
# #             controlnet_conditioning_scale,
# #         ],
# #         fn=infer,
# #         outputs=result,
# #         cache_examples="lazy",
# #     )

# #     # examples = gr.Examples(
# #     #     examples=[
# #     #         #[42, False, "examples/image_1.jpg", 28, 4, 0.6],
# #     #         [42, False, "examples/image_2.jpg", 28, 4, 0.6],
# #     #         #[42, False, "examples/image_3.jpg", 28, 4, 0.6],
# #     #         #[42, False, "examples/image_4.jpg", 28, 4, 0.6],
# #     #         [42, False, "examples/image_5.jpg", 28, 4, 0.6],
# #     #         [42, False, "examples/image_6.jpg", 28, 4, 0.6],
# #     #         [42, False, "examples/image_7.jpg", 28, 4, 0.6],
# #     #     ],
# #     #     inputs=[
# #     #         seed,
# #     #         randomize_seed,
# #     #         input_im,
# #     #         num_inference_steps,
# #     #         upscale_factor,
# #     #         controlnet_conditioning_scale,
# #     #     ],
# #     # )

# #     gr.Markdown("**Disclaimer:**")
# #     gr.Markdown(
# #         "This demo is only for research purpose. Jasper cannot be held responsible for the generation of NSFW (Not Safe For Work) content through the use of this demo. Users are solely responsible for any content they create, and it is their obligation to ensure that it adheres to appropriate and ethical standards. Jasper provides the tools, but the responsibility for their use lies with the individual user."
# #     )
# #     gr.on(
# #         [run_button.click],
# #         fn=infer,
# #         inputs=[
# #             seed,
# #             randomize_seed,
# #             input_im,
# #             num_inference_steps,
# #             upscale_factor,
# #             controlnet_conditioning_scale,
# #         ],
# #         outputs=result,
# #         show_api=False,
# #         # show_progress="minimal",
# #     )

# # demo.queue().launch(share=False, show_api=False)






# import logging
# import random
# import warnings
# import os,shutil,subprocess
# import torch
# import numpy as np
# from diffusers import FluxControlNetModel
# from diffusers.pipelines import FluxControlNetPipeline
# from PIL import Image
# from huggingface_hub import snapshot_download,login
# import io
# import base64
# from flask import Flask, request, jsonify
# from concurrent.futures import ThreadPoolExecutor
# from flask_cors import CORS
# from tqdm import tqdm

# app = Flask(__name__)
# CORS(app)

# # Function to check disk usage
# def check_disk_space():
#     result = subprocess.run(['df', '-h'], capture_output=True, text=True)
#     print(result.stdout)

# # Function to clear Hugging Face cache
# def clear_huggingface_cache():
#     cache_dir = os.path.expanduser('~/.cache/huggingface')
#     if os.path.exists(cache_dir):
#         shutil.rmtree(cache_dir)  # Removes the entire cache directory
#         print(f"Cleared Hugging Face cache at: {cache_dir}")
#     else:
#         print("No Hugging Face cache found.")

# # Check disk space
# check_disk_space()

# # Clear Hugging Face cache
# clear_huggingface_cache()

# # Add config to store base64 images
# app.config['image_outputs'] = {}

# # ThreadPoolExecutor for managing image processing threads
# executor = ThreadPoolExecutor()

# # Determine the device (GPU or CPU)
# if torch.cuda.is_available():
#     device = "cuda"
# else:
#     device = "cpu"

# # Load model from Huggingface Hub
# huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
# if huggingface_token:
#     login(token=huggingface_token)
# else:
#     print("Hugging Face token not found in environment variables.")
# print(huggingface_token)
# with tqdm(total=100, desc="Downloading model", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
#  model_path = snapshot_download(
#     repo_id="black-forest-labs/FLUX.1-dev", 
#     repo_type="model", 
#     ignore_patterns=["*.md", "*..gitattributes"],
#     local_dir="FLUX.1-dev",
#     token=huggingface_token)

# # Load pipeline
# print('controlnet enters')
# with tqdm(total=100, desc="Downloading model", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
#  controlnet = FluxControlNetModel.from_pretrained(
#     "jasperai/Flux.1-dev-Controlnet-Upscaler", torch_dtype=torch.bfloat16
# ).to(device)
# print('controlnet exits')
# print('pipe enters')
# with tqdm(total=100, desc="Downloading model", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
#  pipe = FluxControlNetPipeline.from_pretrained(
#     model_path, controlnet=controlnet, torch_dtype=torch.bfloat16
# ).to(device)
# # pipe.to(device)
# print('pipe exits')

# MAX_SEED = 1000000
# MAX_PIXEL_BUDGET = 1024 * 1024

# def process_input(input_image, upscale_factor):
#     w, h = input_image.size
#     aspect_ratio = w / h
#     was_resized = False

#     # Resize if input size exceeds the maximum pixel budget
#     if w * h * upscale_factor**2 > MAX_PIXEL_BUDGET:
#         warnings.warn(f"Requested output image is too large. Resizing to fit within pixel budget.")
#         input_image = input_image.resize(
#             (
#                 int(aspect_ratio * MAX_PIXEL_BUDGET**0.5 // upscale_factor),
#                 int(MAX_PIXEL_BUDGET**0.5 // aspect_ratio // upscale_factor),
#             )
#         )
#         was_resized = True

#     # Adjust dimensions to be a multiple of 8
#     w, h = input_image.size
#     w = w - w % 8
#     h = h - h % 8

#     return input_image.resize((w, h)), was_resized

# def run_inference(process_id, input_image, upscale_factor, seed, num_inference_steps, controlnet_conditioning_scale):
#     input_image, was_resized = process_input(input_image, upscale_factor)

#     # Rescale image for ControlNet processing
#     w, h = input_image.size
#     control_image = input_image.resize((w * upscale_factor, h * upscale_factor))

#     # Set the random generator for inference
#     generator = torch.Generator().manual_seed(seed)

#     # Perform inference using the pipeline
#     image = pipe(
#         prompt="",
#         control_image=control_image,
#         controlnet_conditioning_scale=controlnet_conditioning_scale,
#         num_inference_steps=num_inference_steps,
#         guidance_scale=3.5,
#         height=control_image.size[1],
#         width=control_image.size[0],
#         generator=generator,
#     ).images[0]

#     # Resize output image back to the original dimensions if needed
#     if was_resized:
#         original_size = (input_image.width * upscale_factor, input_image.height * upscale_factor)
#         image = image.resize(original_size)

#     # Convert the output image to base64
#     buffered = io.BytesIO()
#     image.save(buffered, format="JPEG")
#     image_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")

#     # Store the result in the shared dictionary
#     app.config['image_outputs'][process_id] = image_base64

# @app.route('/infer', methods=['POST'])
# def infer():
#     data = request.json
#     seed = data.get("seed", 42)
#     randomize_seed = data.get("randomize_seed", True)
#     num_inference_steps = data.get("num_inference_steps", 28)
#     upscale_factor = data.get("upscale_factor", 4)
#     controlnet_conditioning_scale = data.get("controlnet_conditioning_scale", 0.6)

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

#     # Load and process the input image
#     input_image_data = base64.b64decode(data['input_image'])
#     input_image = Image.open(io.BytesIO(input_image_data))

#     # Create a unique process ID for this request
#     process_id = str(random.randint(1000, 9999))

#     # Set the status to 'in_progress'
#     app.config['image_outputs'][process_id] = None

#     # Run the inference in a separate thread
#     executor.submit(run_inference, process_id, input_image, upscale_factor, seed, num_inference_steps, controlnet_conditioning_scale)

#     # Return the process ID
#     return jsonify({
#         "process_id": process_id,
#         "message": "Processing started"
#     })

# # Modify status endpoint to receive process_id in request body
# @app.route('/status', methods=['POST'])
# def status():
#     data = request.json
#     process_id = data.get('process_id')

#     # Check if process_id was provided
#     if not process_id:
#         return jsonify({
#             "status": "error",
#             "message": "Process ID is required"
#         }), 400

#     # Check if the process_id exists in the dictionary
#     if process_id not in app.config['image_outputs']:
#         return jsonify({
#             "status": "error",
#             "message": "Invalid process ID"
#         }), 404

#     # Check the status of the image processing
#     image_base64 = app.config['image_outputs'][process_id]
#     if image_base64 is None:
#         return jsonify({
#             "status": "in_progress"
#         })
#     else:
#         return jsonify({
#             "status": "completed",
#             "output_image": image_base64
#         })

# if __name__ == '__main__':
#     app.run(debug=True)





import logging
import random
import warnings
import os
import shutil
import subprocess
import torch
import numpy as np
from diffusers import FluxControlNetModel
from diffusers.pipelines import FluxControlNetPipeline
from PIL import Image
from huggingface_hub import snapshot_download, login
import io
import base64
from flask import Flask, request, jsonify
from concurrent.futures import ThreadPoolExecutor
from flask_cors import CORS
from tqdm import tqdm

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

app = Flask(__name__)
CORS(app)

# Function to check disk usage
def check_disk_space():
    result = subprocess.run(['df', '-h'], capture_output=True, text=True)
    logger.info("Disk space usage:\n%s", result.stdout)

# Function to clear Hugging Face cache
def clear_huggingface_cache():
    cache_dir = os.path.expanduser('~/.cache/huggingface')
    if os.path.exists(cache_dir):
        shutil.rmtree(cache_dir)  # Removes the entire cache directory
        logger.info("Cleared Hugging Face cache at: %s", cache_dir)
    else:
        logger.info("No Hugging Face cache found.")

# Check disk space
check_disk_space()

# Clear Hugging Face cache
clear_huggingface_cache()

# Add config to store base64 images
app.config['image_outputs'] = {}

# ThreadPoolExecutor for managing image processing threads
executor = ThreadPoolExecutor()

# Determine the device (GPU or CPU)
if torch.cuda.is_available():
    device = "cuda"
    logger.info("CUDA is available. Using GPU.")
else:
    device = "cpu"
    logger.info("CUDA is not available. Using CPU.")

# Load model from Huggingface Hub
huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
if huggingface_token:
    login(token=huggingface_token)
    logger.info("Hugging Face token found and logged in.")
else:
    logger.warning("Hugging Face token not found in environment variables.")

logger.info("Hugging Face token: %s", huggingface_token)

# Download model using snapshot_download
with tqdm(total=100, desc="Downloading model", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
    model_path = snapshot_download(
        repo_id="black-forest-labs/FLUX.1-dev", 
        repo_type="model", 
        ignore_patterns=["*.md", "*..gitattributes"],
        local_dir="FLUX.1-dev",
        token=huggingface_token)
    logger.info("Model downloaded to: %s", model_path)

# Load pipeline
logger.info('Loading ControlNet model.')
with tqdm(total=100, desc="Downloading ControlNet model", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
    controlnet = FluxControlNetModel.from_pretrained(
        "jasperai/Flux.1-dev-Controlnet-Upscaler", torch_dtype=torch.bfloat16
    ).to(device)
logger.info("ControlNet model loaded successfully.")

logger.info('Loading pipeline.')
with tqdm(total=100, desc="Downloading pipeline", bar_format="{l_bar}{bar}| {n_fmt}/{total_fmt}") as pbar:
    pipe = FluxControlNetPipeline.from_pretrained(
        model_path, controlnet=controlnet, torch_dtype=torch.bfloat16
    ).to(device)
logger.info("Pipeline loaded successfully.")

MAX_SEED = 1000000
MAX_PIXEL_BUDGET = 1024 * 1024

def process_input(input_image, upscale_factor):
    w, h = input_image.size
    aspect_ratio = w / h
    was_resized = False

    # Resize if input size exceeds the maximum pixel budget
    if w * h * upscale_factor**2 > MAX_PIXEL_BUDGET:
        warnings.warn(f"Requested output image is too large. Resizing to fit within pixel budget.")
        input_image = input_image.resize(
            (
                int(aspect_ratio * MAX_PIXEL_BUDGET**0.5 // upscale_factor),
                int(MAX_PIXEL_BUDGET**0.5 // aspect_ratio // upscale_factor),
            )
        )
        was_resized = True

    # Adjust dimensions to be a multiple of 8
    w, h = input_image.size
    w = w - w % 8
    h = h - h % 8

    return input_image.resize((w, h)), was_resized

def run_inference(process_id, input_image, upscale_factor, seed, num_inference_steps, controlnet_conditioning_scale):
    logger.info("Processing inference for process_id: %s", process_id)
    input_image, was_resized = process_input(input_image, upscale_factor)

    # Rescale image for ControlNet processing
    w, h = input_image.size
    control_image = input_image.resize((w * upscale_factor, h * upscale_factor))

    # Set the random generator for inference
    generator = torch.Generator().manual_seed(seed)

    # Perform inference using the pipeline
    logger.info("Running pipeline for process_id: %s", process_id)
    image = pipe(
        prompt="",
        control_image=control_image,
        controlnet_conditioning_scale=controlnet_conditioning_scale,
        num_inference_steps=num_inference_steps,
        guidance_scale=3.5,
        height=control_image.size[1],
        width=control_image.size[0],
        generator=generator,
    ).images[0]

    # Resize output image back to the original dimensions if needed
    if was_resized:
        original_size = (input_image.width * upscale_factor, input_image.height * upscale_factor)
        image = image.resize(original_size)

    # Convert the output image to base64
    buffered = io.BytesIO()
    image.save(buffered, format="JPEG")
    image_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")

    # Store the result in the shared dictionary
    app.config['image_outputs'][process_id] = image_base64
    logger.info("Inference completed for process_id: %s", process_id)

@app.route('/infer', methods=['POST'])
def infer():
    data = request.json
    seed = data.get("seed", 42)
    randomize_seed = data.get("randomize_seed", True)
    num_inference_steps = data.get("num_inference_steps", 28)
    upscale_factor = data.get("upscale_factor", 4)
    controlnet_conditioning_scale = data.get("controlnet_conditioning_scale", 0.6)

    # Randomize seed if specified
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
        logger.info("Seed randomized to: %d", seed)

    # Load and process the input image
    input_image_data = base64.b64decode(data['input_image'])
    input_image = Image.open(io.BytesIO(input_image_data))

    # Create a unique process ID for this request
    process_id = str(random.randint(1000, 9999))
    logger.info("Process started with process_id: %s", process_id)

    # Set the status to 'in_progress'
    app.config['image_outputs'][process_id] = None

    # Run the inference in a separate thread
    executor.submit(run_inference, process_id, input_image, upscale_factor, seed, num_inference_steps, controlnet_conditioning_scale)

    # Return the process ID
    return jsonify({
        "process_id": process_id,
        "message": "Processing started"
    })

# Modify status endpoint to receive process_id in request body
@app.route('/status', methods=['POST'])
def status():
    data = request.json
    process_id = data.get('process_id')

    # Check if process_id was provided
    if not process_id:
        logger.error("Process ID not provided in request.")
        return jsonify({
            "status": "error",
            "message": "Process ID is required"
        }), 400

    # Check if the process_id exists in the dictionary
    if process_id not in app.config['image_outputs']:
        logger.error("Invalid process ID: %s", process_id)
        return jsonify({
            "status": "error",
            "message": "Invalid process ID"
        }), 404

    # Check the status of the image processing
    image_base64 = app.config['image_outputs'][process_id]
    if image_base64 is None:
        logger.info("Process ID %s is still in progress.", process_id)
        return jsonify({
            "status": "in_progress"
        })
    else:
        logger.info("Process ID %s completed successfully.", process_id)
        return jsonify({
            "status": "completed",
            "output_image": image_base64
        })

if __name__ == '__main__':
    app.run(debug=True)