COPAINT

TLDR: From generated Image -> Copaint PDF

git clone https://github.com/ThibaultGROUEIX/copaint.git
git lfs pull #(install git LFS)

Usage and Install

# Install dependencies
poetry install --with ui

# Run Gradio App on local server
poetry run copaint-app

Alternative Usage and Install (CLI)

# Install dependencies
pip install torch torchvision reportlab PyPDF2 Pillow argparse gradio_pdf

# Run
python copaint.py --input_image data/bear.jpg --copaint_logo copaint/static/logo_copaint.png --outputfolder output

# Using the CLI
pip install --upgrade pip
pip install -e .

# Case 1 : generate all the presets
copaint --input_image data/bear.jpg --copaint_logo copaint/static/logo_copaint.png --outputfolder output --use_presets True

# Case 2 : provide a number of participants, the program will generate the best grid
copaint --input_image data/bear.jpg --copaint_logo copaint/static/logo_copaint.png --outputfolder output --nparticipants 45 

# Case 3 : provide the number of cells in the grid
copaint --input_image data/bear.jpg --copaint_logo copaint/static/logo_copaint.png --outputfolder output --h_cells 3 --w_cells 4

Build and deploy

# Build
poetry build
# Deploy
poetry publish

Time bottleneck

-- the biggest time bottleneck is writing to disk the back image and loading for the PDF creation.

Deploy to huggingface.

git remote add hf_en https://huggingface.co/spaces/groueix/copaint git remote add hf_fr https://huggingface.co/spaces/groueix/copaint_fr

to check remotes : git remote -v

change language in app.py

git push hf_en main git push hf_fr main