handwritten-to-LaTeX-demo

User can draw and brrrr

Config Dataset

Windows:

cd .\Backend\learner\datasets\
kaggle datasets download xainano/handwrittenmathsymbols
tar -xvzf .\handwrittenmathsymbols.zip
tar -xvzf .\data.rar

Linux:

cd ./Backend/learner/datasets
kaggle datasets download xainano/handwrittenmathsymbols
unzip ./handwrittenmathsymbols.zip
rar x data.rar

• Very important make sure the extracted_images has 82 categories, and that will only happen after you extract data.rar .

Training Instructions:

• After installing all the requiremnets by

pip install -r requirements.txt

make sure you are in the Backend directory.

• And run:

python trainmodel.py

Electron build guide

User can build their own electron application to run frontend and backend together

Install deps

npm install

Run electron app

Before running the electron app, you need to build the frontend react app. To do that, follow these:

cd Frontend
npm run build

Now electron is ready to run:

cd .. # don't do that if you are not in frontend's parent directory
npm start
# or
npx electron .

Build binary for your platform

As of now, build binary is facing some issue in starting, but might work on diff platforms

# for linux
npx electron-builder --linux deb tar.xz

# build for Windows ia32
npx electron-builder --win --ia32

# build for macOS, Windows and Linux
npx electron-builder -mwl