Create flowcharts of neural networks and convert the flowcharts into PyTorch for inferencing/training.
Clone this repository
git clone https://github.com/cwc020730/NeuralNetworkBuilder.git
Install Node.js and npm Install Python and pip
Make sure the directory of npm, pip and python executables are in the system environment variable PATH
Go to the frontend directory:
cd frontend
Install relevant packages:
npm install
In a terminal, start the frontend:
npm start
npm start should start the app locally: http://localhost:3000.
In another terminal, start the backend:
pip install -r backend/requirements.txt
pip install -r backend/requirements_pytorch.txt
python -m backend.backend
The backend server should start locally: http://localhost:5000.
You may use File -> Import on http://localhost:3000 and select examples/simple_net_for_mnist.json to load the same neural network.
Click the Run button to start training the neural network.
Currently all the training status will be shown on the terminal where you start python -m backend.backend.
After each training epoch, the data (loss and accuracy) in CELOSS unit will be refreshed and a graph showing the changes will appear.
Welcome any contributions, feedbacks, discussions, or inputs. Submit issue or pull requests to contribute.
- Discord Username: hh4939
- Email: xxzhangjinghexx@gmail.com
- QQ: 1766797600