Skip to content

Neural networks from scratch using numpy and pytorch

Notifications You must be signed in to change notification settings

e-cal/nn-from-scratch

Repository files navigation

Running Locally

  1. Clone the repository: git clone git@github.com:e-cal/nn-from-scratch.git
  2. (Optional) Create a virtual environment
    venv "$VIRTUALENV_HOME/ENVNAME"
    source "$VIRTUALENV_HOME/ENVNAME/bin/activate"
    See the next section for creating a notebook kernel in the virtual environment.
  3. Install the requirements: pip install -r requirements.txt
  4. Launch the notebook server: jupyter notebook
  5. Select the correct kernel to use for the notebook: Kernel > Change kernel > [select the proper kernel]

Setting up a Notebook Kernel in a Virtual Environment

If you install the required packages to a virtual environment, you can add the environment to Jupyter with this command: python3 -m ipykernel install --name=ENVNAME (make sure to replace ENVNAME. You may also need to add the --user flag if you get permission errors.)

Should print something like:

Installed kernelspec ENVNAME in /home/USER/.local/share/jupyter/kernels/ENVNAME

If there is any difficulty getting a kernel set up with the required packages installed, you can install the requirements outside of a virtual environment and use the default kernel.

About

Neural networks from scratch using numpy and pytorch

Resources

Stars

Watchers

Forks