An investigation of the GLEAMS embedder using WeightWatcher, an empirical metric for deep learning complexity based on Random Matrix Theory
Check out the notebook
First, install the Docker client for your system.
Then, in a terminal, change to the project directory (the one containing this file) and:
- Test the installation using
docker info - Run
python build.pyto download data and run any preprocessing steps - Start the notebook container by running
sh start_notebook.shfrom this directory
Now your notebook server is running! Open a browser and point to http://localhost. Next,
- Enter the password token displayed on the terminal
- Click on
notebook.ipynbto open - If you're accessing a finished notebook, you can browse, edit the code, and execute the cells to reproduce or alter the figures.
- If you're starting a new notebook, read the project guidelines in the notebook and start coding!
created with cookiecutter, using the Data science project template