Python implementation of the Grizzly project. I'm currently working to troubleshoot the result of the adaptive attack, as it is incorrect.
Jupyter notebooks were used to help parameter tuning without having to wait for all operations to re-run.
Some changes were made:
get_bytes_class was included as a function in read_metadata.py.