Packages needed: torch
, torchvision
, tqdm
(for progress bar)
To run all experiments, execute:
python3 lab2.py
In lab2.py
, the code to run each experiment is a function (e.g. c5()
, q3()
). You can selectively enable or disable running them by commenting out lines 64-70.
File resnet18.py
contains the model definition, while train.py
contains the training and performance measurement code.