An investigation into the representational power of Batch Normalisation
This repository contains a useful network trainer as well as all the experiment config files needed to run all experiements discussed in the paper
- `conda create --name mlp-cw4`
- `conda activate mlp-cw4`
- `make installEnv`
conda activate /conda_env \
python src/train.py --config configs/experiment_configs/sparse_resnet_56_cifar10.yaml \
python src/graph.py --experiment_name sparse_resnet_56_cifar10 --graph_type all \