Skip to content

Latest commit

 

History

History
12 lines (9 loc) · 650 Bytes

README.md

File metadata and controls

12 lines (9 loc) · 650 Bytes

mbtrim

Mini-batch trimming python implementation (see my ASPAI 2021 paper "Some like it tough: Improving model generalization via progressively increasing the training difficulty", arxiv preprint at https://arxiv.org/abs/2110.13058).

The module in 'mnist_model.py' is a SimpleDensenet Model for MNIST classification. It is a pytorch lightning module, so you must have pytorch lightning installed.

See Pytorch lightning main page