by Lei Wu, Chao Ma, Weinan E
python train.py --dataset fashionmnist --training_size 1000 --model_file net.pkl
python diagnose.py --dataset fashionmnist --training_size 1000 --model_file net.pkl
- pytorch >= 0.4
@inproceedings{leiwu2018,
title={How SGD Selects Global Minima in Over-parameterized Learning: A Dynamical Stability Perspetive},
author={Wu, Lei and Ma, Chao and E, Weinan},
booktitle={Advances in Neural Information Processing Systems},
year={2018}
}