TY  - RPRT
TI  - Optimal checkpointing for heterogeneous chains: how to train deep neural networks with limited memory
AU  - Julien Herrmann
AU  - Olivier Beaumont
AU  - Lionel Eyraud-Dubois
AU  - Julien Hermann
AU  - Alexis Joly
AU  - Alena Shilova
PY  - 2019
UR  - https://arxiv.org/abs/1911.13214
ID  - 1911.13214
ER  -