/checkmate-1

Train huge neural nets without running out of GPU memory

Checkmate

Checkmate helps you train huge neural nets without running out of GPU memory. One line of code will avoid the fatal CUDA_OUT_OF_MEMORY error!

Checkmate will be released in January 2020. Keep an eye out for the upcoming release!

Citation

If you use Checkmate in your work, please cite us with:

@article{jain2019checkmate,
  title={Checkmate: Breaking the Memory Wall with Optimal Tensor Rematerialization},
  author={Jain, Paras and Jain, Ajay and Nrusimha, Aniruddha and Gholami, Amir and Abbeel, Pieter and Keutzer, Kurt and Stoica, Ion and Gonzalez, Joseph E},
  journal={arXiv preprint arXiv:1910.02653},
  year={2019}
}