calgaryml / condensed-sparsity

[ICLR 2024] Dynamic Sparse Training with Structured Sparsity
https://openreview.net/forum?id=kOBkxFRKTA
MIT License
17 stars 3 forks source link

Convert masks for weight matricies into object #7

Closed mklasby closed 1 year ago

mklasby commented 2 years ago

Ideally, the masks used by RigL would be wrapped into an object with some helper functions as we see fit.

Acceptance Critera:

mklasby commented 1 year ago

TODO in future refactor as part of a sparse framework for pytorch