4 ms·
Training Neural Networks with Fixed Sparse Masks
- spekcular 5y agoA Twitter thread by one of the authors, for more context: https://twitter.com/colinraffel/status/1468618801134592012 https://twitter.com/colinraffel/status/1468618801134592012
loading thread…