Announcement_15_iclr2024

Mike Lasby’s collaborative work with researchers at Google DeepMind, MIT and the Vector Institute, “Dynamic Sparse Training with Structured Sparsity” (Lasby et al., 2024), was accepted at the International Conference on Learning Representations (ICLR), 2024! DST methods learn state-of-the-art sparse masks, but accelerating DNNs with unstructured masks is difficult. SRigL learns structured masks, improving real-world CPU/GPU timings!