MaskPro learns categorical distributions over groups of M weights to generate exact (N:M) sparsity via N-way sampling without replacement and stabilizes training with a moving average tracker of loss residuals.
Figure 5: Loss residual curves of training on LLaMA-2-7B model with 1, 32, 128, and 320k samples
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MaskPro: Linear-Space Probabilistic Learning for Strict (N:M)-Sparsity on LLMs
MaskPro learns categorical distributions over groups of M weights to generate exact (N:M) sparsity via N-way sampling without replacement and stabilizes training with a moving average tracker of loss residuals.