ShiftAddLLM: Accelerating Pretrained LLMs via Post-Training Multiplication-Less Reparameterization Paper • 2406.05981 • Published Jun 10, 2024 • 13
SLoPe: Double-Pruned Sparse Plus Lazy Low-Rank Adapter Pretraining of LLMs Paper • 2405.16325 • Published May 25, 2024 • 1
Progressive Gradient Flow for Robust N:M Sparsity Training in Transformers Paper • 2402.04744 • Published Feb 7, 2024 • 2
STEP: Learning N:M Structured Sparsity Masks from Scratch with Precondition Paper • 2302.01172 • Published Feb 2, 2023
ArchGym: An Open-Source Gymnasium for Machine Learning Assisted Architecture Design Paper • 2306.08888 • Published Jun 15, 2023 • 1