conftrace_

subspace descent

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

low-rank training (1) large language model (1) large language model training (1) online pca (1) optimizer state (1) model training (1) memory efficiency (1) singular value decomposition (1) gradient descent (1) neural network optimization (1)

Papers

Memory-Efficient LLM Training with Online Subspace Descent NeurIPS 2024