GaLore performs full-parameter LLM training with up to 65.5% less optimizer memory by projecting gradients onto a low-rank subspace at each step, matching full-rank performance on LLaMA pre-training and RoBERTa fine-tuning.
arXiv preprint arXiv:2209.13569 , year=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
fields
cs.LG 5representative citing papers
Dr. Post-Training reframes general data as a data-induced regularizer for LLM post-training updates, yielding a family of methods that outperform data-selection baselines on SFT, RLHF, and RLVR tasks.
BOOST delivers 1.46-2.27x end-to-end speedups for low-rank bottleneck LLMs by redesigning tensor parallelism around the bottleneck structure plus supporting optimizations.
CR-Net uses cross-layer low-rank residuals in a dual-path network plus specialized recomputation to outperform prior low-rank methods on 60M-7B model pre-training while using less compute and memory.
DLR augments low-rank factorization with a fixed structured residual during training that is absorbed post-training, improving C4 perplexity for LLaMA models from 60M to 7B while preserving exact low-rank inference cost.
citing papers explorer
-
Dr. Post-Training: A Data Regularization Perspective on LLM Post-Training
Dr. Post-Training reframes general data as a data-induced regularizer for LLM post-training updates, yielding a family of methods that outperform data-selection baselines on SFT, RLHF, and RLVR tasks.
-
BOOST: BOttleneck-Optimized Scalable Training Framework for Low-Rank Large Language Models
BOOST delivers 1.46-2.27x end-to-end speedups for low-rank bottleneck LLMs by redesigning tensor parallelism around the bottleneck structure plus supporting optimizations.
-
CR-Net: Scaling Parameter-Efficient Training with Cross-Layer Low-Rank Structure
CR-Net uses cross-layer low-rank residuals in a dual-path network plus specialized recomputation to outperform prior low-rank methods on 60M-7B model pre-training while using less compute and memory.
-
DLR: Zero-Inference-Cost Latent Residuals for Low-Rank Pre-Training
DLR augments low-rank factorization with a fixed structured residual during training that is absorbed post-training, improving C4 perplexity for LLaMA models from 60M to 7B while preserving exact low-rank inference cost.