paper-with-me

Papers

Efficient Backpropagation with Variance-Controlled Adaptive Sampling

2024-02-27 · Ziteng Wang, Jianfei Chen, Jun Zhu

Sampling-based algorithms, which eliminate ''unimportant'' computations during forward and/or back propagation (BP), offer potential solutions to accelerate neural network training. However, since sampling introduces approximations to training, such algorithms may not consistently maintain accuracy across various tasks. In this work, we introduce a variance-controlled adaptive sampling (VCAS) method designed to accelerate BP. VCAS computes an unbiased stochastic gradient with fine-grained layerwise importance sampling in data dimension for activation gradient calculation and leverage score sampling in token dimension for weight gradient calculation. To preserve accuracy, we control the additional variance by learning the sample ratio jointly with model parameters during training. We assessed VCAS on multiple fine-tuning and pre-training tasks in both vision and natural language domains. On all the tasks, VCAS can preserve the original training loss trajectory and validation accuracy with an up to 73.87% FLOPs reduction of BP and 49.58% FLOPs reduction of the whole training process. The implementation is available at https://github.com/thu-ml/VCAS .

📄 PDF Abstract BibTeX arXiv:2402.17227

Code (1)

thu-ml/vcas 공식 구현 pytorch

Similar Papers 제목 키워드 기반

Covariance-Controlled Adaptive Langevin Thermostat for Large-Scale Bayesian Sampling

2015-10-29 · NeurIPS 2015 12 · Xiaocheng Shang, Zhanxing Zhu, Benedict Leimkuhler, Amos J. Storkey

Monte Carlo sampling for Bayesian posterior inference is a common approach used in machine learning. The Markov Chain Monte Carlo procedures that are used are often discrete-time analogues of associated stochastic differ…

BIG-bench Machine Learning

Zero-Order Optimization for LLM Fine-Tuning via Learnable Direction Sampling

2026-02-14 · Valery Parfenov, Grigoriy Evseev, Andrey Veprikov, Nikolay Bushkov 외 arxiv

Fine-tuning large pretrained language models (LLMs) is a cornerstone of modern NLP, yet its growing memory demands (driven by backpropagation and large optimizer States) limit deployment in resource-constrained settings.…

GRADE: Replacing Policy Gradients with Backpropagation for LLM Alignment

2025-12-30 · Lukas Abrie Nel arxiv

Reinforcement learning from human feedback (RLHF) has become the dominant paradigm for aligning large language models with human preferences. However, policy gradient methods such as PPO suffer from high variance gradien…

Reinforcement LearningText Generation

An Investigation of the Bias-Variance Tradeoff in Meta-Gradients

2022-09-22 · Risto Vuorio, Jacob Beck, Shimon Whiteson, Jakob Foerster 외

Meta-gradients provide a general approach for optimizing the meta-parameters of reinforcement learning (RL) algorithms. Estimation of meta-gradients is central to the performance of these meta-algorithms, and has been st…

Meta-LearningReinforcement Learning (RL)

Subset Selection for Stratified Sampling in Online Controlled Experiments

2025-09-19 · Haru Momozu, Yuki Uehara, Naoki Nishimura, Koya Ohashi 외 arxiv

Online controlled experiments, also known as A/B testing, are the digital equivalent of randomized controlled trials for estimating the impact of marketing campaigns on website visitors. Stratified sampling is a traditio…