Výsledky vyhledávání - "Gopalswamy, Karthick"

Report

Collage: Light-Weight Low-Precision Strategy for LLM Training

Autor: Yu, Tao, Gupta, Gaurav, Gopalswamy, Karthick, Mamidala, Amith, Zhou, Hao, Huynh, Jeffrey, Park, Youngsuk, Diamant, Ron, Deoras, Anoop, Huan, Luke

Large models training is plagued by the intense compute cost and limited hardware memory. A practical solution is low-precision representation but is troubled by loss in numerical accuracy and unstable training rendering the model less useful. We arg

Externí odkaz: http://arxiv.org/abs/2405.03637

Zobrazit plný text záznamu

Report

Pessimistic Off-Policy Multi-Objective Optimization

Autor: Alizadeh, Shima, Bhargava, Aniruddha, Gopalswamy, Karthick, Jain, Lalit, Kveton, Branislav, Liu, Ge

Multi-objective optimization is a type of decision making problems where multiple conflicting objectives are optimized. We study offline optimization of multi-objective policies from data collected by an existing policy. We propose a pessimistic esti

Externí odkaz: http://arxiv.org/abs/2310.18617

Zobrazit plný text záznamu

Report

Cross-Frequency Time Series Meta-Forecasting

Autor: Van Ness, Mike, Shen, Huibin, Wang, Hao, Jin, Xiaoyong, Maddix, Danielle C., Gopalswamy, Karthick

Meta-forecasting is a newly emerging field which combines meta-learning and time series forecasting. The goal of meta-forecasting is to train over a collection of source time series and generalize to new time series one-at-a-time. Previous approaches

Externí odkaz: http://arxiv.org/abs/2302.02077

Zobrazit plný text záznamu

Report

First De-Trend then Attend: Rethinking Attention for Time-Series Forecasting

Autor: Zhang, Xiyuan, Jin, Xiaoyong, Gopalswamy, Karthick, Gupta, Gaurav, Park, Youngsuk, Shi, Xingjian, Wang, Hao, Maddix, Danielle C., Wang, Yuyang

Transformer-based models have gained large popularity and demonstrated promising results in long-term time-series forecasting in recent years. In addition to learning attention in time domain, recent works also explore learning attention in frequency

Externí odkaz: http://arxiv.org/abs/2212.08151

Zobrazit plný text záznamu