Opara: Exploiting Operator Parallelism for Expediting DNN Inference on GPUs

1 code implementation • 16 Dec 2023 • Aodong Chen, Fei Xu, Li Han, Yuan Dong, Li Chen, Zhi Zhou, Fangming Liu

GPUs have become the defacto hardware devices to accelerate Deep Neural Network (DNN) inference in deep learning(DL) frameworks.

Scheduling

The Impact of Industrial Zone:Evidence from China's National High-tech Zone Policy

no code implementations • 19 Apr 2023 • Li Han

Based on the statistical yearbook data and related patent data of 287 cities in China from 2000 to 2020, this study regards the policy of establishing the national high-tech zones as a quasi-natural experiment.

Learning on Abstract Domains: A New Approach for Verifiable Guarantee in Reinforcement Learning

no code implementations • 13 Jun 2021 • Peng Jin, Min Zhang, Jianwen Li, Li Han, Xuejun Wen

Formally verifying Deep Reinforcement Learning (DRL) systems is a challenging task due to the dynamic continuity of system behaviors and the black-box feature of embedded neural networks.

reinforcement-learning Reinforcement Learning (RL)

Linear encoding of the spatiotemporal cat

1 code implementation • 6 Dec 2019 • Boris Gutkin, Li Han, Rana Jafari, Adrien K. Saremi, Predrag Cvitanović

The dynamics of an extended, spatiotemporally chaotic system might appear extremely complex.

Chaotic Dynamics Statistical Mechanics High Energy Physics - Lattice Mathematical Physics Mathematical Physics 28D05, 60J10, 70K43, 70S05, 76F20, 82C20