Search Results for author: Paulius Micikevicius

Found 7 papers, 5 papers with code

Accelerating Sparse Deep Neural Networks

no code implementations16 Apr 2021 Asit Mishra, Jorge Albericio Latorre, Jeff Pool, Darko Stosic, Dusan Stosic, Ganesh Venkatesh, Chong Yu, Paulius Micikevicius

We present the design and behavior of Sparse Tensor Cores, which exploit a 2:4 (50%) sparsity pattern that leads to twice the math throughput of dense matrix units.

Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation

1 code implementation20 Apr 2020 Hao Wu, Patrick Judd, Xiaojie Zhang, Mikhail Isaev, Paulius Micikevicius

Quantization techniques can reduce the size of Deep Neural Networks and improve inference latency and throughput by taking advantage of high throughput integer instructions.

Quantization

Mixed-Precision Training for NLP and Speech Recognition with OpenSeq2Seq

4 code implementations25 May 2018 Oleksii Kuchaiev, Boris Ginsburg, Igor Gitman, Vitaly Lavrukhin, Jason Li, Huyen Nguyen, Carl Case, Paulius Micikevicius

We present OpenSeq2Seq - a TensorFlow-based toolkit for training sequence-to-sequence models that features distributed and mixed-precision training.

Automatic Speech Recognition Machine Translation +3

Cannot find the paper you are looking for? You can Submit a new open access paper.