Zobrazeno 1 - 10
of 17
pro vyhledávání: '"Olatunji Ruwase"'
Autor:
Reza Yazdani Aminabadi, Olatunji Ruwase, Minjia Zhang, Yuxiong He, Jose-Maria Arnau, Antonio Gonazalez
Publikováno v:
ACM Transactions on Embedded Computing Systems
The effectiveness of Recurrent Neural Networks (RNNs) for tasks such as Automatic Speech Recognition has fostered interest in RNN inference acceleration. Due to the recurrent nature and data dependencies of RNN computations, prior work has designed c
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::53dcaf60cdc72422386309b88f68f83d
https://hdl.handle.net/2117/389907
https://hdl.handle.net/2117/389907
Publikováno v:
SC
In the last three years, the largest dense deep learning models have grown over 1000x to reach hundreds of billions of parameters, while the GPU memory has only grown by 5x (16 GB to 80 GB). Therefore, the growth in model scale has been supported pri
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::7559ae007f86a11026952f1966ea4ce7
https://doi.org/10.1145/3458817.3476205
https://doi.org/10.1145/3458817.3476205
Publikováno v:
KDD
Explore new techniques in Microsoft's open source library called DeepSpeed, which advances large model training by improving scale, speed, cost, and usability, unlocking the ability to train 100-billion-parameter models. DeepSpeed is compatible with
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::8d8e47466ad5e4bf93349ed52cceb829
https://doi.org/10.1145/3394486.3406703
https://doi.org/10.1145/3394486.3406703
Publikováno v:
IEEE Transactions on Network and Service Management. 15:112-126
The emergence of deep neural networks (DNNs) as a state-of-the-art machine learning technique has enabled a variety of artificial intelligence applications for image recognition, speech recognition and translation, drug discovery, and machine vision.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::039577f42b1e60223494000ccdf4d9c7
https://doi.org/10.1109/tnsm.2018.2808352
https://doi.org/10.1109/tnsm.2018.2808352
Publikováno v:
SC
Large deep learning models offer significant accuracy gains, but training billions to trillions of parameters is challenging. Existing solutions such as data and model parallelisms exhibit fundamental limitations to fit these models into limited devi
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2870098d14ff1ad81f68e71fea157b3d
Publikováno v:
Middleware
The quality of machine learning (ML) and deep learning (DL) models are very sensitive to many different adjustable parameters that are set before training even begins, commonly called hyperparameters. Efficient hyperparameter exploration is of great
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::2f74482782bb752d0629732b4cef60c1
https://doi.org/10.1145/3135974.3135994
https://doi.org/10.1145/3135974.3135994
Publikováno v:
ASPLOS
Convolutional Neural Networks (CNN) are a class of Ar- tificial Neural Networks (ANN) that are highly efficient at the pattern recognition tasks that underlie difficult AI prob- lems in a variety of domains, such as speech recognition, object recogni
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a56a8c4abd2738fb9864fcf0701cc102
https://doi.org/10.1145/3037697.3037745
https://doi.org/10.1145/3037697.3037745
Publikováno v:
SC16: International Conference for High Performance Computing, Networking, Storage and Analysis.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::5ce64af802e71684cbe82c1ad0219aae
https://doi.org/10.1109/sc.2016.25
https://doi.org/10.1109/sc.2016.25
Autor:
Theodoros Strigkos, Phillip B. Gibbons, Vijaya Ramachandran, Michael Kozuch, Todd C. Mowry, Babak Falsafi, Shimin Chen, Evangelos Vlachos, Michael P. Ryan, Olatunji Ruwase
Publikováno v:
IEEE Micro. 29:62-72
Instruction-grain lifeguards monitor executing programs at the granularity of individual instructions to quickly detect bugs and security attacks, but their fine-grain nature incurs high monitoring overheads. This article identifies three common sour
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c83fb487a078de02c59c652777953b92
https://doi.org/10.1109/mm.2009.6
https://doi.org/10.1109/mm.2009.6
Autor:
Todd C. Mowry, Michael Kozuch, Phillip B. Gibbons, Evangelos Vlachos, Olatunji Ruwase, Michael P. Ryan, Shimin Chen, Babak Falsafi, Theodoros Strigkos, Vijaya Ramachandran
Publikováno v:
ISCA
Instruction-grain program monitoring tools, which check and analyze executing programs at the granularity of individual instructions, are invaluable for quickly detecting bugs and security attacks and then limiting their damage (via containment and/o
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::93bb1e1111b07a44083442246e4dfdd7
https://doi.org/10.1145/1394608.1382153
https://doi.org/10.1145/1394608.1382153