A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf
.
Filters
Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling
[article]
2019
arXiv
pre-print
Lingvo is a Tensorflow framework offering a complete solution for collaborative deep learning research, with a particular focus towards sequence-to-sequence models. ...
Lingvo models are composed of modular building blocks that are flexible and easily extensible, and experiment configurations are centralized and highly customizable. ...
Introduction This paper presents the open-source Lingvo framework developed by Google for sequence modeling with deep neural networks. ...
arXiv:1902.08295v1
fatcat:cvpshrgz6zgbjlylc5hkb2ku3i
RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions
[article]
2020
arXiv
pre-print
In this work, we analyze the generalization properties of streaming and non-streaming recurrent neural network transducer (RNN-T) based end-to-end models in order to identify model components that negatively ...
On a long-form YouTube test set, when the nonstreaming RNN-T model is trained with shorter segments of data, the proposed combination improves word error rate (WER) from 22.3% to 14.8%; when the streaming ...
Chen, and et al., “Lingvo:
a modular and scalable framework for sequence-to-
sequence modeling,” 2019.
[32] A. Senior, H. Sak, F. de Chaumont Quitry, T. Sainath,
and K. ...
arXiv:2005.03271v3
fatcat:k73x2lhlfzaujpbwnvu6rexptm
Optimizing Analytical Queries over Semantic Web Sources
[article]
2017
PhD series, Technical Faculty of IT and Design, ˜Aalborg=ålborgœ University
Chapter 2 presented a framework for Exploratory OLAP over LOD sources. We followed the best practices in the system design and constructed the framework on a modular basis. ...
A framework to streamline the ETL process from Linked Open Data to a multidimensional data model is proposed in [13] . ...
doi:10.5278/vbn.phd.tech.00022
fatcat:5d2lmnicazgmze2qhe5cikowwm
Concept Relation Discovery and Innovation Enabling Technology (Cordiet)
2010
Social Science Research Network
It is important for lingvo-didactics, as a basis for creation of various manuals for speech developing, training in Russian for the Russians and the foreigners, and also for translation of legal, psychological ...
Conclusion The methodology presented in this paper provides a framework for solving diverse and very important problems of constructing machine learning algorithms based on a unified logical model in which ...
is partially supported by the Russian Foundation for Basic Research, project No. 08-07-92497 -NTSNIL_a. ...
doi:10.2139/ssrn.1713124
fatcat:f27e6pnel5awtkziz2qxmpqrzq
Transformer-based NMT : modeling, training and implementation
[article]
2021
Transformer-Based NMT: Modeling, Training and Implementation ...
I also thank our supporters, collaborators and anonymous reviewers for their efforts in helping us improve our work. ...
Acknowledgements Acknowledgements I thank my colleagues at Saarland University and DFKI for their help. Working with them is pleasant. ...
doi:10.22028/d291-34998
fatcat:4d226kujkjeodaiwzxltmgluhi
\/)/^<i CJ^ji^^-^n^ A Safe and EfRcient Agent Architecture
1997
Informatica
unpublished
This framework combines the culture support in the domain ontology and the decomposition and causality of partial-order planning. ...
How can an agent maintain its safety enforcement with incomplete Information in a computation tractable manner? This paper presents an agent architecture which can be used to solve the above problem. ...
Fang for sharing his insight into the non-monotonic reasoning, M. Clint, Se Woo Cheon and A. Stewart for their constructive comments on the early versions of this paper, and N. D. N. ...
fatcat:j53p7u2n4jblbe5xtld7refca4