Tensor2Tensor for Neural Machine Translation

E899036

machine learning research research work scientific paper

"Tensor2Tensor for Neural Machine Translation" is a research work introducing a modular, scalable library and methodology for training state-of-the-art neural machine translation models.

Try in SPARQL Jump to: Statements Referenced by

Statements (37)

Predicate	Object
instanceOf	machine learning research ⓘ research work ⓘ scientific paper ⓘ
basedOn	deep learning frameworks ⓘ neural networks ⓘ sequence-to-sequence models ⓘ
contributesTo	advancement of neural machine translation research ⓘ
emphasizes	modularity ⓘ reusability of components ⓘ scalability ⓘ
evaluationDomain	machine translation benchmarks ⓘ
field	deep learning ⓘ natural language processing ⓘ neural machine translation ⓘ
focusesOn	modular library for neural machine translation ⓘ scalable training of neural machine translation models ⓘ standardized training methodology ⓘ state-of-the-art neural machine translation models ⓘ
goal	enable reproducible machine translation experiments ⓘ provide scalable model training infrastructure ⓘ simplify training of neural machine translation models ⓘ
hasMethodology	configuration-driven experiment management ⓘ standardized training and evaluation pipeline ⓘ
introduces	Tensor2Tensor library NERFINISHED ⓘ
language	English ⓘ
provides	data input pipelines for translation tasks ⓘ modular components for model architectures ⓘ training scripts and configurations ⓘ
relatedTo	Tensor2Tensor open-source project NERFINISHED ⓘ Transformer-based translation models ⓘ sequence-to-sequence learning ⓘ
supports	experimentation with different model hyperparameters ⓘ multiple neural machine translation architectures ⓘ
targetsUsers	machine learning researchers ⓘ neural machine translation practitioners ⓘ
usedFor	benchmarking neural machine translation systems ⓘ training translation models on large datasets ⓘ

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

Lukasz Kaiser → coAuthorOf → Tensor2Tensor for Neural Machine Translation ⓘ

subject surface form: Łukasz Kaiser