Depth-adaptive Transformer - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Depth-adaptive Transformer

Résumé

State of the art sequence-to-sequence models for large scale tasks perform a fixed number of computations for each input sequence regardless of whether it is easy or hard to process. In this paper, we train Transformer models which can make output predictions at different stages of the network and we investigate different ways to predict how much computation is required for a particular sequence. Unlike dynamic computation in Universal Transformers, which applies the same set of layers iteratively, we apply different layers at every step to adjust both the amount of computation as well as the model capacity. On IWSLT German-English translation our approach matches the accuracy of a well tuned baseline Transformer while using less than a quarter of the decoder layers.

Mots clés

Fichier principal
Vignette du fichier
paper.pdf (332.46 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02422914 , version 1 (23-12-2019)

Identifiants

  • HAL Id : hal-02422914 , version 1

Citer

Maha Elbayad, Jiatao Gu, Edouard Grave, Michael Auli. Depth-adaptive Transformer. ICLR 2020 - Eighth International Conference on Learning Representations, Apr 2020, Addis Ababa, Ethiopia. pp.1-14. ⟨hal-02422914⟩
331 Consultations
449 Téléchargements

Partager

Gmail Facebook X LinkedIn More