Google Inc.

Lukasz Mieczyslaw Kaiser

Noam M Shazeer

Niki J Parmar

Ashish Teku Vaswani

Jakob David Uszkoreit

Aidan Nicholas Gomez

Llion Owen Jones

Illia Polosukhin

Oriol Vinyals

Peter Junteng Liu

Ben David Goodrich

Mohammad Saleh

Ryan Sepassi

Etienne Pot

Jakob D Uszkoreit

Samuel Bengio

Mostafa Dehghani

Anselm Levskaya

Francois Chollet

Nikita Kitaev

Stephan Gouws

Nitin Kumar Gupta

Yang Li

Ilya Sutskever

Enrique Alfonseca

Ni Lao

Preyas Popat

Aurko Roy

Ekaterina Filippova

Afroz Mohiuddin

Si Si

Carlos Alberto Colmenares Rojas

Mostafa Deghani

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output sequence from an input sequence. One of the methods includes, at each of a plurality of generation time steps: generating a combined sequence for the generation time step that includes the input sequence followed by the output tokens that have already been generated as of the generation time step; processing the combined sequence using a self-attention decoder neural network to generate a time step output that defines a score distribution over a set of possible output tokens; and selecting, using the time step output, an output token from the set of possible output tokens as the next output token in the output sequence.

Attention-based decoder-only sequence transduction neural networks

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output sequence from an input sequence. One of the methods includes receiving the input sequence; processing the input sequence using a latent prediction model configured to autoregressively predict a sequence of discrete latent variables that is shorter than the output sequence and that encodes the output sequence, wherein each discrete latent variable in the sequence is selected from a discrete set of latent variables; and processing the input sequence and the predicted sequence of discrete latent variables using a parallel decoder model configured to generate the outputs in the output sequence in parallel from the input sequence and the predicted sequence of discrete latent variables.

Fast decoding in sequence models using discrete latent variables

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing a machine learning task on a network input to generate a network output. In one aspect, one of the systems includes an attention neural network configured to perform the machine learning task, the attention neural network including one or more LSH attention layers, each LSH attention layer comprising one or more LSH attention sub-layers, each LSH sub-layer configured to: receive a sequence of queries derived from an input sequence to the LSH attention layer, the sequence of queries having a respective query at each of a plurality of input positions; determine one or more respective hash values for each of the respective queries at each of the plurality of input positions; generate a plurality of LSH groupings; and generate an attended input sequence.

Attention neural networks with locality-sensitive hashing

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output sequence from an input sequence. In one aspect, one of the systems includes an encoder neural network configured to receive the input sequence and generate encoded representations of the network inputs, the encoder neural network comprising a sequence of one or more encoder subnetworks, each encoder subnetwork configured to receive a respective encoder subnetwork input for each of the input positions and to generate a respective subnetwork output for each of the input positions, and each encoder subnetwork comprising: an encoder self-attention sub-layer that is configured to receive the subnetwork input for each of the input positions and, for each particular input position in the input order: apply an attention mechanism over the encoder subnetwork inputs using one or more queries derived from the encoder subnetwork input at the particular input position.

Attention-based sequence transduction neural networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for implementing an area attention layer in a neural network system. The area attention layer area implements a way for a neural network model to attend to areas in the memory, where each area contains a group of items that are structurally adjacent.

Neural networks with area attention

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output image. In one aspect, one of the methods includes generating the output image intensity value by intensity value according to a generation order of pixel-color channel pairs from the output image, comprising, for each particular generation order position in the generation order: generating a current output image representation of a current output image, processing the current output image representation using a decoder neural network to generate a probability distribution over possible intensity values for the pixel-color channel pair at the particular generation order position, wherein the decoder neural network includes one or more local masked self-attention sub-layers; and selecting an intensity value for the pixel-color channel pair at the particular generation order position using the probability distribution.

Attention-based image generation neural networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for implementing a sequence to sequence model that is recurrent in depth while employing self-attention to combine information from different parts of sequences.

Universal transformers

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output image. In one aspect, one of the methods includes generating the output image intensity value by intensity value according to a generation order of pixel&#x2014;color channel pairs from the output image, comprising, for each particular generation order position in the generation order: generating a current output image representation of a current output image, processing the current output image representation using a decoder neural network to generate a probability distribution over possible intensity values for the pixel&#x2014;color channel pair at the particular generation order position, wherein the decoder neural network includes one or more local masked self-attention sub-layers; and selecting an intensity value for the pixel&#x2014;color channel pair at the particular generation order position using the probability distribution.

Methods, systems, and apparatus, including computer programs encoded on computer storage media for performing machine translation tasks. One method includes receiving an input text segment in an input language; processing the input text segment using an encoder neural network to generate an encoder neural network output, the encoder neural network comprising multiple depth wise separable convolutional neural network layers; processing the encoder neural network output using an autoregressive decoder neural network to generate a decoder neural network output; and processing the decoder neural network output to generate a predicted output text segment in a target natural language.

Depthwise separable convolutions for neural machine translation

Methods, systems, and apparatus, including computer programs encoded on computer storage media for training a machine learning model to perform multiple machine learning tasks from multiple machine learning domains. One system includes a machine learning model that includes multiple input modality neural networks corresponding to respective different modalities and being configured to map received data inputs of the corresponding modality to mapped data inputs from a unified representation space; an encoder neural network configured to process mapped data inputs from the unified representation space to generate respective encoder data outputs; a decoder neural network configured to process encoder data outputs to generate respective decoder data outputs from the unified representation space; and multiple output modality neural networks corresponding to respective different modalities and being configured to map decoder data outputs to data outputs of the corresponding modality.

Multi-task multi-modal machine learning system

Growing community of inventors

Mountain View, CA, United States of America

Lukasz Mieczyslaw Kaiser