Facebook, Inc.

Meta Platforms, Inc.

Ou Jin

Tianshi Gao

Stuart Michael Bowers

Liang Xiong

Yifei Huang

Ying Zhang

Wenhai Yang

Ahmad Abdulmageed Mohammed Abdulkader

Qiang Wu

Hussein Mohamed Hassan Mehanna

Andrey Malevich

Dmytro Dzhulgakov

Parv Oberoi

Xiangyu Wang

Vignesh Ramanathan

Alexandre Defossez

To select the content to be presented to the user, a first latent vector is determined for a content item based on a first object associated with the content item. A second latent vector is determined for the content item based on a second object associated with the content item. A content item vector is then determined based on the first and second latent vectors. Furthermore, a user vector is determined based on interactions of the user with the first set of content objects and the second set of content objects. A score indicative of the likelihood of the user interacting with the content item is determined based on the content item vector and the user vector.

Deep semantic content selection

In one embodiment, a method includes accessing an input vector representing an input post, wherein: the vector space comprises clusters each associated with a topic; each cluster was determined based on a clustering of training-page vectors corresponding to training pages that each comprise training posts, each training post submitted by a user to a training page and comprises content selected by the user; and each training-page vector was generated by an ANN that was trained, based on the training posts of training pages associated with the ANN, to receive a post and then output a probability that the received post is related to the training posts of the training pages; determining that the input vector is located within a particular cluster in the vector space; and determining a topic of the input post based on the topic associated with the particular cluster that the input vector is located within.

Post topic classification

The present disclosure is directed to a high-capacity training and prediction machine learning platform that can support high-capacity parameter models (e.g., with 10 billion parameters). The platform generates a model for a metric of interest based on a known training set. The model includes parameters indicating importances of different features of the model, taken both singly and in pairs. The model may be applied to predict a value for the metric for given sets of objects, such as for a pair consisting of a user object and a content item object.

High-capacity machine learning system

An optimized computer architecture for training an neural network includes a system having multiple GPUs. The neural network may be divided into separate portions, and a different portion is assigned to each of the multiple GPUs. Within each GPU, its portion is further divided across multiple training worker threads in multiple processing cores, and each processing core has lock-free access to a local parameter memory. The local parameter memory of each GPU is separately, and individually, synchronized with a remote master parameter memory by lock memory access. Each GPU has a separate set of communication worker threads dedicated to data transfer between the GPU and the remote parameter memory so that the GPU's training worker threads are not involved with cross GPU communications.

Sparse neural network training optimization

An online system receives content items from a third party content provider. For each content item, the online system inputs an image into a neural network and extracts a feature vector from a hidden layer of the neural network. The online system compresses each feature vector by assigning a label to each feature value representing whether the feature value was above a threshold value. The online system identifies a set of content items that the user has interacted with and determines a user feature vector by aggregating feature vectors of the set of content items. For a new set of content items, the online system compares the compressed feature vectors of the content item with the user feature vector. The online system selects one or more of the new content items based on the comparison and sends the selected content items to the user.

Neural network based content distribution in an online system

The present disclosure is directed to a high-capacity training and prediction machine learning platform that can support high-capacity parameter models (e.g., with 10 billion weights). The platform implements a generic feature transformation layer for joint updating and a distributed training framework utilizing shard servers to increase training speed for the high-capacity model size. The models generated by the platform can be utilized in conjunction with existing dense baseline models to predict compatibilities between different groupings of objects (e.g., a group of two objects, three objects, etc.).

An online system simplifies modification of features used by machine learned models used by the online system, such as machined learned models with high dimensionality. The online system obtains a superset of features including features used by at least one machine learned model and may include additional features. From the superset of features, the online system generates various groups of features for a machine learned model. The groups of features may be a group including features currently used by the machine learned model, a group including all available features, and one or more intermediate groups. Intermediate groups include various numbers of features from the set selected based on measures of feature impact on the machine learned model associated with various features. A user may select a group of features, test the machine learning model using the selected group, and then launch the tested model based on the results.

Growing community of inventors

Fremont, CA, United States of America

Ou Jin