Patents by Inventor Rihui Peng

Rihui Peng has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Distributed training for deep learning models
    Patent number: 11574253
    Abstract: A computer implemented method trains distributed sets of machine learning models by training each of the distributed machine learning models on different subsets of a set of training data, performing a first layer model synchronization operation in a first layer for each set of machine learning models, wherein each model synchronization operation in the first layer generates first updates for each of the machine learning models in each respective set, updating the machine learning models based on the first updates, performing a second layer model synchronization operation in a second layer for first supersets of the machine learning models wherein each model synchronization in the second layer generates second updates for updating each of the machine learning models in the first supersets based on the second updates such that each machine learning model in a respective first superset is the same.
    Type: Grant
    Filed: August 1, 2019
    Date of Patent: February 7, 2023
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Ivo José Garcia dos Santos, Mehdi Aghagolzadeh, Rihui Peng
  • DISTRIBUTED TRAINING FOR DEEP LEARNING MODELS
    Publication number: 20210035027
    Abstract: A computer implemented method trains distributed sets of machine learning models by training each of the distributed machine learning models on different subsets of a set of training data, performing a first layer model synchronization operation in a first layer for each set of machine learning models, wherein each model synchronization operation in the first layer generates first updates for each of the machine learning models in each respective set, updating the machine learning models based on the first updates, performing a second layer model synchronization operation in a second layer for first supersets of the machine learning models wherein each model synchronization in the second layer generates second updates for updating each of the machine learning models in the first supersets based on the second updates such that each machine learning model in a respective first superset is the same.
    Type: Application
    Filed: August 1, 2019
    Publication date: February 4, 2021
    Inventors: Ivo José Garcia dos Santos, Mehdi Aghagolzadeh, Rihui Peng