利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

8 条记录 (耗时 0.017 秒)

    BI-DIRECTIONAL GRADIENT COMPRESSION FOR DISTRIBUTED AND FEDERATED LEARNING
    3.
    发明公开
    BI-DIRECTIONAL GRADIENT COMPRESSION FOR DISTRIBUTED AND FEDERATED LEARNING 审中-公开

    公开(公告)号：US20240296317A1

    公开(公告)日：2024-09-05

    申请号：US18176867

    申请日：2023-03-01

    申请人： VMware LLC

    发明人： Yaniv Ben-Itzhak ,  Shay Vargaftik

    IPC分类号： G06N3/0495 ,  G06N3/098

    CPC分类号： G06N3/0495 ,  G06N3/098

    摘要： Improved techniques for compressing gradient information that is communicated between clients and a parameter server in a distributed or federated learning training procedure are disclosed. In certain embodiments these techniques enable bi-directional gradient compression, which refers to the compression of both (1) the gradients sent by the participating clients in a given round to the parameter server and (2) the global gradient returned by the parameter server to those clients. In further embodiments, the techniques of the present disclosure eliminate the need for the parameter server to decompress each received gradient as part of computing the global gradient, thereby improving training performance.

    ACCELERATING DATA MESSAGE CLASSIFICATION WITH SMART NICS
    4.
    发明公开
    ACCELERATING DATA MESSAGE CLASSIFICATION WITH SMART NICS 审中-公开

    公开(公告)号：US20240184708A1

    公开(公告)日：2024-06-06

    申请号：US18437627

    申请日：2024-02-09

    申请人： VMware LLC

    发明人： Shay Vargaftik ,  Alex Markuze ,  Yaniv Ben-Itzhak ,  Igor Golikov ,  Avishay Yanai

    IPC分类号： G06F12/0891 ,  G06F13/16

    CPC分类号： G06F12/0891 ,  G06F13/1668 ,  G06F2213/0026 ,  G06F2213/3808

    摘要： Some embodiments provide a method for performing data message processing at a smart NIC of a computer that executes a software forwarding element (SFE). The method determines whether a received data message matches an entry in a data message classification cache stored on the smart NIC based on data message classification results of the SFE. When the data message matches an entry, the method determines whether the matched entry is valid by comparing a timestamp of the entry to a set of rules stored on the smart NIC. When the matched entry is valid, the method processes the data message according to the matched entry without providing the data message to the SFE executing on the computer.

    Accelerating data message classification with smart NICs
    5.
    发明授权
    Accelerating data message classification with smart NICs 有权

    公开(公告)号：US11928062B2

    公开(公告)日：2024-03-12

    申请号：US17845661

    申请日：2022-06-21

    申请人： VMware LLC

    发明人： Shay Vargaftik ,  Alex Markuze ,  Yaniv Ben-Itzhak ,  Igor Golikov ,  Avishay Yanai

    IPC分类号： G06F12/0891 ,  G06F13/16

    CPC分类号： G06F12/0891 ,  G06F13/1668 ,  G06F2213/0026 ,  G06F2213/3808

    摘要： Some embodiments provide a method for performing data message processing at a smart NIC of a computer that executes a software forwarding element (SFE). The method determines whether a received data message matches an entry in a data message classification cache stored on the smart NIC based on data message classification results of the SFE. When the data message matches an entry, the method determines whether the matched entry is valid by comparing a timestamp of the entry to a set of rules stored on the smart NIC. When the matched entry is valid, the method processes the data message according to the matched entry without providing the data message to the SFE executing on the computer.

    COMPRESSION OF MODEL WEIGHTS FOR DISTRIBUTED AND FEDERATED LEARNING
    6.
    发明公开
    COMPRESSION OF MODEL WEIGHTS FOR DISTRIBUTED AND FEDERATED LEARNING 审中-公开

    公开(公告)号：US20240202543A1

    公开(公告)日：2024-06-20

    申请号：US18067503

    申请日：2022-12-16

    申请人： VMware LLC

    发明人： Yaniv Ben-Itzhak ,  Shay Vargaftik ,  Ron Dorfman

    IPC分类号： G06N3/098

    CPC分类号： G06N3/098

    摘要： Techniques for compressing the model weights of an artificial neural network (ANN) in the context of distributed learning (DL) or federated learning (FL) are provided. In one set of embodiments, these techniques include a hybrid approach for compressing the model weights that employs (1) a high complexity compression scheme to compress the ANN's model weights every k rounds of the DL/FL procedure (referred to as anchor rounds), and (2) a low complexity compression scheme to compress accumulated differences in model weights of the ANN for intermediate rounds between the anchor rounds.