公开(公告)号：US20230076457A1

公开(公告)日：2023-03-09

申请号：US17888513

申请日：2022-08-16

Applicant: Zhejiang Lab

Inventor： Feng GAO ,  Wenyuan BAI

IPC: G06N3/04 ,  G06N3/08

Abstract: The present invention discloses an edge calculation-oriented reparametric neural network architecture search method, including the following steps: S1: designing linear operators and multi-branch block structures; S2: constructing a hypernetwork by stacking the multi-branch block structures; S3: training the hypernetwork through a gradient-based first-stage search algorithm; S4: deleting redundant branches in the hypernetwork to construct an optimal subnetwork; S5: converting the multi-branch optimal subnetwork into a single-branch network; and S6: completing task reasoning by using the single-branch network. The method is used to search the neural network structure capable of performing reparameterization, and ensures the reasoning real-time performance and the high efficiency of model operation while ensuring the reasoning precision.

公开(公告)号：US20230350652A1

公开(公告)日：2023-11-02

申请号：US18110910

申请日：2023-02-17

Applicant: ZHEJIANG LAB

Inventor： Wenyuan BAI ,  Feng GAO

IPC: G06F8/41 ,  G06F8/71

CPC classification number: G06F8/443 ,  G06F8/71

Abstract: A job decomposition processing method for distributed computing, which comprises: analyzing a source program to be run by program static analysis to determine a function call graph contained in the source program; determining feature information of functions contained in the source program by program dynamics analysis or/and a program intelligent decomposition algorithm, wherein the feature information of the functions is used to characterize relevant information when each function is being running; decomposing the source program based on the feature information of the functions, a function relationship and available resource information of a computing platform to form an execution recommendation for each function on the computing platform, i.e., which hardware resources are used for computing each function; finally inserting a modifier in the source program and starting computation on the computing platform.

公开(公告)号：US20240111586A1

公开(公告)日：2024-04-04

申请号：US18472648

申请日：2023-09-22

Applicant: ZHEJIANG LAB

Inventor： Shiqiang ZHU ,  Aimin PAN ,  Feng GAO

IPC: G06F9/50

CPC classification number: G06F9/5027

Abstract: The present disclosure belongs to the field of intelligent computing technologies, and relates to a multi-policy intelligent scheduling methods and apparatuses oriented to heterogeneous computing power. The method includes: step 1, setting an execution policy of a task based on heterogeneity of computing clusters, differences of computing tasks and a user requirement, and constructing a Markov decision process model by adopting a reinforcement learning method combined with the execution policy; step 2, adopting a proximal policy optimization to solve an optimal task scheduling policy of the task input by the user based on the constructed Markov decision process model; step 3, scheduling the task to a corresponding computing cluster for execution based on the optimal task scheduling policy.

公开(公告)号：US20240095069A1

公开(公告)日：2024-03-21

申请号：US18554450

申请日：2023-04-13

Applicant: ZHEJIANG LAB

Inventor： Feng GAO ,  Wenyuan BAI

IPC: G06F9/48 ,  G06F9/50

CPC classification number: G06F9/4881 ,  G06F9/5022

Abstract: The present disclosure discloses Spark collaborative computing, job method and apparatus for multiple K8s clusters, and addresses the problem that most of the current multiple K8s clusters adopt the model of federated clusters, and Spark's own method of scheduling and optimization cannot be implemented across domains, by implementing a cross-domain job method, setting the multiple K8s clusters as a master cluster and a slave cluster, with the master cluster being responsible for creating Spark's The master cluster is responsible for creating Spark's Driver container and Pods, and the slave cluster is responsible for creating Spark's Executor container and Pods. After the containers are created, a direct tunnel is established between the master cluster and the slave cluster by aggregating address information and access credentials through the Collaboration Center, and the containers in the slave cluster register with Driver and continuously send heartbeat messages through the tunnel.

公开(公告)号：US20230041862A1

公开(公告)日：2023-02-09

申请号：US17885556

申请日：2022-08-11

Applicant: Zhejiang Lab

Inventor： Feng GAO ,  Wenyuan BAI

IPC: H04L9/32 ,  H04L9/30 ,  H04L9/00

Abstract: The present invention discloses a cloud-side collaborative multi-mode private data circulation method based on a smart contract, including: S1, a system is initialized; S2, the original data are encrypted into private data, an encryption certificate z′ for storage is generated, and z′ includes metadata and a data certificate key′; S3, the DO calls a smart contract program to realize uplink of the encryption certificate z′ and releases z′ to a block chain through a smart contract, wherein the smart contract is open to all user accounts; S4, rapid data circulation is realized: when DO releases the data certificate, DU has been identified, a DU's account IDDU is set through an access policy, the DU obtains an encryption key for data access by executing a smart contract and a key algorithm, private data are obtained through metadata and decrypted to obtain a plaintext; and S5, the data circulation is confirmed.