-
公开(公告)号：US20240281641A1
公开(公告)日：2024-08-22
申请号：US18653096
申请日：2024-05-02
Inventor： Weimin Zhou , Yifeng Liu , Yi Li , Zonghong Dai
IPC: G06N3/045
CPC classification number: G06N3/045
Abstract: A model weight obtaining method includes obtaining structure information of a first neural network model; searching, based on the structure information of the first neural network model, a weight library that stores a plurality of groups of historical weights to obtain a reference weight, where the reference weight is a weight of a second neural network model having a structure similar to that of the first neural network model in the plurality of groups of historical weights; and converting the reference weight to obtain a weight of the first neural network model. In the method, a weight of a neural network model having a structure similar to that of a to-be-trained neural network model is searched for in a weight library, and the weight is converted, to obtain a weight that can be used by the to-be-trained neural network model.
-
公开(公告)号：US20240354580A1
公开(公告)日：2024-10-24
申请号：US18758605
申请日：2024-06-28
Inventor： Zhipeng Liang , Weimin Zhou , Yi Li , Zonghong Dai
IPC: G06N3/086 , G06N3/0475
CPC classification number: G06N3/086 , G06N3/0475
Abstract: A neural network architecture search method includes: receiving an optimization request, where the optimization request includes a model file and an optimization requirement of a to-be-optimized model, and the optimization requirement includes a performance requirement and a hardware requirement; performing neural architecture search processing in search space based on the model file, to obtain a neural network architecture that meets the optimization requirement; and returning the neural network architecture.
-
公开(公告)号：US12032571B2
公开(公告)日：2024-07-09
申请号：US17694970
申请日：2022-03-15
Inventor： Kaiyuan Xie , Xiaolong Bai , Wenqi Fu , Chenghui Yu , Weimin Zhou
IPC: G06F7/00 , G06F16/2453 , G06N20/00
CPC classification number: G06F16/2453 , G06N20/00
Abstract: In a method for AI model optimization, an optimization device receives an original AI model and search configuration information that comprises a plurality of search items each indicating its search categories for performing optimization information search on the original AI model. The device obtains a plurality of search operators corresponding to the plurality of search items, and arranges the search operators in an operation sequence based on the search configuration information. The device then executes the search operators in the arranged operation sequence on the original AI model to obtain an optimized AI model. In the execution of the operation sequence, each search operator, except for the first search operator in the operation sequence, is executed utilizing operation results of a preceding search operator in the operation sequence, the operation results including generated network structures and search space information.
-
公开(公告)号：US20240095529A1
公开(公告)日：2024-03-21
申请号：US18521152
申请日：2023-11-28
Inventor： Weimin Zhou , Yuting Mai , Yi Li , Yijun Guo , Binbin Deng , Zonghong Dai
IPC: G06N3/08
CPC classification number: G06N3/08
Abstract: A neural network optimization method includes receiving a model file of a to-be-optimized neural network; obtaining a search space of a target neural network architecture based on the model file of the to-be-optimized neural network, where the search space includes a value range of each attribute of each neuron in the target neural network architecture; obtaining the target neural network architecture based on the search space; training the target neural network architecture based on the model file of the to-be-optimized neural network, to obtain a model file of a target neural network; and providing the model file of the target neural network to a user.
-
公开(公告)号：US20220414426A1
公开(公告)日：2022-12-29
申请号：US17902206
申请日：2022-09-02
Inventor： Weimin Zhou , Yijun Guo , Yi Li , Yuting Mai , Binbin Deng
Abstract: This application provides a neural architecture search method, applied to a search system. The search system includes a generator and a searcher. The method includes: The generator generates a plurality of neural network architectures based on a search space; the searcher obtains evaluation indicator values of a plurality of child models obtained based on the plurality of neural network architectures on first hardware; and the searcher determines, based on the neural network architectures corresponding to the plurality of child models and the evaluation indicator values of the plurality of child models on the first hardware, a first target neural network architecture that meets a preset condition. In this way, different initial child model training processes are decoupled, and a neural architecture search process is decoupled from an initial child model training process, so that search duration is reduced and search efficiency is improved.
-
公开(公告)号：US20220197901A1
公开(公告)日：2022-06-23
申请号：US17694970
申请日：2022-03-15
Inventor： Kaiyuan Xie , Xiaolong Bai , Wenqi Fu , Chenghui Yu , Weimin Zhou
IPC: G06F16/2453 , G06N20/00
Abstract: In a method for AI model optimization, an optimization device receives an original AI model and search configuration information that comprises a plurality of search items each indicating its search categories for performing optimization information search on the original AI model. The device obtains a plurality of search operators corresponding to the plurality of search items, and arranges the search operators in an operation sequence based on the search configuration information. The device then executes the search operators in the arranged operation sequence on the original AI model to obtain an optimized AI model. In the execution of the operation sequence, each search operator, except for the first search operator in the operation sequence, is executed utilizing operation results of a preceding search operator in the operation sequence, the operation results including generated network structures and search space information.
