首页|Patent Application Titled 'Fly Parameter Compression And Decompression To Facili tate Forward And/Or Back Propagation At Clients During Federated Learning' Publi shed Online (USPTO 20240371362)
Patent Application Titled 'Fly Parameter Compression And Decompression To Facili tate Forward And/Or Back Propagation At Clients During Federated Learning' Publi shed Online (USPTO 20240371362)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Reporters obtained the following quote from the background information supplied by the inventors:“Federated learning of machine learning (ML) model(s) is a tec hnique for training ML model(s) in whichan on-device ML model is stored locally on a client device of a user, and a global ML model, that is aremote-based cou nterpart of the on-device ML model, is stored remotely at a remote system (e.g., aremote server or a cluster of remote servers). The client device, using the o n-device ML model, canprocess user input detected at the client device to gener ate predicted output, and can generate a gradientbased on the predicted output in a supervised or unsupervised manner.
NETL
NSTL
