Patent · US Active

Gradient normalization systems and methods for adaptive loss balancing in deep multitask networks

US11537895B2 · kind B2 · utility

1Cited by
40References
27Claims
0Family size

Assignee

Inventors

Key dates

Filing dateOct 24, 2018
Grant dateDec 27, 2022
Priority date
Expiry dateNov 16, 2040

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N20/00
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Systems and methods for training a multitask network is disclosed. In one aspect, training the multitask network includes determining a gradient norm of a single-task loss adjusted by a task weight for each task, with respect to network weights of the multitask network, and a relative training rate for the task based on the single-task loss for the task. Subsequently, a gradient loss function, comprising a difference between (1) the determined gradient norm for each task and (2) a corresponding target gradient norm, can be determined. An updated task weight for the task can be determined and used in the next iteration of training the multitask network, using a gradient of the gradient loss function with respect to the task weight for the task.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.