Patent · US Active

Compression of machine-learned models via entropy penalized weight reparameterization

US12265898B2 · kind B2 · utility

0Cited by
2References
20Claims
0Family size

Assignee

Inventors

Key dates

Filing dateJan 10, 2024
Grant dateApr 1, 2025
Priority date
Expiry dateJan 10, 2044

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N3/044
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.

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