Graphcore Limited

Smart Impulse

Antoine Labatie

Carlo Luschi

Dominic Alexander Masters

Zachary Frank Eaton-Rosen

A computer-implemented method of training a deep neural network, comprising, for each of one or more batches of training examples: processing the data in a forward pass through the layers of the network, by: applying a set of network weights to the input data to obtain a set of weighted inputs, normalising the weighted inputs based on statistics computed for each training example, transforming the normalised inputs by affine transformation parameters, applying an activation function to the transformed normalised inputs to obtain post-activation values, and normalizing the post-activation values based on one or more proxy variables sampled from a distribution defined by proxy distribution parameters, the normalization applied independently of training example; processing the data in a backward pass through the network to determine updates to learnable parameters comprising network weights, affine transformation parameters, and proxy distribution parameters, and updating the learnable parameters to optimise a predefined loss function.

Processing data batches in a multi-layer network

A system for analyzing electrical consumption of a site equipped with a plurality electrical equipment. The system comprises at least one sensor for periodic measurement of global consumption of the site by measuring current and voltage signals on at least one of the electrical phases upstream of the supply network for the site. The system comprises a processor for decomposing the global consumption of the site to individual consumption of each kind of electrical equipment based on a set of signatures with each signature has power load coefficients representing power loads of the corresponding electrical equipment. The system comprises a processor for optimizing the set of signatures and power load coefficients in comparing a calculated recomposition of the individual consumption with the measured global consumption to have a smallest difference of the comparison result, while respecting physical constraints of the set of signatures and the power load coefficients.

Growing community of inventors

Paris, France

Antoine Labatie