pySigLib -- Fast Signature-Based Computations on CPU and GPU

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shmelev, Daniil, Salvi, Cristopher
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918141069623296
author Shmelev, Daniil
Salvi, Cristopher
author_facet Shmelev, Daniil
Salvi, Cristopher
contents Signature-based methods have recently gained significant traction in machine learning for sequential data. In particular, signature kernels have emerged as powerful discriminators and training losses for generative models on time-series, notably in quantitative finance. However, existing implementations do not scale to the dataset sizes and sequence lengths encountered in practice. We present pySigLib, a high-performance Python library offering optimised implementations of signatures and signature kernels on CPU and GPU, fully compatible with PyTorch's automatic differentiation. Beyond an efficient software stack for large-scale signature-based computation, we introduce a novel differentiation scheme for signature kernels that delivers accurate gradients at a fraction of the runtime of existing libraries.
format Preprint
id arxiv_https___arxiv_org_abs_2509_10613
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle pySigLib -- Fast Signature-Based Computations on CPU and GPU
Shmelev, Daniil
Salvi, Cristopher
Machine Learning
Mathematical Software
60L10, 65Y05, 68T99
Signature-based methods have recently gained significant traction in machine learning for sequential data. In particular, signature kernels have emerged as powerful discriminators and training losses for generative models on time-series, notably in quantitative finance. However, existing implementations do not scale to the dataset sizes and sequence lengths encountered in practice. We present pySigLib, a high-performance Python library offering optimised implementations of signatures and signature kernels on CPU and GPU, fully compatible with PyTorch's automatic differentiation. Beyond an efficient software stack for large-scale signature-based computation, we introduce a novel differentiation scheme for signature kernels that delivers accurate gradients at a fraction of the runtime of existing libraries.
title pySigLib -- Fast Signature-Based Computations on CPU and GPU
topic Machine Learning
Mathematical Software
60L10, 65Y05, 68T99
url https://arxiv.org/abs/2509.10613