pySigLib -- Fast Signature-Based Computations on CPU and GPU
This work addresses a bottleneck for researchers and practitioners in fields like quantitative finance who need efficient signature-based computations on large datasets.
The authors tackled the scalability issue of signature-based methods for sequential data by developing pySigLib, a high-performance library that accelerates signature and signature kernel computations on CPU and GPU, achieving faster gradients with a novel differentiation scheme.
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.