当前位置:
X-MOL 学术
›
arXiv.cs.MS
›
论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Lineax: unified linear solves and linear least-squares in JAX and Equinox
arXiv - CS - Mathematical Software Pub Date : 2023-11-28 , DOI: arxiv-2311.17283 Jason Rader, Terry Lyons, Patrick Kidger
arXiv - CS - Mathematical Software Pub Date : 2023-11-28 , DOI: arxiv-2311.17283 Jason Rader, Terry Lyons, Patrick Kidger
We introduce Lineax, a library bringing linear solves and linear
least-squares to the JAX+Equinox scientific computing ecosystem. Lineax uses
general linear operators, and unifies linear solves and least-squares into a
single, autodifferentiable API. Solvers and operators are user-extensible,
without requiring the user to implement any custom derivative rules to get
differentiability. Lineax is available at https://github.com/google/lineax.
中文翻译:
Lineax:JAX 和 Equinox 中的统一线性求解和线性最小二乘法
我们介绍 Lineax,这是一个为 JAX+Equinox 科学计算生态系统带来线性求解和线性最小二乘的库。Lineax 使用通用线性运算符,并将线性求解和最小二乘统一为单个可自动微分的 API。求解器和运算符是用户可扩展的,无需用户实现任何自定义导数规则即可获得可微性。Lineax 可在 https://github.com/google/lineax 上获取。
更新日期:2023-11-30
中文翻译:
Lineax:JAX 和 Equinox 中的统一线性求解和线性最小二乘法
我们介绍 Lineax,这是一个为 JAX+Equinox 科学计算生态系统带来线性求解和线性最小二乘的库。Lineax 使用通用线性运算符,并将线性求解和最小二乘统一为单个可自动微分的 API。求解器和运算符是用户可扩展的,无需用户实现任何自定义导数规则即可获得可微性。Lineax 可在 https://github.com/google/lineax 上获取。