TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Show HN: Equinox (1.3k stars), a JAX library for neural networks and sciML

1 点作者 patrickkidger超过 1 年前
Hey folks! I wanted to advertise Equinox -- my now-surprisingly-popular ( :D ) JAX library for numerical models. These days that often means &quot;neural networks&quot;, but I like to emphasise that this also includes ODEs&#x2F;SDEs&#x2F;linear solves, etc.<p>For those already using JAX, then Equinox is interesting because (a) it ships with a NN library, and (b) this is built around the idea that &quot;everything is a pytree&quot;, which makes things easy to reason about and easy to compose. Furthermore (c) Equinox offers advanced tools like true runtime errors, out-of-place pytree surgery, and checkpointed while loops, and AFAIK in the JAX ecosystem these are unique to Equinox.<p>For those most familiar with PyTorch: for many use cases (sciML in particular), JAX has a much stronger compiler, more advanced autodiff, etc. And whilst JAX itself is akin to the `torch` namespace, libraries like Equinox are then akin to the `torch.nn` namespace.<p>Because of its speed and features, right now JAX+Equinox is my favourite approach to numerical computing. So I&#x27;d love for some more people to try it. What do you think?

暂无评论

暂无评论