Burn Release 0.9.0: Introducing The Burn Book 🔥, Model Repository, and Framework Improvements

Thanks to the work of its dedicated contributors, the Burn Deep Learning Framework's latest release introduces several key updates, which I am happy to announce:

1. Burn Book: We've created a new, comprehensive guide, which we simply call The Burn Book 🔥. It is available at https://burn-rs.github.io/book/. It targets both beginner and expert users, offering insights from the basic training-to-inference workflow to more advanced topics such as backend extensions and custom kernels. Expect continuous updates.
2. Model Repository: Discover the Burn Model Repository at https://github.com/burn-rs/models. It houses official Burn team-verified models and links to external community projects.
3. Framework Enhancements: We've enhanced Burn's core features with new optimizers, training metrics, tensor operations, and improved ONNX support. We are stunned by the community's engagement and quality contributions! Over 20 individual contributors in the last month only, thank you so much 🙏
4. Backend Improvements: We have done optimizations and additions to some backends, most notably with autotuning in the WGPU backend and preliminary support for the new Candle ML framework.

Check out the release notes for more details: https://github.com/burn-rs/burn/releases/tag/v0.9.0

Let us know what you think of the book and how we may improve it further.