riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 33 Packages Projects Releases Wiki Activity

Add flax/jax quickstart (#12342)

Browse Source
This commit is contained in:
Marc van Zee 2021-06-24 18:04:18 +02:00 committed by GitHub
parent 5b1b5635d3
commit f2c4ce7e33
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
examples/research_projects/jax-projects/README.md
View File

@ -120,7 +120,9 @@ TODO (should be filled by 24.06.)...
## Quickstart flax and jax
TODO (should be filled by 25.06.)...
[JAX](https://jax.readthedocs.io/en/latest/index.html) is Autograd and XLA, brought together for high-performance numerical computing and machine learning research. It provides composable transformations of Python+NumPy programs: differentiate, vectorize, parallelize, Just-In-Time compile to GPU/TPU, and more. A great place for getting started with JAX is the [JAX 101 Tutorial](https://jax.readthedocs.io/en/latest/jax-101/index.html).
[Flax](https://flax.readthedocs.io/en/latest/index.html) is a high-performance neural network library designed for flexibility built on top of JAX. It aims to provide users with full control of their training code and is carefully designed to work well with JAX transformations such as `grad` and `pmap` (see the [Flax philosophy](https://flax.readthedocs.io/en/latest/philosophy.html)). For an introduction to Flax see the [Flax Basics Colab](https://flax.readthedocs.io/en/latest/notebooks/flax_basics.html) or the list of curated [Flax examples](https://flax.readthedocs.io/en/latest/examples.html).
## Quickstart flax and jax in transformers