Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

0.7.1 release #3234

Merged
merged 1 commit into from
Jul 29, 2023
Merged

0.7.1 release #3234

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 29 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,35 @@ vNext
-
-

0.7.1
-----
Breaking changes:
- Migrating Flax from returning FrozenDicts to returning regular dicts. More details can be found in this [announcement](https://github.com/google/flax/discussions/3191)

New features:
- Use pyink
- added dict migration guide to index
- add scan over layers section
- Expose options to customize rich.Table
- add support for initializing carry variables in scan
- Let Flax-Orbax to not port the shape of `target` arrays when they port the `target` shardings.

Bug fixes:
- Use import `orbax.checkpoint` which is a better import pattern.
- Use import `orbax.checkpoint as ocp` to avoid the verbosity of using 'orbax.checkpoint` every time.
- [linen] Add alternative, more numerically stable, variance calculation to `LayerNorm`.
- [linen] Minor cleanup to normalization code.
- Fix norm calculation bug for 0-rank arrays.
- [JAX] Remove references to jax.config.jax_array.
- [linen] Use `stack` instead of `concatenate` in `compute_stats`, to handle scalar stats case.
- [linen] More minor cleanup in normalization `compute_stats`.
- Fix warnings from atari gym.
- Refactor TypeHandler to operate over batches of values, rather than individual ones. This allows more flexibility for implementations that may operate more efficiently on batches.
- Fix carry slice logic
- make flax_basics guide use utility fns
- Fix checkpointing guide error at head
- Improve scan docs

0.7.0
-----
- RNNCellBase refactor.
Expand Down
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -197,7 +197,7 @@ To cite this repository:
author = {Jonathan Heek and Anselm Levskaya and Avital Oliver and Marvin Ritter and Bertrand Rondepierre and Andreas Steiner and Marc van {Z}ee},
title = {{F}lax: A neural network library and ecosystem for {JAX}},
url = {http://github.com/google/flax},
version = {0.7.0},
version = {0.7.1},
year = {2023},
}
```
Expand Down
2 changes: 1 addition & 1 deletion flax/version.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,4 +13,4 @@
# limitations under the License.

"""Current Flax version at head on Github."""
__version__ = "0.7.0"
__version__ = "0.7.1"
Loading