Recent Releases of https://github.com/eserie/wax-ml
https://github.com/eserie/wax-ml - v0.6.4
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.6.3...v0.6.4
- Python
Published by eserie almost 3 years ago
https://github.com/eserie/wax-ml - v0.6.3
What's Changed
- fix vmap module by @eserie in https://github.com/eserie/wax-ml/pull/67
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.6.2...v0.6.3
- Python
Published by eserie over 3 years ago
https://github.com/eserie/wax-ml - v0.6.2
What's Changed
- Refactor vmapliftwith_state by @eserie in https://github.com/eserie/wax-ml/pull/66
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.6.0...v0.6.2
- Python
Published by eserie over 3 years ago
https://github.com/eserie/wax-ml - v0.6.1
What's Changed
- Refactor vmapliftwith_state by @eserie in https://github.com/eserie/wax-ml/pull/66
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.6.0...v0.6.1
- Python
Published by eserie over 3 years ago
https://github.com/eserie/wax-ml - v0.6.0
What's Changed
- replace jax.treemultimap (removed in JAX release v0.3.16) by jax.tr… by @eserie in https://github.com/eserie/wax-ml/pull/60
- Faster unroll: do not propagate params in scan state by @eserie in https://github.com/eserie/wax-ml/pull/61
- correct imports or tree utils to avoid FutureWarnings by @eserie in https://github.com/eserie/wax-ml/pull/63
- Remove dependency with eagerpy by @eserie in https://github.com/eserie/wax-ml/pull/62
- Docs by @eserie in https://github.com/eserie/wax-ml/pull/64
- add stateful.py module with vmapliftwithstate and unrolllift_with… by @eserie in https://github.com/eserie/wax-ml/pull/65
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.5.0...v0.6.0
- Python
Published by eserie over 3 years ago
https://github.com/eserie/wax-ml - v0.5.0
What's Changed
- Rework optimizers by @eserie in https://github.com/eserie/wax-ml/pull/56, https://github.com/eserie/wax-ml/pull/58
- work with Jax v0.3.10 by @eserie in https://github.com/eserie/wax-ml/pull/59
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.4.0...v0.5.0
- Python
Published by eserie almost 4 years ago
https://github.com/eserie/wax-ml - v0.4.0
What's Changed
- EWMA alignement with pandas and speedup by @eserie in https://github.com/eserie/wax-ml/pull/53
- make alpha named arg by @eserie in https://github.com/eserie/wax-ml/pull/54
- Align EWMCov and EWMVar with EWMA by @eserie in https://github.com/eserie/wax-ml/pull/55
- Prepare v0.4.0 by @eserie in https://github.com/eserie/wax-ml/pull/57
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.3.2...v0.4.0
- Python
Published by eserie almost 4 years ago
https://github.com/eserie/wax-ml - v0.3.2
What's Changed
- faster ewma when adjust=True by @eserie in https://github.com/eserie/wax-ml/pull/52
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.3.1...v0.3.2
- Python
Published by eserie almost 4 years ago
https://github.com/eserie/wax-ml - v0.3.1
What's Changed
- Fix rng split in unroll by @eserie in https://github.com/eserie/wax-ml/pull/51
- Python
Published by eserie about 4 years ago
https://github.com/eserie/wax-ml - Realease v0.3.0
What's Changed
- Renew the PRNG key in the unroll operations by @eserie in https://github.com/eserie/wax-ml/pull/45
- Rework notebook 05 by @eserie in https://github.com/eserie/wax-ml/pull/46
- Refactor online optimizer by @eserie in https://github.com/eserie/wax-ml/pull/47
- VMap module works in contexts without PRNG key by @eserie in https://github.com/eserie/wax-ml/pull/48
- OnlineOptimizer returns updated params by @eserie in https://github.com/eserie/wax-ml/pull/49
- Prepare release by @eserie in https://github.com/eserie/wax-ml/pull/50
Full Changelog: https://github.com/eserie/wax-ml/compare/v0.2.0...v0.3.0
- Python
Published by eserie about 4 years ago
https://github.com/eserie/wax-ml - Release v0.2.0
Version v0.2.0 includes new prediction tools for time series analysis and documentation for performing analysis of non-stationary ARMA models.
We have refactored the unroll module and provided a new transformation function: unroll_transform_with_state which behaves very similarly to the haiku transformation function: transform_with_state.
The format and stream modules have been redesigned accordingly.
You can consult the CHANGELOG for more details.
- Python
Published by eserie over 4 years ago