Releases: probabilists/lampe
LAMPE v0.8.0
💥 Breaking news
The required version of Zuko has been upgraded to 0.2.1 to follow API changes. Accordingly, the support for older PyTorch versions (1.11 and earlier) was dropped. PyTorch 1.12 has been around for about a year and is compatible with all CUDA drivers since 10.2.
✨ What's new
- Migrate the documentation to lampe.readthedocs.io (1cf58ef)
- Simplify HDF5 dataset (3b52b13)
- Add in-memory dataset (f1a003a)
- Add the contrastive neural ratio estimation (CNRE) inference algorithm (#7) by @bkmi
- Add the flow matching posterior estimation (FMPE) inference algorithm (#12)
Full Changelog: v0.7.0...v0.8.0
LAMPE v0.7.0
✨ What's new
- Drop support for standardization with respect to the first and second moments (d6fdca7)
- New neural score estimation (NSE) inference algorithm (#7) with the help of @glouppe
🐛 Bug fixes
- Fix incorrect grid shape in
utils.gridapply
for one-dimensional space (#8)
Full Changelog: v0.6.0...v0.7.0
LAMPE v0.6.0
💥 Breaking news
In the v0.4.2 release, LAMPE dropped the nflows
package for built-in normalizing flows. Since then, the flow implementations have improved and new ones have been added, making the lampe.nn.flows
module a respectable normalizing flow collection. In this context, we decided to export said module into a standalone package called Zuko. Doing so, we wish to dissociate the addition/development of simulation-based inference algorithms from the addition of flow architectures.
📝 Documentation
- Use
autosummary
for automatic API generation (822c3cc)
Full Changelog: v0.5.5...v0.6.0
LAMPE v0.5.5
✨ What's new
Full Changelog: v0.5.2...v0.5.5
LAMPE v0.5.2
LAMPE v0.5.0
LAMPE v0.4.4
✨ What's new
- Neural autoregressive flow (NAF) implementation (81a22d3)
Full Changelog: v0.4.2...v0.4.4
LAMPE v0.4.2
✨ What's new
- Move inference components to
lampe.inference
module (257ab9e) - Drop
nflows
dependency for built-in normalizing flows (07f96c3)
🐛 Bug fixes
- Fix missing imports (40f6a94)
Full Changelog: v0.3.0...v0.4.2