Releases: shaham-lab/SpectralNet
Releases · shaham-lab/SpectralNet
v1.0.1
v1.0.1 (2026-03-07)
Bug Fixes
- Add pypi environment to release job for OIDC trusted publishing (
288c464)
Detailed Changes: v1.0.0...v1.0.1
v1.0.0
v1.0.0 (2026-03-07)
Bug Fixes
- Correct four bugs in core clustering and utility code (
778fb05)
Chores
- Add pixi environment, semantic-release config, and dev extras (
3ace05e)
Continuous Integration
- Add GitHub Actions CI and semantic-release workflows (
c7786fc)
Documentation
-
Add pixi development setup instructions to README (
01cde74) -
Update README with Dataset-based large-scale usage pattern (
d2a1e8b)
Features
- Dataset-based interface for true out-of-core scalability (
9586b6a)
Performance Improvements
- Chunked inference in predict/embed and fix get_random_batch (
081ec04)
Refactoring
- Clean up imports, remove dead code, fix docstring (
ebffe72)
Testing
- Add comprehensive test suite (74 tests, 0 warnings) (
e66ed7f)
Detailed Changes: v0.1.2...v1.0.0
SpectralNet 0.1.2
Fixed some installation issues
SpectralNet 0.1.1
Added the option to save the weights of the AE and the Siamese networks
SpectralNet 0.1.0
Mainly bug fixes.
SpectralNet 0.0.8
Happy to announce the release of the 0.0.8 version.
We added an implementation of another kernel function which is called multiquadratic kernel.
You can use it instead of the RBF kernel when computing the affinity matrix.
SpectralNet 0.0.6
Happy to announce the release of the 0.0.6 version.
We added some use-case examples and some bug fixes.