Skip to content

Releases: AI-sandbox/neural-admixture

v1.6.7

29 Sep 17:38

Choose a tag to compare

Fix handling of missing data when reading VCF files.

v1.6.6

03 Sep 17:49

Choose a tag to compare

Fix the README and delete unnecessary code.

v1.6.5

02 Sep 17:46

Choose a tag to compare

  1. Fix inference bug that causes memory crash
  2. Report time correctly
  3. Control the number of threads

v1.6.4

26 Aug 17:07

Choose a tag to compare

Updates:

  1. Ensured GaussianMixture uses float64 for improved numerical stability.
  2. Removed outdated supervised flag from the README.

v1.6.3

27 May 11:25

Choose a tag to compare

What's Changed

  • Include cuda files

v1.6.2

27 May 10:51

Choose a tag to compare

What's Changed

  • MacOS compatibility.

v1.6.1

27 May 09:14
29d98c9

Choose a tag to compare

What's Changed

  • v 1.6.1 by @joansaurina in #36
  • New reader for bed files.
  • 2 bit format for nvidia-gpus speedup using cuda.
  • Mps support.
  • Efficient SVD initialization, log-likelihood and file reading using Cython.

Note: this release introduces breaking changes which might make models trained on prior versions not to work as expected. Please use a version previous to 1.6.1 if that's the case.

Full Changelog: v1.5.1...v1.6.1

v1.5.1

27 Feb 19:35
2b15a85

Choose a tag to compare

  • Bump dependencies

Full Changelog: v1.5.0...v1.5.1

v1.5.0

27 Feb 19:25

Choose a tag to compare

  • Neural network enhancement by changing the input to the PCA space and switching to RMSNorm.
  • Gaussian Mixture and consensus K-Means initialization
  • Improved OOM data handling
  • Multi-GPU training and inference

Note: these releases introduce breaking changes which might make models trained on prior versions not to work as expected. Please use a version previous to 1.5.0 if that's the case.
Full Changelog: v1.4.1...v1.5.0

Update dependencies

25 Jun 10:03

Choose a tag to compare

  • Update dependencies to avoid breaking changes and speedup data reading.
  • Drop columns containing only NaNs from the input matrix.