Skip to content

Releases: genlm/llamppl

v0.2.0: First release of llamppl

10 Apr 04:32
Compare
Choose a tag to compare

This is the first release of the project under its new name: llamppl.

What's Changed

The package was previously published as hfppl.

This release is functionally identical to the final version of hfppl.

No breaking changes — existing users can migrate with minimal effort

# uninstall the old name (optional but recommended)
pip uninstall hfppl

# install the renamed package
pip install llamppl

Import path:

# OLD
import hfppl

# NEW
import llamppl

Full Changelog: v0.1.2...v0.2.0

v0.1.2: Final release under the name `hfppl`

10 Apr 03:14
d558aa9
Compare
Choose a tag to compare

This is the final release of the package under the name hfppl. This project has been renamed and will continue development under the new name: llamppl.

Minimal migration effort required — the API and functionality are identical for now. The only change is the package name.

Please update your dependencies and import paths accordingly:

# OLD
pip install hfppl

# NEW
pip install llamppl
# OLD
import hfppl

# NEW
import llamppl

Full Changelog: v0.1.1...v0.1.2

v0.1.1

03 Apr 20:11
c5e9747
Compare
Choose a tag to compare

What's Changed

Full Changelog: v0.1.0...v0.1.1

v0.1.0, first release!

18 Mar 10:55
fe65965
Compare
Choose a tag to compare

What's Changed

  • Fix bug in getting started example by @gabegrand in #3
  • Fix #4 by @alex-lew in #5
  • Add grammar-constrained SMC to examples by @gabegrand in #8
  • Migrate dependency management to poetry by @gabegrand in #9
  • First pass at integrating Jacob Hoover's and Maddy Bowers's visualization code by @alex-lew in #14
  • Use BitsAndBytesConfig instead of load_in_8bit and token instead of u… by @alex-lew in #15
  • Prevent future merge conflicts by @gabegrand in #17
  • Additional submodels and masks by @gabegrand in #16
  • Run pre-commit on all files by @gabegrand in #18
  • Renormalize LMNextToken.sample() probs to fix floating point errors by @gabegrand in #20
  • Fix ValueError for LMTokenMask.log_prob() when the mask has no support by @gabegrand in #21
  • Integrate AsyncLM models from genlm-backend by @benlebrun in #24
  • Raise a NullMask() error for degenerate token masks by @gabegrand in #27
  • Run tests in CI, add code coverage by @sritchie in #28
  • run pre-commit on all files, convert to ruff by @sritchie in #29
  • Add release workflow, dynamic versioning by @sritchie in #30

New Contributors

Full Changelog: https://github.com/probcomp/hfppl/commits/v0.1.0