Releases: instadeepai/mlipaudit
Releases · instadeepai/mlipaudit
Release list
v0.1.4
v0.1.3
What's Changed
- ci: pin third-party actions to commit SHAs so workflow can start by @lwalew in #130
- ci: enforce develop-only PRs into main and update PR template by @lwalew in #128
- ci: add Claude Code Action workflows for reviews and @claude mentions by @lwalew in #129
- feat: pass charge and multiplicity to models by @lwehrhan in #127
- feat: all neutral protein systems for scaling benchmark by @lwehrhan in #135
- feat: update mlipaudit for mlip 0.2.0 (deps, cuda extra, esen) by @lwalew in #138
Full Changelog: v0.1.2...v0.1.3
v0.1.2
What's Changed
- chore: move magic number by @lwalew in #119
- chore: update ci docker images to slim from slim-bullseye by @lwalew in #120
- ci: run on develop by @lwalew in #121
- fix: add missing continue to ScalingBenchmark by @lwalew in #125
- feat: load models without stress by @lwalew in #124
- feat: release 0.1.2 by @lwalew in #126
Full Changelog: v0.1.1...v0.1.2
v0.1.1
v0.1.0
Initial release with the following features:
- Adding benchmark implementations for 15 benchmarks.
- Adding CLI tool that can (a) run benchmarks and (b) display a GUI for visualization of benchmark results.
- Adding code for public leaderboard page only displayed on the HuggingFace leaderboard of MLIP models.
- Adding extensive code documentation with tutorials.
- Adding extensive unit test coverage.