Repository navigation
chore(release): v0.4.2 - #180
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 29d5d51589
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| saved_model_path("vector_leaves.bin").map_or_else( | ||
| |_| "tests/data/saved".to_owned(), | ||
| |path| { | ||
| path.strip_prefix(env!("CARGO_MANIFEST_DIR")) | ||
| .unwrap_or(&path) | ||
| .display() |
There was a problem hiding this comment.
When the baseline and candidate are built in worktrees whose newest saved-model directories differ (for example, v0.4.1 versus v0.4.2), this emits different source strings. scripts/bench_cpu_round3.py compares the complete --list catalogs at lines 97–99 before honoring --cases, so every comparison aborts—even one selecting only synthetic cases. Keep the catalog label stable and record the resolved fixture path separately if needed.
Useful? React with 👍 / 👎.
| let Ok(version) = name | ||
| .split('.') | ||
| .map(str::parse::<u64>) | ||
| .collect::<Result<Vec<_>, _>>() |
There was a problem hiding this comment.
For a supported prerelease such as 0.4.3-rc.1, parsing every dot-separated component as u64 rejects the current release directory, so the benchmark silently selects an older stable fixture (or reports that none exists) instead of the newest model. The release process explicitly supports prerelease versions and saves their models under the full version string, so this lookup needs SemVer-aware ordering.
AGENTS.md reference: AGENTS.md:L680-L684
Useful? React with 👍 / 👎.
1 participant
Release bump
Updated hessboost to
0.4.2, refreshed Rust/Python lockfiles, and saved this version's native models.After merge, on
main:./release.py --dry-runthen./release.py.