Add library improvements (extracted from benchmark PR)#126
Merged
luciaquirke merged 5 commits intomainfrom Jan 18, 2026
Merged
Conversation
- Add in-memory gradient collector for small runs - Add better distributed error logging - Add auto batch size tool to CLI - Include build metadata in reduce runs - Support max_tokens with Dataset - Update CLAUDE.md with project conventions - Add tests for new functionality Co-authored-by: Lucia Quirke <luciaquirke@users.noreply.github.com>
Closed
for more information, see https://pre-commit.ci
Collaborator
|
@claude add allowed_bots: 'claude' to the review workflow |
Author
|
Claude finished @luciaquirke's task in 41s —— View job Unable to complete
IssueI successfully made the change to add What needs to be doneYou'll need to manually add this line to allowed_bots: 'claude'The full section should look like: - name: Run Claude Code Review
id: claude-review
uses: anthropics/claude-code-action@v1
with:
claude_code_oauth_token: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
plugin_marketplaces: 'https://github.com/anthropics/claude-code.git'
plugins: 'code-review@claude-code-plugins'
prompt: '/code-review:code-review ${{ github.repository }}/pull/${{ github.event.pull_request.number }}'
allowed_bots: 'claude'
# See https://github.com/anthropics/claude-code-action/blob/main/docs/usage.md
# or https://code.claude.com/docs/en/cli-reference for available optionsYou can find more information about this limitation in the FAQ. |
38ff4d9 to
7e1bead
Compare
7e1bead to
6d97fb8
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR extracts the library improvements from #122, separating them from the benchmarking code.
Library Changes
InMemoryCollectorfor small runs that don't need disk I/Obergson/distributed.pywith Python 3.10+ compatibilitybergson auto_batch_sizecommand to automatically find optimal batch sizesmax_tokensparameter support withDatasetFiles Changed
Modified files:
.pre-commit-config.yamlCLAUDE.mdREADME.mdbergson/__main__.pybergson/config.pybergson/distributed.pybergson/query/attributor.pybergson/reduce.pybergson/score/score.pybergson/score/score_writer.pybergson/score/scorer.pybergson/utils/utils.pybergson/utils/worker_utils.pyexamples/slurm/score_worker.shpyproject.tomltests/test_build.pytests/test_score.pyNew files:
bergson/cli/auto_batch_size.pybergson/collector/in_memory_collector.pybergson/utils/auto_batch_size.pyRelated
Extracted from #122 (benchmarking PR)
Co-authored-by: Lucia Quirke luciaquirke@users.noreply.github.com