Repository navigation
docs: roadmap — simplify each component before the job loop - #294
Merged
Merged
Conversation
Records the plan: three behavior-preserving PRs before step 4 (delete legacy leftovers; generation takes GenerationConfig; judging takes JudgingConfig), then split out each unit of work as part of the loop PRs, one command at a time. Waits for the step-3 stack (#289-#293). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Oct 27, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Docs only: records in
docs/roadmap.mdthe plan to simplify generation and judging. The work waits for the step-3 stack (#289, #290, #291, #293) to merge, because it rewrites the same files.The problem. Each component reaches its work through layers that existed for the legacy scripts:
vera_cli→run_for_user_models→run_generation→ConversationRunner→ its pool → one conversationvera_cli→run_judging→judge_conversations→batch_evaluate_with_individual_judges→ its pool → one evaluationScoring is already one layer (
run_scoring) and is the model.The target comes from #266 and is the same everywhere:
<component>/run.pyholds onerun_<verb>(config)entry point, a module holds one unit of work, plusutils.pyandviz.py.Two phases:
Before the loop, three PRs that leave behavior unchanged:
GenerationConfigdirectly. This is the Model spec boundary phase.JudgingConfigdirectly.None of these depend on the loop design, and they shrink step 4.
With the loop (step 4): splitting out "one conversation" and "one evaluation" is the same work as defining docs: design draft — one job loop for generate and judge #266's jobs, so it happens in the loop PRs. Judge goes first, then generate, then the pipeline, and each PR deletes that component's
runner.py.Why not all at once: one PR that changes structure and concurrency together would be the hardest to review, and the hardest to bisect if scores moved.
I didn't renumber the step table, so this won't conflict with #289's edits to those rows on
release/2.1.🤖 Generated with Claude Code