feat: add centralized eval queue with submitit backend#33
Open
BobbyZhouZijian wants to merge 1 commit intomainfrom
Open
feat: add centralized eval queue with submitit backend#33BobbyZhouZijian wants to merge 1 commit intomainfrom
BobbyZhouZijian wants to merge 1 commit intomainfrom
Conversation
Replace direct in-process grading with a manager-based task queue. Agents submit eval requests to .coral/queue/pending/ and poll for results, while the manager's QueueDispatcher dispatches grader jobs via submitit (LocalExecutor or SlurmExecutor). - Add coral/queue/ package (types, client, dispatcher, worker, counter) - Add QueueConfig nested under GraderConfig with strategy (fair/fifo/ priority), rate limiting, max concurrency, and executor selection - Queue timeout defaults to grader.timeout + 60s overhead - Replace non-atomic eval counter with fcntl.flock-based counter - Remove direct multiprocessing grading path from post_commit.py - Integrate dispatcher into AgentManager.monitor_loop() - Add 22 tests covering types, counter, client, scheduling, config Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Closes #21
.coral/queue/pending/as JSON, poll.coral/queue/results/for grading resultsQueueDispatcherpicks up pending requests, applies scheduling policy (fair/fifo/priority), and submits grader jobs via submitit (LocalExecutororSlurmExecutor)grader.queue— timeout defaults tograder.timeout + 60sfcntl.flock()replaces the racy_increment_eval_count()Test plan
uv run pytest tests/test_queue.py -v— 22 queue-specific tests pass (types, counter concurrency, client submit/poll/timeout, dispatcher scheduling strategies, rate limiting, config integration)uv run pytest tests/ -v— 101 tests pass, no regressionsuv run ruff check— lint clean on all changed filescoral start -c examples/kernel_builder/task.yaml agents.count=3— agents queue and take turns grading🤖 Generated with Claude Code