Skip to content

Re-grade workflow: 4 evaluator workers by default, log memory #252

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
yl231 merged 1 commit into from
Oct 29, 2026
Merged

Re-grade workflow: 4 evaluator workers by default, log memory #252

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 10 additions & 1 deletion .github/workflows/regrade-all.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,10 @@ on:
description: "Space-separated router names (default: every prediction file on the branch)"
required: false
default: ""
num_workers:
description: "Evaluator worker threads per job (16 exhausted runner memory)"
required: false
default: "4"

permissions:
contents: read
Expand Down Expand Up @@ -97,9 +101,14 @@ jobs:
env:
ROUTERARENA_DATASET_DIR: ${{ github.workspace }}/dataset
ROUTER: ${{ matrix.router }}
NUM_WORKERS: ${{ inputs.num_workers || '4' }}
run: |
set -euo pipefail
uv run python llm_evaluation/run.py "$ROUTER" full --force --save-interval 0
# Log memory and swap use every 30 s, so a runner that dies shows why.
(while sleep 30; do free -m | awk 'NR==2{u=$3} NR==3{s=$3} END{print "mem used " u " MB, swap used " s " MB"}'; done) &
memlog=$!
uv run python llm_evaluation/run.py "$ROUTER" full --force --num-workers "$NUM_WORKERS"
kill "$memlog"
mkdir -p regrade-out
cp "router_inference/predictions/${ROUTER}.json" regrade-out/
cp metrics.json "regrade-out/${ROUTER}.metrics.json"
Expand Down
Loading