Skip to content

Commit 23a25f5

Browse files
committed
chore: restore e2e-test.yml for demo
Made-with: Cursor
1 parent 87542d9 commit 23a25f5

File tree

1 file changed

+40
-0
lines changed

1 file changed

+40
-0
lines changed

.github/workflows/e2e-test.yml

Lines changed: 40 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,40 @@
1+
name: VerifyWise LLM Evaluation
2+
3+
on:
4+
workflow_dispatch:
5+
6+
jobs:
7+
evaluate:
8+
name: VerifyWise LLM Evaluation (gpt-4o-mini)
9+
runs-on: ubuntu-latest
10+
steps:
11+
- uses: actions/checkout@v4
12+
13+
- name: Run evaluation
14+
id: eval
15+
uses: ./
16+
with:
17+
api_url: https://29ae92ae4bd295.lhr.life
18+
project_id: project_20260403_144244_140917
19+
dataset_id: "52"
20+
metrics: answer_relevancy,bias
21+
model_name: gpt-4o-mini
22+
model_provider: openai
23+
threshold: "0.5"
24+
timeout_minutes: "5"
25+
poll_interval_seconds: "10"
26+
experiment_name: "GH-Actions-E2E-Test"
27+
fail_on_threshold: "false"
28+
vw_api_token: ${{ secrets.VW_API_TOKEN }}
29+
llm_api_key: ${{ secrets.LLM_API_KEY }}
30+
# judge_api_key not set — defaults to llm_api_key
31+
32+
- name: Show results
33+
if: always()
34+
run: |
35+
echo "Passed: ${{ steps.eval.outputs.passed }}"
36+
echo "Experiment: ${{ steps.eval.outputs.experiment_id }}"
37+
if [ -f "${{ steps.eval.outputs.results_path }}" ]; then
38+
echo "--- Results ---"
39+
cat "${{ steps.eval.outputs.results_path }}"
40+
fi

0 commit comments

Comments
 (0)