/ github.com / gke-labs / eval-runner

pr-391-b2-base Completed

← Back to Runs
Repository
https://github.com/gke-labs/eval-runner.git
Commit
0a3e6b0
Model
google/gemini-3.5-flash-lite
Age
4h
Tasks
1/1 (1 ok, 0 fail)
Total Samples
8 (samples 9-16)
Mean Success Rate
0.12
Total Cost
$0.0037
Tokens Consumed
2322
Latency (mean of per-task p50 / p95)
3.27s / 4.25s

Eval Tasks

eval phase rate 95% CI (0-1 axis) n method log path
docs/readme_comprehension_blind Completed 0.12 [0.02, 0.47]
1/8 (samples 9-16) wilson gs://gkelabs-evalrunner-dev-eval-runner/runs/0a3e6b01043cf8c6eaebcb676001a692ab71c350/docs/readme_comprehension_blind/2026-10-27T15-47-17-00-00_readme-comprehension-blind_jckuZue53pdFHPmtjEutTh.json ↗