Hacktoberfest 2026:维护者为十月标记出来的 issue,仍然开放、适合新手。 浏览 Hacktoberfest issue

Scorer-eval GUI drops the coverage fields the docs say to compare

未关闭
#3,325 2 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 2 天内回复

还没有人认领这个 Issue。

评估

难度
3/5
预计耗时
1-2 天
新手友好度
72/100
Issue 类型
功能
描述清晰度
描述清楚
活跃度
活跃
技术栈
python, react, typescript
领域
ai, backend, frontend

调研方向

Start with scenario_progress_read_model.py around lines 600–612 and ObjectiveScorerDetails.tsx; then check the scorer-evaluation docs at doc/code/scoring/4_scorer_metrics.py to confirm how coverage is defined. Carry num_responses and num_input_responses through the read model and show coverage with accuracy in scorer details. Done when the GUI exposes coverage so users can distinguish abstention from a perfect score.

由索引模型根据 Issue 内容生成。

描述

Scorer-eval GUI drops the coverage fields the docs say to compare

Summary

The scorer-evaluation docs (doc/code/scoring/4_scorer_metrics.py) correctly tell readers to compare num_responses against num_input_responses to see coverage, because responses with no verdict are excluded from accuracy. The exclusion itself is intended and test-pinned.

What breaks the guidance: the GUI backend read model (scenario_progress_read_model.py:600-612) drops both fields at the API boundary, and ObjectiveScorerDetails.tsx renders the masked accuracy with no way to surface a caveat.

Demonstration (real ObjectiveScorerEvaluator.evaluate_dataset_async, 10-item dataset): a scorer returning no verdict on the 3 hard items reports accuracy 1.0000 / F1 1.0000 at 7/10 coverage (identical to a perfect scorer) while an honest scorer attempting all 10 at 70% correct reports 0.7000. Abstention is strictly rewarded. The GUI cannot show why.

Suggested direction: carry num_responses / num_input_responses through the read model and render coverage alongside accuracy in the scorer-details view.

主要语言
Python
星标
4.6k
派生
962
平均合并
2 天 15 小时
30 天内合并 PR
390

环境准备

在 Codespaces 中打开

在浏览器里用你自己的 GitHub 账号启动这个项目的开发容器。

  • 没有 Dockerfile 或 Docker Compose 文件
  • 有 Pull Request 模板
  • 没有贡献指南

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

microsoft/PyRIT 的其他 Issue

查看 microsoft/PyRIT 的全部 Issue

相似的 Issue

更多 Python Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。