Overview
Who this competition fits
适合已有中文机器阅读理解经验、能使用 PaddlePaddle 2.0 及以上版本训练端到端模型的个人参赛者。尤其适合希望学习模型可解释性,并能处理答案抽取与 token 级证据输出的人。
Read the original official blurb
本次比赛聚焦于阅读理解任务的可解释评测,要求模型根据问题从文本中找出预测的答案以及预测依赖的证据。 Official tags: 阅读理解,可解释性. Sign-ups: 867.
Resources
What this competition gives you
There is a baseline you can run as-is — starting from it is usually far faster than building from scratch.
Preparation
From registration to a first submission
- 01
PaddlePaddle 2.0+
- 02
机器阅读理解
- 03
中文文本分词与 token 对齐
- 04
证据抽取(rationale)
- 05
TrustAI 解释方法
- 06
TSV/ZIP 提交文件生成
Before you commit: 难点不只是答对阅读理解答案,还要从给定的 sent_token 中按重要度顺序输出支持答案的 token id 序列;证据既要贴近人工标注,也要在扰动样本下保持一致。提交还必须同时满足指定的三列文本格式和 ZIP 文件命名要求。
Source
How this page was assembled
Competition information is structured from the official page. Scores are platform estimates for decision support; official rules take precedence.
- Official competition page
- Baidu AI Studio
- Last checked
- Not recorded