Evaluation Dataset
The following datasets are all transformed into standard Evaluation Prompts before evaluation.
Dataset1（CLCC）
Data description：
Chinese Linguistics & Cognition Challenge (CLCC) mainly consists of two parts:
CLCC-H (190): Sampling from open source data set and manual screening (manual evaluation required)
CLCC-H-v2.0：CLCC-H-v2.0 screened and modified the data on CLCC-H-v1.0, while adding a large number of questions formulated manually. The evaluation results were all generated by manual evaluation.
Dataset structure：
Data detail：
| KEYS | EXPLAIN |
|---|---|
| ID | question id |
| question | question |
| answer | reference answer |
| source | data source |
| ability | corresponding ability label |