Update README.md
Browse files
README.md
CHANGED
|
@@ -2,8 +2,6 @@
|
|
| 2 |
|
| 3 |
# CL²GEC:A Multi-Discipline Benchmark for Continual Learning in Chinese Literature Grammatical Error Correction
|
| 4 |
|
| 5 |
-
## Dataset Summary
|
| 6 |
-
|
| 7 |
**CL²GEC** is a benchmark for **Chinese grammatical error correction (GEC)** in **scholarly writing** with a **continual-learning** protocol. The corpus covers **10 first-level disciplines** (Law, Management, Education, Economics, Natural Sciences, History, Agricultural Sciences, Literature, Arts, Philosophy). Each sample contains an errorful sentence (`source`) and one or more corrected references (`references`). Standard **train / validation / test** splits are provided and may be used **per-discipline** to study sequential/continual learning behavior such as forgetting and transfer.
|
| 8 |
|
| 9 |
---
|
|
|
|
| 2 |
|
| 3 |
# CL²GEC:A Multi-Discipline Benchmark for Continual Learning in Chinese Literature Grammatical Error Correction
|
| 4 |
|
|
|
|
|
|
|
| 5 |
**CL²GEC** is a benchmark for **Chinese grammatical error correction (GEC)** in **scholarly writing** with a **continual-learning** protocol. The corpus covers **10 first-level disciplines** (Law, Management, Education, Economics, Natural Sciences, History, Agricultural Sciences, Literature, Arts, Philosophy). Each sample contains an errorful sentence (`source`) and one or more corrected references (`references`). Standard **train / validation / test** splits are provided and may be used **per-discipline** to study sequential/continual learning behavior such as forgetting and transfer.
|
| 6 |
|
| 7 |
---
|