Update README.md

README.md

@@ -17,8 +17,8 @@ DiscoX is a benchmark for the evaluation of  LLMs on discourse- and expert-level
 - **Languages**: English ⇄ Chinese (100 English→Chinese tasks, 100 Chinese→English tasks)
 - **Total samples**: 200 discourse- and exprt-level translation items
 - **Average passage length**: ~1.7k characters (min 0.73k, max 3.04k)
-- **Meta fields**: primary & secondary domain labels, structured rubrics, prompt IDs,etc.
-- **Reference Rubrics**: every task ships with multiple rubrics annotated by experts, capturing key points for evaluating translation quality.
+- **Meta fields**: primary & secondary domain labels, structured rubrics, prompt IDs,etc
+- **Reference Rubrics**: every task ships with multiple rubrics annotated by experts, capturing key points for evaluating translation quality
 Primary domain coverage:
 | Primary Domain | Samples | Share |
 | --- | --- | --- |