Update README.md

README.md

@@ -75,8 +75,8 @@ and is comparable with Mistral-7B-Instruct-v0.1 on MMLU and MT-Bench in English.
 |                                                     |TC, Knowledge |TC, Reasoning|TC, Chat Behavior  |EN, Knowledge|EN, Chat Behavior|
 |                                                     | 5 shot       | 3 shot    | 0 shot              | 5 shot     | 0 shot           |
 | MediaTek-Research/Breeze-7B-Instruct-v0.1           |              |           |     5.7             |            |                  |
-| yentinglin/Taiwan-LLM-7B-v2.1-chat                  | 27.25        |           |     4.2             | 26.47      |                  |
-| yentinglin/Taiwan-LLM-13B-v2.0-chat                 | 31.92        |           |     5.0             |            |                  |
+| yentinglin/Taiwan-LLM-7B-v2.1-chat                  | 27.25        |           |     4.2             | 26.47      |     -*           |
+| yentinglin/Taiwan-LLM-13B-v2.0-chat                 | 31.92        |           |     5.0             |            |     -*           |
 | 01-ai/Yi-6B-Chat                                    | 29.28        |           |     5.0             |            |                  |
 | 01-ai/Yi-34B-Chat                                   | 54.11        |           |     6.9             |            |                  |
 | Qwen/Qwen-7B-Chat                                   | 28.30        |           |     5.4             |            |                  |
@@ -95,6 +95,9 @@ and is comparable with Mistral-7B-Instruct-v0.1 on MMLU and MT-Bench in English.
 | Qwen/Qwen-14B-Chat                                  | 46.51        | 58.20          | 51.12      | 49.38      |
 | gpt-3.5-turbo-1106                                  |              |                |            |            |
 
+
+* Taiwan-LLM models responds to multi-turn questions (English) in Traditional Chinese.
+
 ## Use in Transformers
 
 First install direct dependencies: