Update README.md

README.md

@@ -53,6 +53,8 @@ and is comparable with Mistral-7B-Instruct-v0.1 on MMLU and MT-Bench in English.
 
 
 ## Inference Performance
+In this test, we use the first 1500 characters of one of 201802最高法院民事裁判書 as input and ask the model to rewrite the article.
+The max_new_tokens is set to 1000 (except Qwen/Qwen-7B, which is set to 400). All models were inferenced with `vllm` on 2 A6000 (TP=2 ).
 
 | Models                                                             | Speed (char/sec)  |Estimated Max Input Length (TC Char)|
 |--------------------------------------------------------------------|-------------------|--------------------------|
@@ -65,11 +67,6 @@ and is comparable with Mistral-7B-Instruct-v0.1 on MMLU and MT-Bench in English.
 | Taiwan-LLM-7B-v2.1-base                                 |   16.23           |    2.6k                  |
 | Yi-34B                                                       |   15.18           |    4.4k                  |
 
-\*Test Input: 201802最高法院民事裁判書(first 1500 characters)  
-\*Test Prompt: 請重新寫出這篇文章  
-\*Output Max Token: 1000  (For Qwen/Qwen-7B, max_new_tokens is set to 400.)   
-\*Using vllm, TP=2 on 2 A6000  
-
 
 ## Chat Model Performance