update description

Signed-off-by: Zhang Jun <[email protected]>

README.md

@@ -1,5 +1,5 @@
 ---
-title: Paddlepaddle Meets Llm
+title: Paddlepaddle Meets LLM
 emoji: 💬
 colorFrom: yellow
 colorTo: purple

app.py

@@ -17,10 +17,14 @@ def inference(input_text):
 
 
 title = 'PaddlePaddle Meets LLM'
-description = 'What is special: underlying execution is using PaddlePaddle and PaddleNLP!'
-article = "<p style='text-align: center'> PaddleNLP <a href='https://github.com/PaddlePaddle/PaddleNLP'>Github Repo</a></p>"
+description = '''
+- The underlying execution framework is based on [PaddlePaddle](https://github.com/PaddlePaddle/Paddle) and [PaddleNLP](https://github.com/PaddlePaddle/PaddleNLP).
+- PaddleNLP supports a wide range of open-source LLMs. Check out the [full model list](https://github.com/PaddlePaddle/PaddleNLP?tab=readme-ov-file#%E6%A8%A1%E5%9E%8B%E6%94%AF%E6%8C%81).
+- We chose QWEN2-0.5B as the model for this use case due to limited computational resources.
+- [ERNIE 4.5](https://yiyan.baidu.com/) was trained with PaddlePaddle, [give it a try](https://huggingface.co/spaces/PaddlePaddle/ernie_demo)!
+'''
 
-examples = ['请自我介绍一下。', '今天吃什么好呢？']
+examples = ['请自我介绍一下。']
 
 demo = gr.Interface(
     inference,
@@ -28,7 +32,6 @@ demo = gr.Interface(
     outputs="text",
     title=title,
     description=description,
-    article=article,
     examples=examples,
     )