Update app.py

app.py

@@ -6,9 +6,34 @@ from threading import Thread
 from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
 
 
-TITLE = "Chat with Llama3-8B-Chinese"
+TITLE = "<center>Chat with Llama3-8B-Chinese</center>"
+
+DESCRIPTION = "<h3><center>Visit <a href='https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat' target='_blank'>our model page</a> for details.</center></h3>"
+
+TOOL_EXAMPLE = '''You have access to the following tools:
+```python
+def google_search(keywords: List[str]):
+    """
+    Search on the Internet based on the keywords.
+    
+    Args:
+        keywords (List[str]): Keywords for the search engine.
+    """
+    pass
+```
+
+Write "Action:" followed by a list of actions in JSON that you want to call, e.g.
+Action:
+```json
+[
+    {
+        "name": "tool name (one of [google_search])",
+        "arguments": "the input to the tool"
+    }
+]
+```
+'''
 
-DESCRIPTION = "Visit <a href='https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat' target='_blank'>our model page</a> for details."
 
 tokenizer = AutoTokenizer.from_pretrained("shenzhi-wang/Llama3-8B-Chinese-Chat")
 model = AutoModelForCausalLM.from_pretrained("shenzhi-wang/Llama3-8B-Chinese-Chat", device_map="auto")
@@ -40,7 +65,7 @@ def stream_chat(message: str, history: list, system: str, temperature: float, ma
 
     output = ""
     for new_token in streamer:
-        outputs += new_token
+        output += new_token
         yield output
 
 
@@ -73,8 +98,10 @@ with gr.Blocks(fill_height=True) as demo:
             ),
         ],
         examples=[
-            ["我的蓝牙耳机坏了，我该去看牙科还是耳鼻喉科？"],
-            ["今日行军进展如何", "扮演诸葛亮和我对话。"],
+            ["我的蓝牙耳机坏了，我该去看牙科还是耳鼻喉科？", "You are a helpful assistant."],
+            ["在一道没有余数的除法算式里，被除数(不为零)加上除数和商的积，再除以被除数，所得的商是多少？", "You are a helpful assistant."],
+            ["今日行军进展如何？", "扮演诸葛亮和我对话。"],
+            ["羊驼的寿命有多久？", TOOL_EXAMPLE],
         ],
         cache_examples=False,
         title=TITLE,