Update app.py

app.py

@@ -1,16 +1,13 @@
 import os
 import gradio as gr
-from http import HTTPStatus
 from typing import List, Optional, Tuple, Dict
 import dashscope
 from dashscope import Generation
 from dashscope.api_entities.dashscope_response import Role
-import requests  # <-- Add this line to import the requests library
-
 
 # Configuration
 default_system = 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.'
-dashscope.api_key = os.getenv('HF_TOKEN')  # Replace 'YOUR_API_TOKEN' with your actual API token.
+dashscope.api_key = os.getenv('HF_TOKEN')
 
 # Typing definitions
 History = List[Tuple[str, str]]
@@ -55,46 +52,37 @@ def model_chat(query: Optional[str], history: Optional[History], system: str) ->
         query = ''
     if history is None:
         history = []
-
-    # Ensure the query is clearly asking for numbers
-    if 'next numbers' in query or 'give me numbers after' in query:
-        query = "Please give me the next 10 numbers after 10, starting from 11."
     
     messages = history_to_messages(history, system)
-    messages.append({'role': 'user', 'content': query})
-    
-    payload = {"inputs": query, "parameters": {"max_new_tokens": 150}, "history": messages}
-    headers = {"Authorization": f"Bearer {os.getenv('HF_TOKEN')}"}
+    messages.append({'role': Role.USER, 'content': query})
     
     try:
-        response = requests.post(f"https://api-inference.huggingface.co/models/Qwen/Qwen2.5-72B-Instruct", 
-                                 json=payload, headers=headers)
+        gen = Generation.call(
+            model='qwen2.5-72b-instruct',
+            messages=messages,
+            result_format='message',
+            stream=True
+        )
         
-        if response.status_code == 200:
-            response_data = response.json()
-            
-            if isinstance(response_data, list):
-                response_text = response_data[0].get('generated_text', '')
+        for response in gen:
+            if response.status_code == 200:
+                response_text = response.output.choices[0].message.content
+                
+                # Log the chat to file
+                log_history_to_file(query, response_text)
+                
+                # Update history with the new assistant response
+                history.append((query, response_text))
+                return response_text, history, system
             else:
-                response_text = response_data.get('generated_text', '')
-            
-            # Log the chat to file
-            log_history_to_file(query, response_text)
-            
-            # Update history with the new assistant response and return it
-            history.append([query, response_text])
-            return response_text, history, system
-        else:
-            error_message = f"Error {response.status_code}: {response.json().get('error', response.text)}"
-            log_history_to_file(query, error_message)
-            return error_message, history, system
+                error_message = f"Error: {response.status_code} - {response.message}"
+                log_history_to_file(query, error_message)
+                return error_message, history, system
     except Exception as e:
         error_message = f"Exception: {str(e)}"
         log_history_to_file(query, error_message)
         return error_message, history, system
 
-
-
 # Gradio Interface Setup
 with gr.Blocks() as demo:
     gr.Markdown("<center><font size=8>Qwen2.5-72B-Instruct👾</center>")
@@ -113,7 +101,6 @@ with gr.Blocks() as demo:
         clear_history = gr.Button("🧹 Clear history")
         submit = gr.Button("🚀 Send")
 
-    # Link buttons to functions
     textbox.submit(model_chat,
                    inputs=[textbox, chatbot, system_state],
                    outputs=[textbox, chatbot, system_input],