Update backend/query_llm.py

backend/query_llm.py

@@ -5,7 +5,7 @@ import gradio as gr
 
 from os import getenv
 from typing import Any, Dict, Generator, List
-
+from gradio_client import Client
 from huggingface_hub import InferenceClient
 from transformers import AutoTokenizer
 
@@ -83,8 +83,24 @@ def generate_hf(prompt: str, history: str, temperature: float = 0.5, max_new_tok
     formatted_prompt = format_prompt(prompt, "hf")
 
     try:
-        stream = hf_client.text_generation(formatted_prompt, **generate_kwargs,
-                                            stream=True, details=True, return_full_text=False)
+        # stream = hf_client.text_generation(formatted_prompt, **generate_kwargs,
+        #                                     stream=True, details=True, return_full_text=False)
+        client_Qwen = Client("Qwen/Qwen1.5-110B-Chat-demo")
+        response = client_Qwen.predict(
+                    query=prompt,
+                    history=[],
+                    system="You are a helpful assistant.",
+                    api_name="/model_chat"
+                )
+                 # Extract the API output text
+        api_output = response[1] if response and len(response) > 1 else "No output received from the API."
+        
+                #chatbot_responses.append((txt, api_output[0][1]))
+        print(response)
+        
+        # Print the generated code
+        print(api_output[0][1])
+        stream=api_output[0][1]
         output = ""
         for response in stream:
             output += response.token.text