Update app.py

app.py

@@ -2,13 +2,9 @@ import gradio as gr
 from duckduckgo_search import DDGS
 from typing import List, Dict
 import os
-from langchain.llms import HuggingFaceHub
-from langchain.chains import ConversationChain
-from langchain.memory import ConversationBufferMemory
-from langchain.prompts import PromptTemplate
 
 # Environment variables and configurations
-os.environ["HUGGINGFACEHUB_API_TOKEN"] = os.environ.get("HUGGINGFACE_TOKEN")  # Replace with your actual HuggingFace API token
+huggingface_token = os.environ.get("HUGGINGFACE_TOKEN")
 
 MODELS = [
     "mistralai/Mistral-7B-Instruct-v0.3",
@@ -26,7 +22,7 @@ def get_web_search_results(query: str, max_results: int = 10) -> List[Dict[str,
         print(f"An error occurred during web search: {str(e)}")
         return [{"error": f"An error occurred during web search: {str(e)}"}]
 
-def summarize_results(query: str, search_results: List[Dict[str, str]]) -> str:
+def summarize_results(query: str, search_results: List[Dict[str, str]], model: str) -> str:
     try:
         context = "\n\n".join([f"Title: {result['title']}\nContent: {result['body']}" for result in search_results])
         
@@ -44,44 +40,21 @@ def summarize_results(query: str, search_results: List[Dict[str, str]]) -> str:
     except Exception as e:
         return f"An error occurred during summarization: {str(e)}"
 
-# Initialize LangChain components
-llm = HuggingFaceHub(repo_id="mistralai/Mistral-7B-Instruct-v0.3", model_kwargs={"temperature": 0.7, "max_length": 512})
-
-template = """You are an AI assistant that helps with web searches and summarizes information. 
-Use the provided web search results to answer questions.
-
-Current conversation:
-{history}
-Human: {input}
-AI Assistant:"""
-
-prompt = PromptTemplate(input_variables=["history", "input"], template=template)
-
-conversation = ConversationChain(
-    llm=llm,
-    memory=ConversationBufferMemory(ai_prefix="AI Assistant"),
-    prompt=prompt,
-    verbose=True
-)
-
 def respond(message, chat_history, model, temperature, num_api_calls):
-    search_results = get_web_search_results(message)
-    
-    if not search_results:
-        response = f"No search results found for the query: {message}"
-    elif "error" in search_results[0]:
-        response = search_results[0]["error"]
-    else:
-        summary = summarize_results(message, search_results)
-        
-        # Use LangChain to generate a response based on the summary
-        langchain_response = conversation.predict(input=f"Based on this summary, answer the user's question: {summary}\n\nUser's question: {message}")
+    final_summary = ""
+    for _ in range(num_api_calls):
+        search_results = get_web_search_results(message)
         
-        response = langchain_response
+        if not search_results:
+            final_summary += f"No search results found for the query: {message}\n\n"
+        elif "error" in search_results[0]:
+            final_summary += search_results[0]["error"] + "\n\n"
+        else:
+            summary = summarize_results(message, search_results, model)
+            final_summary += summary + "\n\n"
 
-    return response
+    return final_summary if final_summary else "Unable to generate a response. Please try a different query."
 
-# The rest of your Gradio setup remains the same
 css = """
 Your custom CSS here
 """
@@ -110,7 +83,7 @@ theme = gr.themes.Soft(
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Dropdown(choices=MODELS, label="Select Model", value=MODELS[0]),
+        gr.Dropdown(choices=MODELS, label="Select Model", value=MODELS[2]),
         gr.Slider(minimum=0.1, maximum=1.0, value=0.2, step=0.1, label="Temperature"),
         gr.Slider(minimum=1, maximum=5, value=1, step=1, label="Number of API Calls")
     ],
@@ -134,4 +107,4 @@ demo = gr.ChatInterface(
     )
 )
 
-demo.launch()
+demo.launch()