made google work, added llama

model.py

@@ -2,11 +2,11 @@ from dotenv import load_dotenv, find_dotenv
 import os
 from langchain_google_genai import ChatGoogleGenerativeAI
 from langchain_huggingface import HuggingFaceEndpoint, ChatHuggingFace
-from langchain_community.tools import DuckDuckGoSearchRun
+from langchain_community.tools.google_search.tool import GoogleSearchAPIWrapper
 from langchain_tavily import TavilySearch
 from langchain_community.document_loaders import AsyncHtmlLoader
 from langchain.tools import tool
-from langchain.prompts import ChatPromptTemplate
+from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
 from langchain.agents import AgentExecutor, create_tool_calling_agent
 from csv_cache import CSVSCache
 from prompt import get_prompt
@@ -64,15 +64,18 @@ class LLMProvider(Enum):
     corresponding environment variable names for API keys.
     """
     HUGGINGFACE = ("HuggingFace", "HF_TOKEN")
+    HUGGINGFACE_LLAMA = ("HUGGINGFACE_LLAMA", "HF_TOKEN")
     GOOGLE_GEMINI = ("Google Gemini", "GOOGLE_API_KEY")
 
 class Model:
-    def __init__(self):        
-        #load_dotenv(find_dotenv())
-        self.token = os.getenv("HF_TOKEN")
+    def __init__(self, provider: LLMProvider = LLMProvider.HUGGINGFACE):        
+        load_dotenv(find_dotenv())
         self.system_prompt = get_prompt()
         print(f"system_prompt: {self.system_prompt}")
+        self.provider = provider
         self.agent_executor = self.setup_model()
+        
+
 
     def get_answer(self, question: str) -> str:
         try:
@@ -107,10 +110,19 @@ class Model:
         if provider == LLMProvider.HUGGINGFACE:
             llm = HuggingFaceEndpoint(
             repo_id="Qwen/Qwen3-Next-80B-A3B-Thinking",
-            huggingfacehub_api_token=self.token,
+            huggingfacehub_api_token=api_token,
+            temperature=0
+        )
+            return ChatHuggingFace(llm=llm).bind_tools(tools)
+        
+        if provider == LLMProvider.HUGGINGFACE_LLAMA:
+            llm = HuggingFaceEndpoint(
+            repo_id="meta-llama/Llama-2-7b-chat-hf",
+            huggingfacehub_api_token=api_token,
             temperature=0
         )
             return ChatHuggingFace(llm=llm).bind_tools(tools)
+           
             
         
         elif provider == LLMProvider.GOOGLE_GEMINI:
@@ -118,14 +130,17 @@ class Model:
                 model="gemini-2.5-flash",
                 temperature=0
             )
-            # Define the Google Search tool using the Gemini API's built-in tool
-            google_search_tool = Tool(
-                name="google_search",
-                description="Search Google for recent information.",
-                func=lambda query: chat.invoke(query, tools=[{"googleSearch": {}}]).content
-                )
-            tools = tools.append(google_search_tool)
-            return chat.bind(tools)
+            # search = GoogleSearchAPIWrapper()
+
+            # # Define the Google Search tool correctly
+            # google_search_tool = Tool(
+            #     name="Google Search",
+            #     description="Search Google for recent information.",
+            #     func=search.run, # Use the run method to execute the search directly
+            # )
+
+            # tools.append(google_search_tool)
+            return chat.bind_tools(tools)
     
         else:
             raise ValueError(f"Unknown LLM provider: {provider}")
@@ -148,14 +163,15 @@ class Model:
             tavily_search_tool,
             arxiv_search,
         ]
-        chat = self.get_chat_with_tools(LLMProvider.HUGGINGFACE, tools)
+        chat = self.get_chat_with_tools(self.provider, tools)
         
         # Create the ReAct prompt template
         prompt = ChatPromptTemplate.from_messages(
             [
                 ("system", self.system_prompt),  # Use the new, detailed ReAct prompt
-                ("placeholder", "{agent_scratchpad}"),
                 ("human", "{input}"),
+                MessagesPlaceholder(variable_name="agent_scratchpad"),
+
             ]
         )
 
@@ -193,10 +209,14 @@ def update_mode(model):
 
 def main():
     load_dotenv(find_dotenv())
-    model = Model()
+    csv = CSVSCache()
+    df = csv.get_all_entries()
+    model = Model(LLMProvider.HUGGINGFACE)
     #update_mode(model)
-    response = model.get_answer("Where were the Vietnamese specimens described by Kuznetzov in Nedoshivina's 2010 paper eventually deposited? Just give me the city name without abbreviations.")
-    print(f"the output is: {response}")
+    test_questions = [0, 6, 10, 12, 15]
+    for row in test_questions:
+        response = model.get_answer(df.iloc[row]['question'])
+        print(f"the output is: {response}")
 
 if __name__ == "__main__":
     main()