Доработка: 1. В vectorstore.py заменить импорт Ollama на from langchain

agent.py

@@ -0,0 +1,133 @@
+import os
+from dotenv import load_dotenv
+from langchain_ollama import OllamaLLM, OllamaEmbeddings
+from langchain_chroma import Chroma
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import DirectoryLoader, TextLoader, UnstructuredMarkdownLoader
+from langchain.tools import Tool
+from langchain_community.tools.tavily_search import TavilySearchResults
+from langchain.agents import initialize_agent, AgentType
+
+# Load environment variables
+load_dotenv()
+TAVILY_API_KEY = os.getenv("TAVILY_API_KEY")
+if not TAVILY_API_KEY:
+    raise ValueError("TAVILY_API_KEY not found in .env file")
+
+# Configuration
+PERSIST_DIRECTORY = "./chroma_db"
+DOCUMENTS_DIR = "./documents"
+EMBEDDING_MODEL = "nomic-embed-text"
+LLM_MODEL = "llama3"
+
+def setup_vectorstore():
+    """Initialize or load ChromaDB vectorstore with Ollama embeddings."""
+    embeddings = OllamaEmbeddings(model=EMBEDDING_MODEL)
+    
+    if os.path.exists(PERSIST_DIRECTORY) and os.listdir(PERSIST_DIRECTORY):
+        vectorstore = Chroma(
+            persist_directory=PERSIST_DIRECTORY,
+            embedding_function=embeddings
+        )
+    else:
+        # Load and process documents
+        loader = DirectoryLoader(
+            DOCUMENTS_DIR,
+            glob="**/*",
+            loader_cls=lambda path: TextLoader(path, encoding="utf-8") if path.endswith(".txt") 
+                                    else UnstructuredMarkdownLoader(path) if path.endswith(".md") 
+                                    else None,
+            show_progress=True,
+            use_multithreading=True
+        )
+        documents = loader.load()
+        
+        text_splitter = RecursiveCharacterTextSplitter(
+            chunk_size=1000,
+            chunk_overlap=200,
+            length_function=len
+        )
+        texts = text_splitter.split_documents(documents)
+        
+        vectorstore = Chroma.from_documents(
+            documents=texts,
+            embedding=embeddings,
+            persist_directory=PERSIST_DIRECTORY
+        )
+        vectorstore.persist()
+    
+    return vectorstore
+
+# Initialize vectorstore
+vectorstore = setup_vectorstore()
+
+# Define tools
+def search_local_kb(query: str) -> str:
+    """Search local knowledge base using ChromaDB."""
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 3})
+    docs = retriever.get_relevant_documents(query)
+    return "\n\n".join([doc.page_content for doc in docs])
+
+def web_search(query: str) -> str:
+    """Search the web using Tavily."""
+    search = TavilySearchResults(tavily_api_key=TAVILY_API_KEY, max_results=3)
+    results = search.run(query)
+    return "\n\n".join([result["content"] for result in results])
+
+# Create LangChain tools
+local_tool = Tool(
+    name="search_local_kb",
+    func=search_local_kb,
+    description="Useful for answering questions about local documents stored in the knowledge base."
+)
+
+web_tool = Tool(
+    name="web_search",
+    func=web_search,
+    description="Useful for answering questions about current events, news, or general knowledge from the internet."
+)
+
+tools = [local_tool, web_tool]
+
+# Initialize LLM
+llm = OllamaLLM(model=LLM_MODEL)
+
+# Create agent with routing logic
+agent = initialize_agent(
+    tools=tools,
+    llm=llm,
+    agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION,
+    verbose=True,
+    handle_parsing_errors=True,
+    system_message="""You are an AI agent that must choose between two tools:
+    1. search_local_kb: For questions about local documents (concepts, notes, stored information)
+    2. web_search: For questions requiring up-to-date information from the internet
+    
+    When answering, ALWAYS specify your information source:
+    - If using local knowledge base: [Source: chromadb]
+    - If using web search: [Source: tavily]
+    
+    Be concise and accurate in your responses."""
+)
+
+# Chat interface
+def main():
+    print("RAG Agent with ChromaDB and Tavily Web Search")
+    print("Type 'exit' to quit\n")
+    
+    while True:
+        query = input("Запрос: ").strip()
+        if query.lower() == "exit":
+            break
+        
+        if not query:
+            continue
+            
+        try:
+            response = agent.run(query)
+            print(f"\n{response}\n")
+        except Exception as e:
+            print(f"Error: {str(e)}\n")
+
+if __name__ == "__main__":
+    main()