Update chatbot_backend.py

chatbot_backend.py

@@ -1,49 +1,41 @@
 import os
-import asyncio
+from langchain_community.llms import Ollama
+from langchain_community.embeddings import OllamaEmbeddings
 from langchain_community.document_loaders import TextLoader
-from langchain.text_splitter import CharacterTextSplitter
-from langchain_google_genai import ChatGoogleGenerativeAI, GoogleGenerativeAIEmbeddings
-from langchain.vectorstores import FAISS
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+from langchain.memory import ConversationBufferMemory
 from langchain.chains import ConversationalRetrievalChain
 
-google_api_key = os.getenv("GOOGLE_API_KEY")
+# ✅ Memory
+memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
 
-# 🟢 Event loop safe embeddings initializer
-def get_embeddings():
-    try:
-        asyncio.get_running_loop()
-    except RuntimeError:
-        loop = asyncio.new_event_loop()
-        asyncio.set_event_loop(loop)
-
-    return GoogleGenerativeAIEmbeddings(
-        model="models/embedding-001",
-        google_api_key=google_api_key
-    )
+# ✅ LLM (Ollama model, e.g. llama3, mistral, phi3)
+llm = Ollama(model="llama3")
 
-# 🟢 Use loader safely
-loader = TextLoader("data.txt")
+# ✅ Load + Split Docs
+loader = TextLoader("data.txt", encoding="utf-8")
 docs = loader.load()
-
-# 🟢 Split text into chunks
-text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
-documents = text_splitter.split_documents(docs)
-
-# 🟢 Create vectorstore with embeddings
-embeddings = get_embeddings()
-db = FAISS.from_documents(documents, embeddings)
-
-# 🟢 Conversational chain
-qa = ConversationalRetrievalChain.from_llm(
-    ChatGoogleGenerativeAI(model="gemini-1.5-pro", google_api_key=google_api_key),
-    db.as_retriever()
+splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
+chunks = splitter.split_documents(docs)
+
+# ✅ Embeddings + VectorStore
+embeddings = OllamaEmbeddings(model="llama3")
+vectorstore = FAISS.from_documents(documents=chunks, embedding=embeddings)
+retriever = vectorstore.as_retriever(search_kwargs={"k": 2})
+
+# ✅ Conversational Retrieval Chain (with memory)
+conversational_chain = ConversationalRetrievalChain.from_llm(
+    llm=llm,
+    retriever=retriever,
+    memory=memory,
+    return_source_documents=False
 )
 
-# 🟢 Function to interact with bot
-chat_history = []
-
-def ask_bot(query: str):
-    global chat_history
-    result = qa({"question": query, "chat_history": chat_history})
-    chat_history.append((query, result["answer"]))
-    return result["answer"]
+# ✅ Ask bot function
+def ask_bot(query: str) -> str:
+    try:
+        response = conversational_chain.invoke({"question": query})
+        return response["answer"]
+    except Exception as e:
+        return f"Error: {str(e)}"