Update app.py

app.py

@@ -1,5 +1,3 @@
-
-
 import os
 import streamlit as st
 from transformers import pipeline
@@ -8,13 +6,12 @@ from langchain.prompts import PromptTemplate
 from langchain.chains.question_answering import load_qa_chain
 from langchain.vectorstores import Chroma
 from langchain.retrievers import mmr_retriever
-from utills import load_txt_documents , split_docs, chroma_db,
-
+from utills import load_txt_documents , split_docs, chroma_db, load_uploaded_documents
 
 # Initialize variables and paths
 script_dir = os.path.dirname(os.path.abspath(__file__))
 data_path = "./data/"
-model_path = os.path.join(script_dir, 'mistral-7b-v0.1-layla-v4-Q4_K_M.gguf.2')
+model_path = os.path.join(script_dir, 'qwen2-0_5b-instruct-q4_0.gguf')
 store = {}
 
 # Set up HuggingFace embeddings
@@ -42,22 +39,32 @@ def load_txt_documents(data_path):
             documents.extend(TextLoader(file_path).load())
     return documents
 
-documents = load_txt_documents(data_path)
-
-def split_docs(documents, chunk_size, overlap):
-    # Your implementation here
-    pass
+@st.cache_data
+def load_uploaded_documents(uploaded_files):
+    documents = []
+    for uploaded_file in uploaded_files:
+        content = uploaded_file.read().decode("utf-8")
+        documents.append({"content": content, "filename": uploaded_file.name})
+    return documents
 
-docs = split_docs(documents, 450, 20)
 
 
+documents = load_documents(data_path)
+docs = split_docs(documents, 450, 20)
 
-chroma_db = chroma_db(docs, hf)
+@st.cache_resource
+def create_chroma_db(docs, hf):
+    return Chroma(docs, hf)
 
+chroma_db = create_chroma_db(docs, hf)
 
-retriever = retriever_from_chroma(chroma_db,"mmr",6)
+@st.cache_resource
+def create_retriever(chroma_db):
+    return mmr_retriever(chroma_db, "mmr", 6)
 
+retriever = create_retriever(chroma_db)
 
+# Set up LlamaCpp model
 callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
 
 @st.cache_resource
@@ -75,7 +82,7 @@ def load_llm(model_path):
         verbose=False,
     )
 
-llm = load_llm()
+llm = load_llm(model_path)
 
 contextualize_q_system_prompt = """Given a context, chat history and the latest user question
 which maybe reference context in the chat history, formulate a standalone question
@@ -138,8 +145,8 @@ def display_documents(docs, on_click=None):
                 if st.button(f"Expand Article {i+1}"):
                     on_click(i)
 
-def main(conversational_rag_chain):
-    """Main function for the Streamlit app."""
+def main_page(conversational_rag_chain):
+    """Main page for the Streamlit app."""
     msgs = st.session_state.get("chat_history", StreamlitChatMessageHistory(key="special_app_key"))
     chain_with_history = conversational_rag_chain
 
@@ -164,5 +171,27 @@ def main(conversational_rag_chain):
 
     st.session_state["chat_history"] = msgs
 
+def upload_page():
+    """Page for uploading and viewing documents."""
+    st.title("Upload and Check Documents")
+
+    uploaded_files = st.file_uploader("Upload Text Files", type=["txt"], accept_multiple_files=True)
+
+    if uploaded_files:
+        documents = load_uploaded_documents(uploaded_files)
+        for document in documents:
+            st.write(f"**Filename: {document['filename']}**")
+            st.text(document['content'])
+
+def main():
+    """Main function for the Streamlit app with page navigation."""
+    st.sidebar.title("Navigation")
+    page = st.sidebar.radio("Go to", ["Chatbot", "Upload Documents"])
+
+    if page == "Chatbot":
+        main_page(conversational_rag_chain)
+    elif page == "Upload Documents":
+        upload_page()
+
 if __name__ == "__main__":
-    main(conversational_rag_chain)
+    main()