Update app.py

app.py

@@ -1,5 +1,6 @@
 import os
 import streamlit as st
+from dotenv import load_dotenv
 from transformers import pipeline
 from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain_community.llms import llamacpp
@@ -14,6 +15,7 @@ from langchain.prompts import PromptTemplate
 from langchain.chains import create_history_aware_retriever, create_retrieval_chain
 from langchain.vectorstores import Chroma
 from utills import load_txt_documents , split_docs,  load_uploaded_documents,retriever_from_chroma
+from langchain.text_splitter import TokenTextSplitter,RecursiveCharacterTextSplitter
 
 
 script_dir = os.path.dirname(os.path.abspath(__file__))
@@ -32,66 +34,93 @@ hf = HuggingFaceEmbeddings(
     model_kwargs=model_kwargs,
     encode_kwargs=encode_kwargs
 )
+def get_vectorstore(text_chunks):
+    model_name = "sentence-transformers/all-mpnet-base-v2"
+    model_kwargs = {'device': 'cpu'}
+    encode_kwargs = {'normalize_embeddings': True}
+    hf = HuggingFaceEmbeddings(
+        model_name=model_name,
+        model_kwargs=model_kwargs,
+        encode_kwargs=encode_kwargs
+)
+
+    vectorstore = Chroma.from_documents(texts=text_chunks, embedding=hf,persist_directory="docs/chroma/")
+    return vectorstore
 
 
 
 
+def get_pdf_text(pdf_docs):
+    text = ""
+    for pdf in pdf_docs:
+        pdf_reader = PdfReader(pdf)
+        for page in pdf_reader.pages:
+            text += page.extract_text()
+    return text
 
-documents = load_txt_documents(data_path)
-docs = split_docs(documents, 450, 20)
 
+def get_text_chunks(text):
+    text_splitter = RecursiveCharacterTextSplitter.from_tiktoken_encoder(
+        separator="\n",
+        chunk_size=1000,
+        chunk_overlap=200,
+        length_function=len
+    )
+    chunks = text_splitter.split_text(text)
+    return chunks
 
 
-retriever = retriever_from_chroma(docs, hf, "mmr", 6)
 
+def create_conversational_rag_chain(vectorstore):
+    retriever = retriever_from_chroma(docs, hf, "mmr", 6)
 
-callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
+    callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
 
-llm = llamacpp.LlamaCpp(
-            model_path="/kaggle/working/phi-2-layla-v1-chatml-Q8_0.gguf",
-            n_gpu_layers=1,
-            temperature=0.1,
-            top_p = 0.9,
-            n_ctx=22000,
-            max_tokens=200,
-            repeat_penalty=1.7,
-            callback_manager = callback_manager,
-            verbose=False,
+    llm = LlamaCpp(
+                model_path="/kaggle/working/phi-2-layla-v1-chatml-Q8_0.gguf",
+                n_gpu_layers=1,
+                temperature=0.1,
+                top_p=0.9,
+                n_ctx=22000,
+                max_tokens=200,
+                repeat_penalty=1.7,
+                callback_manager=callback_manager,
+                verbose=False,
             )
 
-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
-which can be understood without the chat history. Do NOT answer the question,
-just reformulate it if needed and otherwise return it as is."""
+    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
+    which can be understood without the chat history. Do NOT answer the question,
+    just reformulate it if needed and otherwise return it as is."""
 
-ha_retriever = history_aware_retriever(llm, retriever, contextualize_q_system_prompt)
+    ha_retriever = history_aware_retriever(llm, retriever, contextualize_q_system_prompt)
 
-qa_system_prompt = """You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, just say that you don't know. Be as informative as possible, be polite and formal.\n{context}"""
+    qa_system_prompt = """You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, just say that you don't know. Be as informative as possible, be polite and formal.\n{context}"""
 
-qa_prompt = ChatPromptTemplate.from_messages(
-    [
-        ("system", qa_system_prompt),
-        MessagesPlaceholder("chat_history"),
-        ("human", "{input}"),
-    ]
-)
-
-question_answer_chain = create_stuff_documents_chain(llm, qa_prompt)
+    qa_prompt = ChatPromptTemplate.from_messages(
+        [
+            ("system", qa_system_prompt),
+            MessagesPlaceholder("chat_history"),
+            ("human", "{input}"),
+        ]
+    )
 
+    question_answer_chain = create_stuff_documents_chain(llm, qa_prompt)
 
-rag_chain = create_retrieval_chain(ha_retriever, question_answer_chain)
-msgs = StreamlitChatMessageHistory(key="special_app_key")
+    rag_chain = create_retrieval_chain(ha_retriever, question_answer_chain)
+    msgs = StreamlitChatMessageHistory(key="special_app_key")
 
-@st.cache_resource
-def create_conversational_rag_chain():
-    return RunnableWithMessageHistory(
-        rag_chain,
-        lambda session_id: msgs,
-        input_messages_key="input",
-        history_messages_key="chat_history",
-        output_messages_key="answer",
-    )
+    @st.cache_resource
+    def create_rag_chain():
+        return RunnableWithMessageHistory(
+            rag_chain,
+            lambda session_id: msgs,
+            input_messages_key="input",
+            history_messages_key="chat_history",
+            output_messages_key="answer",
+        )
 
+    return create_rag_chain()
 conversational_rag_chain = create_conversational_rag_chain()
 
 def display_chat_history(chat_history):
@@ -102,6 +131,7 @@ def display_chat_history(chat_history):
 
 
 def main_page(conversational_rag_chain):
+    load_dotenv()
     """Main page for the Streamlit app."""
     msgs = st.session_state.get("chat_history", StreamlitChatMessageHistory(key="special_app_key"))
     chain_with_history = conversational_rag_chain