Create app.py

adding first version

app.py

@@ -0,0 +1,94 @@
+import streamlit as st
+from langchain_community.llms import HuggingFaceEndpoint
+from langchain_core.runnables import RunnablePassthrough
+from langchain_core.output_parsers import StrOutputParser
+from langchain.prompts import ChatPromptTemplate
+from PyPDF2 import PdfReader
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+import os
+from langchain_community.vectorstores import Chroma
+from langchain.chains.question_answering import load_qa_chain
+from langchain.prompts import PromptTemplate
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_chroma import Chroma
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from transformers import pipeline
+
+def get_pdf(pdf_docs):
+    docs=[]
+    for pdf in pdf_docs:
+        temp_file = "./temp.pdf"
+        # Delete the existing temp.pdf file if it exists
+        if os.path.exists(temp_file):
+            os.remove(temp_file)
+        with open(temp_file, "wb") as file:
+            file.write(pdf.getvalue())
+            file_name = pdf.name
+        loader = PyPDFLoader(temp_file)
+        docs.extend(loader.load())
+    return  docs
+    
+def text_splitter(text):
+    text_splitter = RecursiveCharacterTextSplitter(
+    # Set a really small chunk size, just to show.
+    chunk_size=10000,
+    chunk_overlap=500,
+    separators=["\n\n","\n"," ",".",","])
+    chunks=text_splitter.split_documents(text)
+    return chunks
+
+
+def get_conversational_chain(retriever):
+    prompt_template = """
+    Given the following extracted parts of a long document and a question, create a final answer.
+    Answer the question as detailed as possible from the provided context, make sure to provide all the details, if the answer is not in
+    provided context just say, "answer is not available in the context", and then ignore the context and add the answer from your knowledge like a simple llm prompt.
+    Try to give atleast the basic information.Do not return blank answer.\n\n
+    Make sure to understand the question and answer as per the question.
+    The answer should be a detailed one and try to incorporate examples for better understanding.
+    If the question involves terms like detailed or explained , give answer which involves complete detail about the question.\n\n
+    Context:\n {context}?\n
+    Question: \n{question}\n
+    Answer:
+    """
+    pipeline("text-generation", model="nvidia/Llama3-ChatQA-1.5-8B")
+
+    pt = ChatPromptTemplate.from_template(prompt_template)
+    # Retrieve and generate using the relevant snippets of the blog.
+    #retriever = db.as_retriever()
+    rag_chain = (
+    {"context": retriever, "question": RunnablePassthrough()}
+    | pt
+    | llm
+    | StrOutputParser()
+     )
+    return rag_chain
+
+def embedding(chunk,query):
+    embeddings=HuggingFaceEmbeddings()
+    db = Chroma.from_documents(chunk,embeddings)
+    doc = db.similarity_search(query)
+    chain = get_conversational_chain(db.as_retriever())
+    response = chain.invoke(query)
+    return response
+
+if 'messages' not in st.session_state:
+    st.session_state.messages = [{'role': 'assistant', "content": 'Hello! Upload a PDF and ask me questions.'}]
+
+    
+st.header("Chat with your pdf")
+with st.sidebar:
+    st.title("PDF FILE UPLOAD:")
+    pdf_docs = st.file_uploader("Upload your PDF File and Click on the Submit Button", accept_multiple_files=True, key="pdf_uploader")
+    
+query = st.chat_input("Ask a Question from the PDF File")    
+if query:
+    raw_text = get_pdf(pdf_docs)
+    text_chunks = text_splitter(raw_text)
+    st.session_state.messages.append({'role': 'user', "content": query})
+    response = embedding(text_chunks,query)
+    st.session_state.messages.append({'role': 'assistant', "content": response})
+
+for message in st.session_state.messages:
+    with st.chat_message(message['role']):
+        st.write(message['content'])