rsaketh02 commited on
Commit
f1fca27
·
verified ·
1 Parent(s): ef25e33

Upload 2 files

Browse files
Files changed (2) hide show
  1. app.py +72 -0
  2. requirements.txt +11 -0
app.py ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import streamlit as st
2
+ from langchain_groq import ChatGroq
3
+ from langchain_huggingface import HuggingFaceEmbeddings
4
+ from langchain_text_splitters import RecursiveCharacterTextSplitter
5
+ from langchain.chains.combine_documents import create_stuff_documents_chain
6
+ from langchain_core.prompts import ChatPromptTemplate
7
+ from langchain.chains import create_retrieval_chain
8
+ from langchain_community.vectorstores import FAISS
9
+ from langchain_community.document_loaders import PyPDFLoader
10
+
11
+ import os
12
+ from dotenv import load_dotenv
13
+ import tempfile
14
+ import time
15
+
16
+ load_dotenv()
17
+
18
+ ## Langsmith Tracking
19
+ os.environ['LANGCHAIN_API_KEY'] = os.getenv('LANGCHAIN_API_KEY')
20
+ os.environ['LANGCHAIN_TRACING_V2'] = 'true'
21
+ os.environ['LANGCHAIN_PROJECT'] = "Simple Q&A Chatbot With OpenAI"
22
+ os.environ['GROQ_API_KEY'] = os.getenv('GROQ_API_KEY')
23
+ os.environ["HF_TOKEN"] = os.getenv('HF_TOKEN')
24
+ os.environ["TOKENIZERS_PARALLELISM"] = "false"
25
+
26
+ llm = ChatGroq(model="llama-3.1-70b-Versatile")
27
+
28
+ prompt = ChatPromptTemplate.from_template(
29
+ """
30
+ Answer the question based on provided context only.
31
+ Please provide the most accurate response based on the question
32
+ <context>
33
+ {context}
34
+ </context>
35
+ Question: {input}
36
+ """
37
+ )
38
+
39
+ def create_vector_embeddings(pdf_file_path):
40
+ st.session_state.embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
41
+ st.session_state.loader = PyPDFLoader(pdf_file_path)
42
+ st.session_state.docs = st.session_state.loader.load()
43
+ st.session_state.text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
44
+ st.session_state.final_documents = st.session_state.text_splitter.split_documents(st.session_state.docs)
45
+ st.session_state.vectors = FAISS.from_documents(st.session_state.final_documents, st.session_state.embeddings)
46
+
47
+ uploaded_file = st.file_uploader("Upload a PDF", type="pdf", key="pdf_uploader")
48
+
49
+ user_prompt = st.text_input("Enter your Query about PDF here:")
50
+
51
+ if st.button("Document Embedding") and uploaded_file is not None:
52
+ with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
53
+ tmp_file.write(uploaded_file.getvalue())
54
+ tmp_file_path = tmp_file.name
55
+ create_vector_embeddings(tmp_file_path)
56
+ st.write("Vector Database is ready")
57
+
58
+ if user_prompt and "vectors" in st.session_state:
59
+ document_chain = create_stuff_documents_chain(llm, prompt)
60
+ retriever = st.session_state.vectors.as_retriever()
61
+ retrieval_chain = create_retrieval_chain(retriever, document_chain)
62
+
63
+ start = time.process_time()
64
+ response = retrieval_chain.invoke({"input": user_prompt})
65
+ st.write(f"Response Time: {time.process_time() - start}")
66
+
67
+ st.write(response["answer"])
68
+
69
+ with st.expander("Document Similarity Search"):
70
+ for i, doc in enumerate(response["context"]):
71
+ st.write(doc.page_content)
72
+ st.write("---------------------------------------")
requirements.txt ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ langchain_groq
2
+ langchain
3
+ python-dotenv
4
+ langchain_community
5
+ langchain_core
6
+ streamlit
7
+ langchain_huggingface
8
+ langchain-text-splitters
9
+ pypdf
10
+ faiss-cpu
11
+ langchain-openai