initial commit

app.py

@@ -0,0 +1,101 @@
+import streamlit as st
+from langchain import PromptTemplate
+from langchain.llms import HuggingFaceHub
+from langchain.chains import RetrievalQA
+from langchain.embeddings import SentenceTransformerEmbeddings
+from qdrant_client import QdrantClient
+from langchain.vectorstores import Qdrant
+from huggingface_hub import login
+from langchain.document_loaders import PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.vectorstores import Chroma
+import os
+
+# Set up Streamlit UI
+st.title("HuggingFace QA with Langchain and Qdrant")
+st.write("This app leverages a Language Model to provide answers to your questions using retrieved context.")
+
+# Load HuggingFace token from environment variable for HuggingFace Space
+huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
+
+# Log in to HuggingFace Hub
+if huggingface_token:
+    login(token=huggingface_token)
+else:
+    st.error("HuggingFace token not found. Please set the HUGGINGFACE_TOKEN environment variable.")
+
+# HuggingFace Inference API Configuration
+config = {
+    'max_new_tokens': 1024,
+    'temperature': 0.1,
+    'top_k': 50,
+    'top_p': 0.9
+}
+
+# Use HuggingFaceHub for LLM
+llm = HuggingFaceHub(repo_id="stanford-crfm/BioMedLM", model_kwargs=config, huggingfacehub_api_token=huggingface_token)
+
+st.write("LLM Initialized....")
+
+prompt_template = """Use the following pieces of information to answer the user's question.
+If you don't know the answer, just say that you don't know, don't try to make up an answer.
+
+Context: {context}
+Question: {question}
+
+Only return the helpful answer below and nothing else.
+Helpful answer:
+"""
+
+embeddings = SentenceTransformerEmbeddings(model_name="NeuML/pubmedbert-base-embeddings")
+
+# PDF Loader and Document Processing
+uploaded_file = st.file_uploader("Upload a PDF file", type=["pdf"])
+if uploaded_file is not None:
+    loader = PyPDFLoader(uploaded_file)
+    documents = loader.load()
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+    docs = text_splitter.split_documents(documents)
+
+    # Create Chroma Vector Store from PDF
+    db = Chroma.from_documents(docs, embeddings)
+    retriever = db.as_retriever(search_kwargs={"k": 1})
+else:
+    # Use Qdrant if no PDF is uploaded
+    url = "http://localhost:6333"
+    client = QdrantClient(
+        url=url, prefer_grpc=False
+    )
+    db = Qdrant(client=client, embeddings=embeddings, collection_name="vector_db")
+    retriever = db.as_retriever(search_kwargs={"k": 1})
+
+prompt = PromptTemplate(template=prompt_template, input_variables=['context', 'question'])
+
+# Streamlit Form to get user input
+with st.form(key='query_form'):
+    query = st.text_input("Enter your question here:")
+    submit_button = st.form_submit_button(label='Get Answer')
+
+# Handle form submission
+if submit_button and query:
+    chain_type_kwargs = {"prompt": prompt}
+    qa = RetrievalQA.from_chain_type(
+        llm=llm,
+        chain_type="stuff",
+        retriever=retriever,
+        return_source_documents=True,
+        chain_type_kwargs=chain_type_kwargs,
+        verbose=True
+    )
+    response = qa(query)
+    answer = response['result']
+    source_document = response['source_documents'][0].page_content
+    doc = response['source_documents'][0].metadata.get('source', 'Uploaded PDF')
+
+    # Display the results
+    st.write("## Answer:")
+    st.write(answer)
+    st.write("## Source Document:")
+    st.write(source_document)
+    st.write("## Document Source:")
+    st.write(doc)