initial commit

.gitignore

@@ -0,0 +1 @@
+.streamlit

app.py

@@ -0,0 +1,63 @@
+import os
+import pinecone
+import openai
+
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.text_splitter import CharacterTextSplitter
+from langchain.vectorstores import Pinecone
+from langchain.document_loaders import PyPDFLoader
+from langchain.chains import ConversationalRetrievalChain
+from langchain.llms import OpenAI
+import streamlit as st
+
+
+def _initialize_env():
+    pinecone.init(
+        api_key=st.secrets["pinecone_api_key"],
+        environment=st.secrets["pinecone_env"]
+    )
+    openai.api_key = st.secrets["openai_api_key"]
+
+
+def _initialize_indexes():
+    loader = PyPDFLoader("./etf-book.pdf")
+    documents = loader.load()
+    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
+    docs = text_splitter.split_documents(documents)
+    embeddings = OpenAIEmbeddings()
+
+    PINECONE_TABLE_NAME = os.getenv('PINECONE_TABLE_NAME')
+    db = Pinecone.from_documents(docs, embeddings, index_name=PINECONE_TABLE_NAME)
+    return db
+
+
+def _initialize_retriever(db):
+    retriever = db.as_retriever(search_type="similarity", search_kwargs={"k": 2})
+    qa = ConversationalRetrievalChain.from_llm(OpenAI(temperature=0, model_name="gpt-3.5-turbo"), retriever)
+    return qa
+
+
+chat_history = []
+
+def answer(input):
+    result = qa({
+        "question": input,
+        "chat_history": chat_history
+    })
+    chat_history = chat_history.append((input, result["answer"]))
+    chat_history = chat_history[-10:]
+    st.write(
+        "Bot: ",
+        result["answer"]
+    )
+
+
+_initialize_env()
+db = _initialize_indexes()
+qa = _initialize_retriever(db)
+
+question = st.text_input('Question')
+st.button(
+    'Answer',
+    on_click=lambda: answer(question)
+)

requirements.txt

@@ -0,0 +1,4 @@
+langchain==0.0.178
+openai==0.27.7
+pinecone==0.1.0
+streamlit==1.22.0