Spaces:
Sleeping
Sleeping
File size: 1,974 Bytes
5ed07e1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 |
from langchain.callbacks.base import BaseCallbackHandler
from langchain.schema import ChatMessage
from langchain_openai import ChatOpenAI
import streamlit as st
st.set_page_config(page_title="Streamlit + Langchain")
st.title("Basic Chatbot with Streamlit and Langchain")
st.caption("Features text streaming")
class StreamHandler(BaseCallbackHandler):
def __init__(self, container, text=""):
self.container = container
self.text = text
def on_llm_new_token(self, token: str, **kwargs) -> None:
self.text += token
self.container.markdown(self.text)
# Text input to enter OpenAI API key
with st.sidebar:
OPENAI_API_KEY = st.text_input("Enter your OpenAI API Key", type="password")
# Streamlit session state
if "messages" not in st.session_state:
st.session_state["messages"] = [
ChatMessage(role="assistant", content="How can I help you?")
]
# Display all chat messages from session state
for message in st.session_state.messages:
st.chat_message(message.role).write(message.content)
# If user submits a prompt in the text input, continue
if prompt := st.chat_input():
if not OPENAI_API_KEY:
st.error("Please add your OpenAI API key to continue.")
st.stop()
# Add user's prompt to the chat messages
st.session_state.messages.append(ChatMessage(role="user", content=prompt))
st.chat_message("user").write(prompt)
# Display the assistant's response with langchain query
with st.chat_message("assistant"):
stream_handler = StreamHandler(st.empty())
llm = ChatOpenAI(
model="gpt-4o-mini",
openai_api_key=OPENAI_API_KEY,
streaming=True,
callbacks=[stream_handler],
)
response = llm.invoke(st.session_state.messages)
st.session_state.messages.append(
ChatMessage(role="assistant", content=response.content)
)
|