Skip to content

armgit / burr

Public Imported from GitHub
0

Build applications that make decisions (chatbots, agents, simulations, etc...). Monitor, trace, persist, and execute on your own infrastructure.

4.0 KB
Raw Blame
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.

from typing import Optional

import application as chatbot_application
import streamlit as st

from burr.integrations.streamlit import (
    AppState,
    Record,
    get_state,
    render_explorer,
    set_slider_to_current,
    update_state,
)

st.set_page_config(layout="wide")

st.markdown(
    "This is a prototype streamlit app -- this is a UI for demonstration purposes, but you can"
    " run the real UI by running `burr` -- a window will pop up and you can follow the results live!"
)


def render_chat_message(record: Record):
    if record.action in ["prompt", "response"]:
        recent_chat_message = record.state["chat_history"][-1]
        content = recent_chat_message["content"]
        content_type = recent_chat_message["type"]
        role = recent_chat_message["role"]
        with st.chat_message(role):
            if content_type == "image":
                st.image(content)
            elif content_type == "code":
                st.code(content)
            elif content_type == "text":
                st.write(content)
            elif content_type == "error":
                st.error(content)


def retrieve_state():
    if "burr_state" not in st.session_state:
        # TODO --enable usage of hamilton. Currently its not wiring in inputs
        # But it should be easy enough
        state = AppState.from_empty(
            app=chatbot_application.application(),
        )
    else:
        state = get_state()
    return state


def chatbot_step(app_state: AppState, prompt: Optional[str]) -> bool:
    """Pushes state forward for the chatbot. Returns whether or not to rerun the app.

    :param app_state: State of the app
    :param prompt: Prompt to set the chatbot to. If this is None it means it should continue and not be reset.
    :return:
    """
    inputs = None
    if prompt is not None:
        # We need to update
        inputs = {"prompt": prompt}
        # app_state.app.update_state(app_state.app.state.update(prompt=prompt))
        st.session_state.running = True  # set to running
    # if its not running this is a no-op
    if not st.session_state.get("running", False):
        return False
    application = app_state.app
    step_output = application.step(inputs=inputs)
    if step_output is None:
        st.session_state.running = False
        return False
    action, result, state = step_output
    app_state.history.append(Record(state.get_all(), action.name, result))
    set_slider_to_current()
    if action.name == "response":
        # we've gotten to the end
        st.session_state.running = False
        return True  # run one last time
    return True


def main():
    st.title("Chatbot example with Burr")
    app_state = retrieve_state()  # retrieve first so we can use for the ret of the step
    columns = st.columns(2)
    with columns[0]:
        prompt = st.chat_input(
            "...", disabled=st.session_state.get("running", False), key="chat_input"
        )
        should_rerun = chatbot_step(app_state, prompt)
        with st.container(height=850):
            for item in app_state.history:
                render_chat_message(item)
    with columns[1]:
        render_explorer(app_state)
    update_state(app_state)  # update so the next iteration knows what to do
    if should_rerun:
        st.rerun()


if __name__ == "__main__":
    main()