Spaces:
Sleeping
Sleeping
File size: 5,134 Bytes
b092c58 1c2e145 b092c58 |
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 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 |
from dataclasses import asdict
from io import StringIO
import json
import os
import streamlit as st
from data_driven_characters.character import generate_character_definition, Character
from data_driven_characters.corpus import (
generate_corpus_summaries,
generate_docs,
)
from data_driven_characters.chatbots import (
SummaryChatBot,
RetrievalChatBot,
SummaryRetrievalChatBot,
)
from data_driven_characters.interfaces import reset_chat, clear_user_input, converse
@st.cache_resource()
def create_chatbot(character_definition, corpus_summaries, chatbot_type):
if chatbot_type == "summary":
chatbot = SummaryChatBot(character_definition=character_definition)
elif chatbot_type == "retrieval":
chatbot = RetrievalChatBot(
character_definition=character_definition,
documents=corpus_summaries,
)
elif chatbot_type == "summary with retrieval":
chatbot = SummaryRetrievalChatBot(
character_definition=character_definition,
documents=corpus_summaries,
)
else:
raise ValueError(f"Unknown chatbot type: {chatbot_type}")
return chatbot
@st.cache_data(persist="disk")
def process_corpus(corpus):
# load docs
docs = generate_docs(
corpus=corpus,
chunk_size=2048,
chunk_overlap=64,
)
# generate summaries
corpus_summaries = generate_corpus_summaries(docs=docs, summary_type="map_reduce")
return corpus_summaries
@st.cache_data(persist="disk")
def get_character_definition(name, corpus_summaries):
character_definition = generate_character_definition(
name=name,
corpus_summaries=corpus_summaries,
)
return asdict(character_definition)
def main():
st.title("Data-Driven Characters")
st.write(
"Upload a corpus in the sidebar to generate a character chatbot that is grounded in the corpus content."
)
openai_api_key = st.text_input(
label="Your OpenAI API KEY",
placeholder="Your OpenAI API KEY",
type="password",
)
os.environ["OPENAI_API_KEY"] = openai_api_key
with st.sidebar:
uploaded_file = st.file_uploader("Upload corpus")
if uploaded_file is not None:
corpus_name = os.path.splitext(os.path.basename(uploaded_file.name))[0]
# read file
stringio = StringIO(uploaded_file.getvalue().decode("utf-8"))
corpus = stringio.read()
# scrollable text
st.markdown(
f"""
<div style='overflow: auto; height: 200px; border: 1px solid gray; border-radius: 5px; padding: 10px'>
{corpus}</div>
""",
unsafe_allow_html=True,
)
st.divider()
# get character name
character_name = st.text_input(f"Enter a character name from {corpus_name}")
if character_name:
if not openai_api_key:
st.error(
"You must enter an API key to use the OpenAI API. Please enter an API key in the sidebar."
)
return
if (
"character_name" in st.session_state
and st.session_state["character_name"] != character_name
):
clear_user_input()
reset_chat()
st.session_state["character_name"] = character_name
with st.spinner("Processing corpus (this will take a while)..."):
corpus_summaries = process_corpus(corpus)
with st.spinner("Generating character definition..."):
# get character definition
character_definition = get_character_definition(
name=character_name,
corpus_summaries=corpus_summaries,
)
print(json.dumps(character_definition, indent=4))
chatbot_type = st.selectbox(
"Select a memory type",
options=["summary", "retrieval", "summary with retrieval"],
index=2,
)
if (
"chatbot_type" in st.session_state
and st.session_state["chatbot_type"] != chatbot_type
):
clear_user_input()
reset_chat()
st.session_state["chatbot_type"] = chatbot_type
st.markdown(
f"[Export to character.ai](https://beta.character.ai/editing):"
)
st.write(character_definition)
if uploaded_file is not None and character_name:
st.divider()
chatbot = create_chatbot(
character_definition=Character(**character_definition),
corpus_summaries=corpus_summaries,
chatbot_type=chatbot_type,
)
converse(chatbot)
if __name__ == "__main__":
main()
|