Spaces:
Running
Running
Upload 3 files
Browse files- app.py +131 -0
- dataset_loader.py +18 -0
- requirements.txt +11 -0
app.py
ADDED
@@ -0,0 +1,131 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import pickle
|
3 |
+
from pathlib import Path
|
4 |
+
import streamlit as st
|
5 |
+
import streamlit_authenticator as stauth
|
6 |
+
import traceback
|
7 |
+
from dataset_loader import load_dataset
|
8 |
+
from dotenv import load_dotenv
|
9 |
+
|
10 |
+
load_dotenv()
|
11 |
+
|
12 |
+
if not os.path.isdir("spsetia_chatbot_data"):
|
13 |
+
load_dataset()
|
14 |
+
|
15 |
+
from spsetia_chatbot_data.utils import (
|
16 |
+
run_conversation,
|
17 |
+
record_db,
|
18 |
+
hashed_passwords,
|
19 |
+
)
|
20 |
+
|
21 |
+
|
22 |
+
st.set_page_config(page_title="S P Setia Chatbot", page_icon=":robot:", layout="wide")
|
23 |
+
|
24 |
+
|
25 |
+
# --- USER AUTHENTICATION ---
|
26 |
+
names = ["S P Setia Admin"]
|
27 |
+
usernames = ["spsetia_admin"]
|
28 |
+
|
29 |
+
authenticator = stauth.Authenticate(
|
30 |
+
names,
|
31 |
+
usernames,
|
32 |
+
hashed_passwords,
|
33 |
+
"spsetia_chatbot",
|
34 |
+
"eqiddnksadjw",
|
35 |
+
cookie_expiry_days=30,
|
36 |
+
)
|
37 |
+
|
38 |
+
name, authentication_status, username = authenticator.login("Login", "main")
|
39 |
+
|
40 |
+
if authentication_status == False:
|
41 |
+
st.error("Username/password is incorrect")
|
42 |
+
|
43 |
+
if authentication_status == None:
|
44 |
+
st.warning("Please enter your username and password")
|
45 |
+
|
46 |
+
if authentication_status:
|
47 |
+
st.sidebar.title(f"S P Setia Chatbot")
|
48 |
+
st.sidebar.header(f"Welcome {name}")
|
49 |
+
authenticator.logout("Logout", "sidebar")
|
50 |
+
if st.sidebar.button("Refresh Chat"):
|
51 |
+
st.session_state.messages = [
|
52 |
+
{"role": "assistant", "content": "How may I help you?"}
|
53 |
+
]
|
54 |
+
if "messages" not in st.session_state:
|
55 |
+
st.session_state.messages = [
|
56 |
+
{"role": "assistant", "content": "How may I help you?"}
|
57 |
+
]
|
58 |
+
|
59 |
+
for message in st.session_state.messages:
|
60 |
+
with st.chat_message(message["role"]):
|
61 |
+
st.write(message["content"])
|
62 |
+
|
63 |
+
if prompt := st.chat_input():
|
64 |
+
st.session_state.messages.append({"role": "user", "content": prompt})
|
65 |
+
with st.chat_message("user"):
|
66 |
+
st.write(prompt)
|
67 |
+
|
68 |
+
if st.session_state.messages[-1]["role"] != "assistant":
|
69 |
+
try:
|
70 |
+
with st.chat_message("assistant"):
|
71 |
+
with st.spinner("Thinking..."):
|
72 |
+
response = run_conversation(
|
73 |
+
st.session_state.messages, prompt, username
|
74 |
+
)
|
75 |
+
qa_id = response[4]
|
76 |
+
prompt_tokens = response[5]
|
77 |
+
completion_tokens = response[6]
|
78 |
+
if response[3]:
|
79 |
+
if isinstance(response[0], list):
|
80 |
+
answer = ""
|
81 |
+
for item in response[0]:
|
82 |
+
answer += f"""{item[0]}\n"""
|
83 |
+
st.text(item[0])
|
84 |
+
for image in item[1]:
|
85 |
+
if image:
|
86 |
+
st.image(image)
|
87 |
+
reference = f""" \n\nReference \nfolder: {response[2]["folder"]} \nfile: {response[2]["file"]}"""
|
88 |
+
st.text(reference)
|
89 |
+
answer += reference
|
90 |
+
record_db(
|
91 |
+
content=answer,
|
92 |
+
qa_id=qa_id,
|
93 |
+
msg_mode=1,
|
94 |
+
recipient_id=username,
|
95 |
+
prompt_tokens=prompt_tokens,
|
96 |
+
completion_tokens=completion_tokens,
|
97 |
+
)
|
98 |
+
else:
|
99 |
+
answer = response[0]
|
100 |
+
reference = f""" \n\nReference \nfolder: {response[2]["folder"]} \nfile: {response[2]["file"]}"""
|
101 |
+
answer += reference
|
102 |
+
st.text(answer)
|
103 |
+
record_db(
|
104 |
+
content=answer,
|
105 |
+
qa_id=qa_id,
|
106 |
+
msg_mode=1,
|
107 |
+
recipient_id=username,
|
108 |
+
prompt_tokens=prompt_tokens,
|
109 |
+
completion_tokens=completion_tokens,
|
110 |
+
)
|
111 |
+
else:
|
112 |
+
answer = response[0]
|
113 |
+
st.text(answer)
|
114 |
+
if images := response[1]:
|
115 |
+
if isinstance(images, list):
|
116 |
+
for image in images:
|
117 |
+
st.image(image)
|
118 |
+
record_db(
|
119 |
+
content=answer,
|
120 |
+
qa_id=qa_id,
|
121 |
+
msg_mode=1,
|
122 |
+
recipient_id=username,
|
123 |
+
prompt_tokens=prompt_tokens,
|
124 |
+
completion_tokens=completion_tokens,
|
125 |
+
)
|
126 |
+
|
127 |
+
message = {"role": "assistant", "content": answer}
|
128 |
+
st.session_state.messages.append(message)
|
129 |
+
except:
|
130 |
+
st.error("An error occurred. Please try again.")
|
131 |
+
traceback.print_exc()
|
dataset_loader.py
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from huggingface_hub import Repository
|
2 |
+
from huggingface_hub import login
|
3 |
+
import os
|
4 |
+
from dotenv import load_dotenv
|
5 |
+
|
6 |
+
load_dotenv()
|
7 |
+
|
8 |
+
|
9 |
+
def load_dataset():
|
10 |
+
login(token=os.getenv("HUB_TOKEN"))
|
11 |
+
|
12 |
+
repo = Repository(
|
13 |
+
local_dir="spsetia_chatbot_data",
|
14 |
+
repo_type="dataset",
|
15 |
+
clone_from="cjzhi98/spsetia-chatbot",
|
16 |
+
token=True,
|
17 |
+
)
|
18 |
+
repo.git_pull()
|
requirements.txt
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
openai
|
2 |
+
pandas
|
3 |
+
fuzzywuzzy
|
4 |
+
python-dotenv
|
5 |
+
numpy
|
6 |
+
python-Levenshtein
|
7 |
+
streamlit
|
8 |
+
streamlit-authenticator==0.1.5
|
9 |
+
prettyprint
|
10 |
+
boto3
|
11 |
+
huggingface_hub
|