Update app.py
Browse files
app.py
CHANGED
@@ -1,87 +1,43 @@
|
|
1 |
import gradio as gr
|
2 |
-
import gdown
|
3 |
import pandas as pd
|
4 |
import os
|
5 |
-
import shutil
|
6 |
import threading
|
7 |
import schedule
|
8 |
import time
|
9 |
from datetime import datetime
|
10 |
-
from llama_cpp import Llama
|
11 |
from llama_index.core import SimpleDirectoryReader, VectorStoreIndex, Settings, Document
|
12 |
-
from llama_index.core.
|
13 |
-
from llama_index.llms.llama_cpp import LlamaCPP
|
14 |
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
|
15 |
from huggingface_hub import hf_hub_download
|
16 |
-
from llama_index.
|
17 |
-
|
18 |
-
# Fungsi untuk mengunduh model Llama
|
19 |
-
def initialize_llama_model():
|
20 |
-
model_path = hf_hub_download(
|
21 |
-
repo_id="TheBloke/zephyr-7b-beta-GGUF",
|
22 |
-
filename="zephyr-7b-beta.Q4_K_M.gguf",
|
23 |
-
cache_dir="./models"
|
24 |
-
)
|
25 |
-
return model_path
|
26 |
-
|
27 |
-
# Fungsi untuk mengatur konfigurasi Settings
|
28 |
-
def initialize_settings(model_path):
|
29 |
-
Settings.llm = LlamaCPP(
|
30 |
-
model_path=model_path,
|
31 |
-
temperature=0.7,
|
32 |
-
)
|
33 |
-
|
34 |
-
# Fungsi untuk mengunduh file CSV terbaru dari Google Drive
|
35 |
-
def download_csv_from_drive():
|
36 |
-
csv_url = "https://drive.google.com/uc?id=1UIx369_8GlzPiKArMVg8v-IwC6hYTYA0" # Ganti dengan ID file terbaru
|
37 |
-
output_folder = "bahandokumen"
|
38 |
-
output_csv = os.path.join(output_folder, "data.csv") # Path lengkap ke dalam folder
|
39 |
-
|
40 |
-
# Pastikan folder bahandokumen/ ada
|
41 |
-
os.makedirs(output_folder, exist_ok=True)
|
42 |
|
43 |
-
|
44 |
-
|
45 |
-
|
|
|
46 |
|
47 |
-
|
|
|
48 |
try:
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
else:
|
53 |
-
print("β Gagal mengunduh file. Cek kembali link Google Drive.")
|
54 |
except Exception as e:
|
55 |
-
print(f"β
|
56 |
-
|
57 |
-
return output_csv
|
58 |
|
59 |
-
# Fungsi untuk
|
60 |
-
def save_to_huggingface():
|
61 |
-
src = "bahandokumen/data.csv"
|
62 |
-
dst = "/home/user/app/bahandokumen/data.csv" # Jalur sesuai dengan HF Spaces
|
63 |
-
|
64 |
-
# Pastikan file ada sebelum menyalin
|
65 |
-
if os.path.exists(src):
|
66 |
-
if os.path.abspath(src) != os.path.abspath(dst): # Cek apakah lokasi berbeda
|
67 |
-
shutil.copy(src, dst)
|
68 |
-
print(f"β
File {src} berhasil disalin ke {dst}")
|
69 |
-
else:
|
70 |
-
print("β οΈ File sudah berada di lokasi yang sama, tidak perlu disalin.")
|
71 |
-
else:
|
72 |
-
print("β File tidak ditemukan, pastikan berhasil diunduh.")
|
73 |
-
|
74 |
-
# Fungsi untuk update index chatbot
|
75 |
def update_index():
|
76 |
print(f"π [{datetime.now()}] Mengupdate index dengan data terbaru...")
|
77 |
-
|
78 |
-
|
|
|
|
|
|
|
79 |
|
80 |
-
# Baca CSV dengan Pandas
|
81 |
-
df = pd.read_csv(csv_file)
|
82 |
documents = [Document(text=" | ".join(map(str, row.values))) for _, row in df.iterrows()]
|
83 |
|
84 |
-
# Tambahkan
|
85 |
text_documents = SimpleDirectoryReader(input_files=[
|
86 |
"bahandokumen/K3.txt",
|
87 |
"bahandokumen/bonus.txt",
|
@@ -103,42 +59,39 @@ def update_index():
|
|
103 |
index = VectorStoreIndex(nodes)
|
104 |
print("β
Index berhasil diperbarui!")
|
105 |
|
106 |
-
# Inisialisasi
|
107 |
def initialize_chat_engine():
|
108 |
return index.as_chat_engine(chat_mode="condense_plus_context", similarity_top_k=3)
|
109 |
|
110 |
-
# Fungsi untuk
|
111 |
def generate_response(message, history):
|
112 |
chat_engine = initialize_chat_engine()
|
113 |
response = chat_engine.stream_chat(message)
|
114 |
-
text = "".join(response.response_gen)
|
115 |
-
history.append((message, text))
|
116 |
return history
|
117 |
|
118 |
-
# Scheduler untuk update otomatis
|
119 |
def schedule_update():
|
120 |
schedule.every().day.at("00:00").do(update_index)
|
121 |
while True:
|
122 |
schedule.run_pending()
|
123 |
-
time.sleep(60)
|
124 |
|
125 |
-
#
|
126 |
def start_scheduler():
|
127 |
thread = threading.Thread(target=schedule_update, daemon=True)
|
128 |
thread.start()
|
129 |
|
130 |
-
# Inisialisasi
|
131 |
def launch_gradio():
|
132 |
with gr.Blocks() as demo:
|
133 |
-
|
134 |
-
chat_interface = gr.ChatInterface(lambda message, history: generate_response(message, history))
|
135 |
demo.launch()
|
136 |
|
137 |
-
# Fungsi
|
138 |
def main():
|
139 |
global index # Agar index bisa diperbarui secara global
|
140 |
-
model_path = initialize_llama_model()
|
141 |
-
initialize_settings(model_path)
|
142 |
|
143 |
print("π Inisialisasi index pertama kali...")
|
144 |
update_index()
|
@@ -149,4 +102,4 @@ def main():
|
|
149 |
launch_gradio()
|
150 |
|
151 |
if __name__ == "__main__":
|
152 |
-
main()
|
|
|
1 |
import gradio as gr
|
|
|
2 |
import pandas as pd
|
3 |
import os
|
|
|
4 |
import threading
|
5 |
import schedule
|
6 |
import time
|
7 |
from datetime import datetime
|
|
|
8 |
from llama_index.core import SimpleDirectoryReader, VectorStoreIndex, Settings, Document
|
9 |
+
from llama_index.core.node_parser import SentenceSplitter
|
|
|
10 |
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
|
11 |
from huggingface_hub import hf_hub_download
|
12 |
+
from llama_index.llms.llama_cpp import LlamaCPP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
|
14 |
+
# Fungsi untuk membaca Google Sheets sebagai DataFrame
|
15 |
+
def fetch_google_sheets_data():
|
16 |
+
sheet_id = "1740C15sMgOowI5wv5Q-kmYQUXMJUEyRSrFXRwIMc4Z0" # Ganti dengan ID Google Sheets Anda
|
17 |
+
sheet_name = "datatarget" # Ganti dengan nama sheet
|
18 |
|
19 |
+
url = f"https://docs.google.com/spreadsheets/d/1740C15sMgOowI5wv5Q-kmYQUXMJUEyRSrFXRwIMc4Z0/edit?gid=1097907532#gid=1097907532"
|
20 |
+
|
21 |
try:
|
22 |
+
df = pd.read_csv(url)
|
23 |
+
print("β
Data berhasil diambil dari Google Sheets!")
|
24 |
+
return df
|
|
|
|
|
25 |
except Exception as e:
|
26 |
+
print(f"β Gagal mengambil data: {e}")
|
27 |
+
return None
|
|
|
28 |
|
29 |
+
# Fungsi untuk update index chatbot dengan data terbaru dari Google Sheets
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
def update_index():
|
31 |
print(f"π [{datetime.now()}] Mengupdate index dengan data terbaru...")
|
32 |
+
|
33 |
+
df = fetch_google_sheets_data()
|
34 |
+
if df is None:
|
35 |
+
print("β οΈ Tidak bisa memperbarui index karena gagal membaca Google Sheets.")
|
36 |
+
return
|
37 |
|
|
|
|
|
38 |
documents = [Document(text=" | ".join(map(str, row.values))) for _, row in df.iterrows()]
|
39 |
|
40 |
+
# Tambahkan dokumen tambahan dari teks lokal
|
41 |
text_documents = SimpleDirectoryReader(input_files=[
|
42 |
"bahandokumen/K3.txt",
|
43 |
"bahandokumen/bonus.txt",
|
|
|
59 |
index = VectorStoreIndex(nodes)
|
60 |
print("β
Index berhasil diperbarui!")
|
61 |
|
62 |
+
# Inisialisasi Chatbot
|
63 |
def initialize_chat_engine():
|
64 |
return index.as_chat_engine(chat_mode="condense_plus_context", similarity_top_k=3)
|
65 |
|
66 |
+
# Fungsi untuk mendapatkan respons chatbot
|
67 |
def generate_response(message, history):
|
68 |
chat_engine = initialize_chat_engine()
|
69 |
response = chat_engine.stream_chat(message)
|
70 |
+
text = "".join(response.response_gen)
|
71 |
+
history.append((message, text))
|
72 |
return history
|
73 |
|
74 |
+
# Scheduler untuk update otomatis
|
75 |
def schedule_update():
|
76 |
schedule.every().day.at("00:00").do(update_index)
|
77 |
while True:
|
78 |
schedule.run_pending()
|
79 |
+
time.sleep(60)
|
80 |
|
81 |
+
# Menjalankan scheduler di thread terpisah
|
82 |
def start_scheduler():
|
83 |
thread = threading.Thread(target=schedule_update, daemon=True)
|
84 |
thread.start()
|
85 |
|
86 |
+
# Inisialisasi Gradio
|
87 |
def launch_gradio():
|
88 |
with gr.Blocks() as demo:
|
89 |
+
gr.ChatInterface(lambda message, history: generate_response(message, history))
|
|
|
90 |
demo.launch()
|
91 |
|
92 |
+
# Fungsi utama
|
93 |
def main():
|
94 |
global index # Agar index bisa diperbarui secara global
|
|
|
|
|
95 |
|
96 |
print("π Inisialisasi index pertama kali...")
|
97 |
update_index()
|
|
|
102 |
launch_gradio()
|
103 |
|
104 |
if __name__ == "__main__":
|
105 |
+
main()
|