Update app.py
Browse files
app.py
CHANGED
@@ -2,6 +2,7 @@ import gradio as gr
|
|
2 |
import gdown
|
3 |
import pandas as pd
|
4 |
import os
|
|
|
5 |
import threading
|
6 |
import schedule
|
7 |
import time
|
@@ -17,7 +18,7 @@ from llama_index.core.node_parser import SentenceSplitter
|
|
17 |
# Fungsi untuk mengunduh model Llama
|
18 |
def initialize_llama_model():
|
19 |
model_path = hf_hub_download(
|
20 |
-
repo_id="
|
21 |
filename="zephyr-7b-beta.Q4_K_M.gguf",
|
22 |
cache_dir="./models"
|
23 |
)
|
@@ -32,25 +33,43 @@ def initialize_settings(model_path):
|
|
32 |
|
33 |
# Fungsi untuk mengunduh file CSV terbaru dari Google Drive
|
34 |
def download_csv_from_drive():
|
35 |
-
csv_url = "https://drive.google.com/uc?id=1UIx369_8GlzPiKArMVg8v-IwC6hYTYA0" # Ganti dengan ID file
|
36 |
output_csv = "bahandokumen/data.csv"
|
37 |
|
38 |
if os.path.exists(output_csv):
|
39 |
os.remove(output_csv) # Hapus file lama
|
40 |
|
41 |
print("π Mengunduh file CSV terbaru...")
|
42 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
return output_csv
|
44 |
|
45 |
-
# Fungsi untuk
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
46 |
def update_index():
|
47 |
print(f"π [{datetime.now()}] Mengupdate index dengan data terbaru...")
|
48 |
csv_file = download_csv_from_drive()
|
49 |
-
|
|
|
50 |
# Baca CSV dengan Pandas
|
51 |
df = pd.read_csv(csv_file)
|
52 |
-
|
53 |
-
# Konversi isi CSV menjadi dokumen teks
|
54 |
documents = [Document(text=" | ".join(map(str, row.values))) for _, row in df.iterrows()]
|
55 |
|
56 |
# Tambahkan file dokumen lain
|
@@ -65,16 +84,12 @@ def update_index():
|
|
65 |
]).load_data()
|
66 |
|
67 |
documents.extend(text_documents)
|
68 |
-
|
69 |
-
# Parsing dokumen menjadi nodes
|
70 |
parser = SentenceSplitter(chunk_size=150, chunk_overlap=10)
|
71 |
nodes = parser.get_nodes_from_documents(documents)
|
72 |
|
73 |
-
# Gunakan model embedding
|
74 |
embedding = HuggingFaceEmbedding("firqaaa/indo-sentence-bert-base")
|
75 |
Settings.embed_model = embedding
|
76 |
|
77 |
-
# Buat index vektor
|
78 |
global index
|
79 |
index = VectorStoreIndex(nodes)
|
80 |
print("β
Index berhasil diperbarui!")
|
@@ -85,7 +100,7 @@ def initialize_chat_engine():
|
|
85 |
|
86 |
# Fungsi untuk menghasilkan respons chatbot
|
87 |
def generate_response(message, history):
|
88 |
-
chat_engine = initialize_chat_engine()
|
89 |
response = chat_engine.stream_chat(message)
|
90 |
text = "".join(response.response_gen)
|
91 |
history.append((message, text))
|
@@ -117,12 +132,12 @@ def main():
|
|
117 |
initialize_settings(model_path)
|
118 |
|
119 |
print("π Inisialisasi index pertama kali...")
|
120 |
-
update_index()
|
121 |
|
122 |
print("β³ Menjalankan scheduler update index (setiap jam 12 malam)...")
|
123 |
-
start_scheduler()
|
124 |
|
125 |
-
launch_gradio()
|
126 |
|
127 |
if __name__ == "__main__":
|
128 |
main()
|
|
|
2 |
import gdown
|
3 |
import pandas as pd
|
4 |
import os
|
5 |
+
import shutil
|
6 |
import threading
|
7 |
import schedule
|
8 |
import time
|
|
|
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 |
)
|
|
|
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_csv = "bahandokumen/data.csv"
|
38 |
|
39 |
if os.path.exists(output_csv):
|
40 |
os.remove(output_csv) # Hapus file lama
|
41 |
|
42 |
print("π Mengunduh file CSV terbaru...")
|
43 |
+
try:
|
44 |
+
gdown.download(csv_url, output_csv, quiet=False)
|
45 |
+
if os.path.exists(output_csv):
|
46 |
+
print(f"β
File berhasil diunduh: {output_csv}")
|
47 |
+
else:
|
48 |
+
print("β Gagal mengunduh file. Cek kembali link Google Drive.")
|
49 |
+
except Exception as e:
|
50 |
+
print(f"β Terjadi kesalahan saat mengunduh file: {e}")
|
51 |
+
|
52 |
return output_csv
|
53 |
|
54 |
+
# Fungsi untuk menyimpan file ke Hugging Face Spaces
|
55 |
+
def save_to_huggingface():
|
56 |
+
src = "bahandokumen/data.csv"
|
57 |
+
dst = "/home/user/app/bahandokumen/data.csv" # Jalur sesuai dengan HF Spaces
|
58 |
+
|
59 |
+
if os.path.exists(src):
|
60 |
+
shutil.copy(src, dst)
|
61 |
+
print(f"β
File {src} berhasil disalin ke {dst}")
|
62 |
+
else:
|
63 |
+
print("β File tidak ditemukan, pastikan berhasil diunduh.")
|
64 |
+
|
65 |
+
# Fungsi untuk update index chatbot
|
66 |
def update_index():
|
67 |
print(f"π [{datetime.now()}] Mengupdate index dengan data terbaru...")
|
68 |
csv_file = download_csv_from_drive()
|
69 |
+
save_to_huggingface()
|
70 |
+
|
71 |
# Baca CSV dengan Pandas
|
72 |
df = pd.read_csv(csv_file)
|
|
|
|
|
73 |
documents = [Document(text=" | ".join(map(str, row.values))) for _, row in df.iterrows()]
|
74 |
|
75 |
# Tambahkan file dokumen lain
|
|
|
84 |
]).load_data()
|
85 |
|
86 |
documents.extend(text_documents)
|
|
|
|
|
87 |
parser = SentenceSplitter(chunk_size=150, chunk_overlap=10)
|
88 |
nodes = parser.get_nodes_from_documents(documents)
|
89 |
|
|
|
90 |
embedding = HuggingFaceEmbedding("firqaaa/indo-sentence-bert-base")
|
91 |
Settings.embed_model = embedding
|
92 |
|
|
|
93 |
global index
|
94 |
index = VectorStoreIndex(nodes)
|
95 |
print("β
Index berhasil diperbarui!")
|
|
|
100 |
|
101 |
# Fungsi untuk menghasilkan respons chatbot
|
102 |
def generate_response(message, history):
|
103 |
+
chat_engine = initialize_chat_engine()
|
104 |
response = chat_engine.stream_chat(message)
|
105 |
text = "".join(response.response_gen)
|
106 |
history.append((message, text))
|
|
|
132 |
initialize_settings(model_path)
|
133 |
|
134 |
print("π Inisialisasi index pertama kali...")
|
135 |
+
update_index()
|
136 |
|
137 |
print("β³ Menjalankan scheduler update index (setiap jam 12 malam)...")
|
138 |
+
start_scheduler()
|
139 |
|
140 |
+
launch_gradio()
|
141 |
|
142 |
if __name__ == "__main__":
|
143 |
main()
|