Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -8,12 +8,29 @@ import librosa
|
|
| 8 |
import soundfile as sf
|
| 9 |
import pandas as pd
|
| 10 |
import numpy as np
|
| 11 |
-
from typing import List,
|
| 12 |
import shutil
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
# Khởi tạo model Whisper
|
| 15 |
model = WhisperModel("large-v3-turbo", device="cpu", compute_type="int8")
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
def extract_audio_files(input_file: str, temp_dir: str) -> List[str]:
|
| 18 |
"""Giải nén file zip hoặc copy file audio đơn"""
|
| 19 |
audio_files = []
|
|
@@ -64,7 +81,6 @@ def merge_short_segments(segments: List[dict], min_duration: float = 2.0) -> Lis
|
|
| 64 |
current_duration = current['end'] - current['start']
|
| 65 |
|
| 66 |
if current_duration < min_duration:
|
| 67 |
-
# Gộp với segment tiếp theo
|
| 68 |
current['end'] = seg['end']
|
| 69 |
current['text'] = current['text'] + ' ' + seg['text']
|
| 70 |
else:
|
|
@@ -121,12 +137,10 @@ def save_to_parquet(records: List[dict], output_dir: str, max_size_mb: int = 500
|
|
| 121 |
parquet_files = []
|
| 122 |
|
| 123 |
if file_size_mb <= max_size_mb:
|
| 124 |
-
|
| 125 |
-
output_path = os.path.join(output_dir, 'dataset.parquet')
|
| 126 |
df.to_parquet(output_path, engine='pyarrow')
|
| 127 |
parquet_files.append(output_path)
|
| 128 |
else:
|
| 129 |
-
# Chia nhỏ thành nhiều parts
|
| 130 |
num_parts = int(np.ceil(file_size_mb / max_size_mb))
|
| 131 |
chunk_size = len(df) // num_parts + 1
|
| 132 |
|
|
@@ -135,22 +149,38 @@ def save_to_parquet(records: List[dict], output_dir: str, max_size_mb: int = 500
|
|
| 135 |
end_idx = min((i + 1) * chunk_size, len(df))
|
| 136 |
|
| 137 |
df_chunk = df.iloc[start_idx:end_idx]
|
| 138 |
-
output_path = os.path.join(output_dir, f'
|
| 139 |
df_chunk.to_parquet(output_path, engine='pyarrow')
|
| 140 |
parquet_files.append(output_path)
|
| 141 |
|
| 142 |
return parquet_files
|
| 143 |
|
| 144 |
-
def
|
| 145 |
-
"""
|
| 146 |
-
|
| 147 |
-
|
| 148 |
-
|
| 149 |
-
|
| 150 |
-
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 154 |
|
| 155 |
os.makedirs(extract_dir, exist_ok=True)
|
| 156 |
os.makedirs(audio_output_dir, exist_ok=True)
|
|
@@ -160,12 +190,24 @@ def process_audio(input_file):
|
|
| 160 |
audio_files = extract_audio_files(input_file, extract_dir)
|
| 161 |
|
| 162 |
if not audio_files:
|
| 163 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 164 |
|
| 165 |
all_records = []
|
| 166 |
|
| 167 |
# Xử lý từng file audio
|
| 168 |
-
for audio_file in audio_files:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 169 |
base_name = Path(audio_file).stem
|
| 170 |
|
| 171 |
# Transcribe
|
|
@@ -184,6 +226,10 @@ def process_audio(input_file):
|
|
| 184 |
|
| 185 |
all_records.extend(records)
|
| 186 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 187 |
# Lưu vào parquet
|
| 188 |
parquet_files = save_to_parquet(all_records, final_output_dir)
|
| 189 |
|
|
@@ -192,7 +238,7 @@ def process_audio(input_file):
|
|
| 192 |
shutil.copytree(audio_output_dir, final_audio_dir)
|
| 193 |
|
| 194 |
# Tạo file zip
|
| 195 |
-
zip_path = os.path.join(
|
| 196 |
with zipfile.ZipFile(zip_path, 'w', zipfile.ZIP_DEFLATED) as zipf:
|
| 197 |
# Thêm audio files
|
| 198 |
for root, _, files in os.walk(final_audio_dir):
|
|
@@ -205,57 +251,206 @@ def process_audio(input_file):
|
|
| 205 |
for pq_file in parquet_files:
|
| 206 |
zipf.write(pq_file, os.path.basename(pq_file))
|
| 207 |
|
| 208 |
-
#
|
| 209 |
-
|
| 210 |
-
shutil.copy(zip_path, final_zip)
|
| 211 |
|
| 212 |
-
|
| 213 |
-
|
| 214 |
-
|
| 215 |
-
|
| 216 |
-
|
| 217 |
-
|
| 218 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 219 |
|
| 220 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 221 |
|
| 222 |
# Tạo giao diện Gradio
|
| 223 |
-
with gr.Blocks(title="Audio Transcription & Dataset Creator") as app:
|
| 224 |
gr.Markdown("""
|
| 225 |
-
# 🎙️ Audio Transcription & Dataset Creator
|
| 226 |
-
Upload file audio hoặc file zip
|
| 227 |
-
Hệ thống sẽ:
|
| 228 |
-
1. Transcribe bằng Whisper Large-v3-Turbo
|
| 229 |
-
2. Cắt audio theo timestamps (gộp câu ngắn)
|
| 230 |
-
3. Tạo dataset Parquet chuẩn với audio bytes
|
| 231 |
""")
|
| 232 |
|
| 233 |
-
with gr.
|
| 234 |
-
|
| 235 |
-
|
| 236 |
-
|
| 237 |
-
|
| 238 |
-
)
|
| 239 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 240 |
|
| 241 |
-
|
| 242 |
-
|
| 243 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 244 |
|
| 245 |
-
|
| 246 |
-
|
|
|
|
| 247 |
inputs=input_file,
|
| 248 |
-
outputs=[
|
| 249 |
)
|
| 250 |
|
| 251 |
-
|
| 252 |
-
|
| 253 |
-
|
| 254 |
-
|
| 255 |
-
|
| 256 |
-
|
| 257 |
-
|
| 258 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 259 |
|
| 260 |
if __name__ == "__main__":
|
| 261 |
app.launch()
|
|
|
|
| 8 |
import soundfile as sf
|
| 9 |
import pandas as pd
|
| 10 |
import numpy as np
|
| 11 |
+
from typing import List, Dict
|
| 12 |
import shutil
|
| 13 |
+
import threading
|
| 14 |
+
import time
|
| 15 |
+
from datetime import datetime
|
| 16 |
+
import json
|
| 17 |
+
import traceback
|
| 18 |
|
| 19 |
# Khởi tạo model Whisper
|
| 20 |
model = WhisperModel("large-v3-turbo", device="cpu", compute_type="int8")
|
| 21 |
|
| 22 |
+
# Lưu trữ tasks và history
|
| 23 |
+
TASKS = {}
|
| 24 |
+
TASK_LOCK = threading.Lock()
|
| 25 |
+
STORAGE_DIR = "task_storage"
|
| 26 |
+
os.makedirs(STORAGE_DIR, exist_ok=True)
|
| 27 |
+
|
| 28 |
+
class TaskStatus:
|
| 29 |
+
WAITING = "⏳ Đang chờ"
|
| 30 |
+
PROCESSING = "🔄 Đang xử lý"
|
| 31 |
+
SUCCESS = "✅ Thành công"
|
| 32 |
+
ERROR = "❌ Lỗi"
|
| 33 |
+
|
| 34 |
def extract_audio_files(input_file: str, temp_dir: str) -> List[str]:
|
| 35 |
"""Giải nén file zip hoặc copy file audio đơn"""
|
| 36 |
audio_files = []
|
|
|
|
| 81 |
current_duration = current['end'] - current['start']
|
| 82 |
|
| 83 |
if current_duration < min_duration:
|
|
|
|
| 84 |
current['end'] = seg['end']
|
| 85 |
current['text'] = current['text'] + ' ' + seg['text']
|
| 86 |
else:
|
|
|
|
| 137 |
parquet_files = []
|
| 138 |
|
| 139 |
if file_size_mb <= max_size_mb:
|
| 140 |
+
output_path = os.path.join(output_dir, 'train-00000-of-00001.parquet')
|
|
|
|
| 141 |
df.to_parquet(output_path, engine='pyarrow')
|
| 142 |
parquet_files.append(output_path)
|
| 143 |
else:
|
|
|
|
| 144 |
num_parts = int(np.ceil(file_size_mb / max_size_mb))
|
| 145 |
chunk_size = len(df) // num_parts + 1
|
| 146 |
|
|
|
|
| 149 |
end_idx = min((i + 1) * chunk_size, len(df))
|
| 150 |
|
| 151 |
df_chunk = df.iloc[start_idx:end_idx]
|
| 152 |
+
output_path = os.path.join(output_dir, f'train-{i:05d}-of-{num_parts:05d}.parquet')
|
| 153 |
df_chunk.to_parquet(output_path, engine='pyarrow')
|
| 154 |
parquet_files.append(output_path)
|
| 155 |
|
| 156 |
return parquet_files
|
| 157 |
|
| 158 |
+
def update_task_status(task_id: str, status: str, details: dict = None):
|
| 159 |
+
"""Cập nhật trạng thái task"""
|
| 160 |
+
with TASK_LOCK:
|
| 161 |
+
if task_id in TASKS:
|
| 162 |
+
TASKS[task_id]['status'] = status
|
| 163 |
+
TASKS[task_id]['updated_at'] = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
|
| 164 |
+
if details:
|
| 165 |
+
TASKS[task_id].update(details)
|
| 166 |
+
|
| 167 |
+
# Lưu vào file
|
| 168 |
+
with open(os.path.join(STORAGE_DIR, f"{task_id}.json"), 'w', encoding='utf-8') as f:
|
| 169 |
+
json.dump(TASKS[task_id], f, ensure_ascii=False, indent=2)
|
| 170 |
+
|
| 171 |
+
def process_audio_background(task_id: str, input_file: str, original_filename: str):
|
| 172 |
+
"""Xử lý audio trong background"""
|
| 173 |
+
try:
|
| 174 |
+
update_task_status(task_id, TaskStatus.PROCESSING, {
|
| 175 |
+
'progress': 'Đang giải nén và phát hiện file audio...'
|
| 176 |
+
})
|
| 177 |
+
|
| 178 |
+
task_dir = os.path.join(STORAGE_DIR, task_id)
|
| 179 |
+
os.makedirs(task_dir, exist_ok=True)
|
| 180 |
+
|
| 181 |
+
extract_dir = os.path.join(task_dir, 'extracted')
|
| 182 |
+
audio_output_dir = os.path.join(task_dir, 'audio')
|
| 183 |
+
final_output_dir = os.path.join(task_dir, 'output')
|
| 184 |
|
| 185 |
os.makedirs(extract_dir, exist_ok=True)
|
| 186 |
os.makedirs(audio_output_dir, exist_ok=True)
|
|
|
|
| 190 |
audio_files = extract_audio_files(input_file, extract_dir)
|
| 191 |
|
| 192 |
if not audio_files:
|
| 193 |
+
update_task_status(task_id, TaskStatus.ERROR, {
|
| 194 |
+
'error': 'Không tìm thấy file audio nào trong file tải lên!'
|
| 195 |
+
})
|
| 196 |
+
return
|
| 197 |
+
|
| 198 |
+
update_task_status(task_id, TaskStatus.PROCESSING, {
|
| 199 |
+
'progress': f'Tìm thấy {len(audio_files)} file audio. Đang transcribe...',
|
| 200 |
+
'total_files': len(audio_files)
|
| 201 |
+
})
|
| 202 |
|
| 203 |
all_records = []
|
| 204 |
|
| 205 |
# Xử lý từng file audio
|
| 206 |
+
for idx, audio_file in enumerate(audio_files):
|
| 207 |
+
update_task_status(task_id, TaskStatus.PROCESSING, {
|
| 208 |
+
'progress': f'Đang xử lý file {idx+1}/{len(audio_files)}: {Path(audio_file).name}'
|
| 209 |
+
})
|
| 210 |
+
|
| 211 |
base_name = Path(audio_file).stem
|
| 212 |
|
| 213 |
# Transcribe
|
|
|
|
| 226 |
|
| 227 |
all_records.extend(records)
|
| 228 |
|
| 229 |
+
update_task_status(task_id, TaskStatus.PROCESSING, {
|
| 230 |
+
'progress': f'Đã tạo {len(all_records)} segments. Đang lưu vào Parquet...'
|
| 231 |
+
})
|
| 232 |
+
|
| 233 |
# Lưu vào parquet
|
| 234 |
parquet_files = save_to_parquet(all_records, final_output_dir)
|
| 235 |
|
|
|
|
| 238 |
shutil.copytree(audio_output_dir, final_audio_dir)
|
| 239 |
|
| 240 |
# Tạo file zip
|
| 241 |
+
zip_path = os.path.join(task_dir, 'dataset_output.zip')
|
| 242 |
with zipfile.ZipFile(zip_path, 'w', zipfile.ZIP_DEFLATED) as zipf:
|
| 243 |
# Thêm audio files
|
| 244 |
for root, _, files in os.walk(final_audio_dir):
|
|
|
|
| 251 |
for pq_file in parquet_files:
|
| 252 |
zipf.write(pq_file, os.path.basename(pq_file))
|
| 253 |
|
| 254 |
+
# Tính kích thước file
|
| 255 |
+
zip_size_mb = os.path.getsize(zip_path) / (1024 * 1024)
|
|
|
|
| 256 |
|
| 257 |
+
update_task_status(task_id, TaskStatus.SUCCESS, {
|
| 258 |
+
'progress': 'Hoàn thành!',
|
| 259 |
+
'input_files': len(audio_files),
|
| 260 |
+
'total_segments': len(all_records),
|
| 261 |
+
'parquet_files': len(parquet_files),
|
| 262 |
+
'output_zip': zip_path,
|
| 263 |
+
'zip_size_mb': round(zip_size_mb, 2)
|
| 264 |
+
})
|
| 265 |
+
|
| 266 |
+
except Exception as e:
|
| 267 |
+
error_msg = f"{str(e)}\n\n{traceback.format_exc()}"
|
| 268 |
+
update_task_status(task_id, TaskStatus.ERROR, {
|
| 269 |
+
'error': error_msg
|
| 270 |
+
})
|
| 271 |
+
|
| 272 |
+
def submit_task(input_file):
|
| 273 |
+
"""Submit task mới"""
|
| 274 |
+
if input_file is None:
|
| 275 |
+
return "❌ Vui lòng upload file audio hoặc file zip!", ""
|
| 276 |
+
|
| 277 |
+
task_id = f"task_{int(time.time() * 1000)}"
|
| 278 |
+
original_filename = Path(input_file).name
|
| 279 |
+
|
| 280 |
+
task_info = {
|
| 281 |
+
'task_id': task_id,
|
| 282 |
+
'status': TaskStatus.WAITING,
|
| 283 |
+
'created_at': datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
|
| 284 |
+
'updated_at': datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
|
| 285 |
+
'original_filename': original_filename,
|
| 286 |
+
'progress': 'Task đã được tạo, đang chờ xử lý...'
|
| 287 |
+
}
|
| 288 |
+
|
| 289 |
+
with TASK_LOCK:
|
| 290 |
+
TASKS[task_id] = task_info
|
| 291 |
+
|
| 292 |
+
# Chạy background thread
|
| 293 |
+
thread = threading.Thread(
|
| 294 |
+
target=process_audio_background,
|
| 295 |
+
args=(task_id, input_file, original_filename),
|
| 296 |
+
daemon=True
|
| 297 |
+
)
|
| 298 |
+
thread.start()
|
| 299 |
+
|
| 300 |
+
return f"✅ Task {task_id} đã được tạo và đang xử lý trong background!", task_id
|
| 301 |
+
|
| 302 |
+
def load_all_tasks():
|
| 303 |
+
"""Load tất cả tasks từ storage"""
|
| 304 |
+
with TASK_LOCK:
|
| 305 |
+
for file in os.listdir(STORAGE_DIR):
|
| 306 |
+
if file.endswith('.json'):
|
| 307 |
+
task_id = file.replace('.json', '')
|
| 308 |
+
if task_id not in TASKS:
|
| 309 |
+
with open(os.path.join(STORAGE_DIR, file), 'r', encoding='utf-8') as f:
|
| 310 |
+
TASKS[task_id] = json.load(f)
|
| 311 |
+
|
| 312 |
+
def get_task_list():
|
| 313 |
+
"""Lấy danh sách tasks để hiển thị trong dropdown"""
|
| 314 |
+
load_all_tasks()
|
| 315 |
+
with TASK_LOCK:
|
| 316 |
+
task_list = [(f"{task['task_id']} - {task['status']} - {task['original_filename']}",
|
| 317 |
+
task['task_id'])
|
| 318 |
+
for task in sorted(TASKS.values(),
|
| 319 |
+
key=lambda x: x['created_at'],
|
| 320 |
+
reverse=True)]
|
| 321 |
+
return task_list
|
| 322 |
+
|
| 323 |
+
def get_task_info(task_id):
|
| 324 |
+
"""Lấy thông tin chi tiết của task"""
|
| 325 |
+
if not task_id:
|
| 326 |
+
return "Chọn task để xem thông tin", None
|
| 327 |
+
|
| 328 |
+
load_all_tasks()
|
| 329 |
+
|
| 330 |
+
with TASK_LOCK:
|
| 331 |
+
if task_id not in TASKS:
|
| 332 |
+
return "Task không tồn tại!", None
|
| 333 |
|
| 334 |
+
task = TASKS[task_id]
|
| 335 |
+
|
| 336 |
+
info = f"""
|
| 337 |
+
## 📋 Thông tin Task: {task_id}
|
| 338 |
+
|
| 339 |
+
**Trạng thái:** {task['status']}
|
| 340 |
+
**File gốc:** {task.get('original_filename', 'N/A')}
|
| 341 |
+
**Thời gian tạo:** {task['created_at']}
|
| 342 |
+
**Cập nhật lần cuối:** {task['updated_at']}
|
| 343 |
+
|
| 344 |
+
---
|
| 345 |
+
|
| 346 |
+
### 📊 Chi tiết
|
| 347 |
+
|
| 348 |
+
**Tiến trình:** {task.get('progress', 'N/A')}
|
| 349 |
+
"""
|
| 350 |
+
|
| 351 |
+
if task['status'] == TaskStatus.SUCCESS:
|
| 352 |
+
info += f"""
|
| 353 |
+
**Số file audio đầu vào:** {task.get('input_files', 'N/A')}
|
| 354 |
+
**Tổng số segments:** {task.get('total_segments', 'N/A')}
|
| 355 |
+
**Số file Parquet:** {task.get('parquet_files', 'N/A')}
|
| 356 |
+
**Kích thước ZIP:** {task.get('zip_size_mb', 'N/A')} MB
|
| 357 |
+
"""
|
| 358 |
+
zip_path = task.get('output_zip')
|
| 359 |
+
if zip_path and os.path.exists(zip_path):
|
| 360 |
+
return info, zip_path
|
| 361 |
+
|
| 362 |
+
elif task['status'] == TaskStatus.ERROR:
|
| 363 |
+
info += f"""
|
| 364 |
+
**Lỗi:**
|
| 365 |
+
```
|
| 366 |
+
{task.get('error', 'Unknown error')}
|
| 367 |
+
```
|
| 368 |
+
"""
|
| 369 |
+
|
| 370 |
+
return info, None
|
| 371 |
+
|
| 372 |
+
def refresh_task_list():
|
| 373 |
+
"""Refresh danh sách tasks"""
|
| 374 |
+
choices = get_task_list()
|
| 375 |
+
return gr.Dropdown(choices=choices, value=choices[0][1] if choices else None)
|
| 376 |
+
|
| 377 |
+
# Load tasks khi khởi động
|
| 378 |
+
load_all_tasks()
|
| 379 |
|
| 380 |
# Tạo giao diện Gradio
|
| 381 |
+
with gr.Blocks(title="Audio Transcription & Dataset Creator", theme=gr.themes.Soft()) as app:
|
| 382 |
gr.Markdown("""
|
| 383 |
+
# 🎙️ Audio Transcription & Dataset Creator with Background Processing
|
| 384 |
+
Upload file audio hoặc file zip - Hệ thống xử lý trong background và lưu lịch sử
|
|
|
|
|
|
|
|
|
|
|
|
|
| 385 |
""")
|
| 386 |
|
| 387 |
+
with gr.Tabs():
|
| 388 |
+
# Tab Upload
|
| 389 |
+
with gr.Tab("📤 Upload & Submit"):
|
| 390 |
+
gr.Markdown("### Tải lên file và submit task")
|
| 391 |
+
|
| 392 |
+
with gr.Row():
|
| 393 |
+
with gr.Column():
|
| 394 |
+
input_file = gr.File(
|
| 395 |
+
label="Upload Audio File hoặc ZIP",
|
| 396 |
+
file_types=['.wav', '.mp3', '.flac', '.ogg', '.m4a', '.aac', '.zip']
|
| 397 |
+
)
|
| 398 |
+
submit_btn = gr.Button("🚀 Submit Task", variant="primary", size="lg")
|
| 399 |
+
|
| 400 |
+
with gr.Column():
|
| 401 |
+
submit_status = gr.Textbox(label="📋 Trạng thái Submit", lines=3)
|
| 402 |
+
current_task_id = gr.Textbox(label="Task ID", visible=False)
|
| 403 |
+
|
| 404 |
+
gr.Markdown("""
|
| 405 |
+
### ℹ️ Hướng dẫn:
|
| 406 |
+
1. Upload file audio hoặc ZIP chứa nhiều file audio
|
| 407 |
+
2. Click "Submit Task" - task sẽ chạy trong background
|
| 408 |
+
3. Chuyển sang tab "History" để xem tiến trình và tải kết quả
|
| 409 |
+
""")
|
| 410 |
|
| 411 |
+
# Tab History
|
| 412 |
+
with gr.Tab("📜 History"):
|
| 413 |
+
gr.Markdown("### Xem lại lịch sử tasks và tải kết quả")
|
| 414 |
+
|
| 415 |
+
with gr.Row():
|
| 416 |
+
refresh_btn = gr.Button("🔄 Refresh", size="sm")
|
| 417 |
+
task_dropdown = gr.Dropdown(
|
| 418 |
+
label="Chọn Task",
|
| 419 |
+
choices=get_task_list(),
|
| 420 |
+
value=get_task_list()[0][1] if get_task_list() else None,
|
| 421 |
+
interactive=True
|
| 422 |
+
)
|
| 423 |
+
|
| 424 |
+
task_info_display = gr.Markdown("Chọn task để xem thông tin")
|
| 425 |
+
|
| 426 |
+
download_btn = gr.File(label="📦 Tải về Dataset ZIP")
|
| 427 |
+
|
| 428 |
+
# Auto refresh mỗi 3 giây
|
| 429 |
+
timer = gr.Timer(3)
|
| 430 |
|
| 431 |
+
# Event handlers
|
| 432 |
+
submit_btn.click(
|
| 433 |
+
fn=submit_task,
|
| 434 |
inputs=input_file,
|
| 435 |
+
outputs=[submit_status, current_task_id]
|
| 436 |
)
|
| 437 |
|
| 438 |
+
refresh_btn.click(
|
| 439 |
+
fn=refresh_task_list,
|
| 440 |
+
outputs=task_dropdown
|
| 441 |
+
)
|
| 442 |
+
|
| 443 |
+
task_dropdown.change(
|
| 444 |
+
fn=get_task_info,
|
| 445 |
+
inputs=task_dropdown,
|
| 446 |
+
outputs=[task_info_display, download_btn]
|
| 447 |
+
)
|
| 448 |
+
|
| 449 |
+
timer.tick(
|
| 450 |
+
fn=get_task_info,
|
| 451 |
+
inputs=task_dropdown,
|
| 452 |
+
outputs=[task_info_display, download_btn]
|
| 453 |
+
)
|
| 454 |
|
| 455 |
if __name__ == "__main__":
|
| 456 |
app.launch()
|