Ubuntu
47 lines
1.3 KiB
Python
47 lines
1.3 KiB
Python
#!/usr/bin/env python3
|
|
import warnings
|
|
warnings.filterwarnings('ignore')
|
|
|
|
from funasr import AutoModel
|
|
import os
|
|
import subprocess
|
|
|
|
print("🔄 加载 FunASR 模型...")
|
|
print(" 首次会自动下载模型（约100MB）到 ~/.cache/modelscope")
|
|
print(" 请等待...\n")
|
|
|
|
model = AutoModel(
|
|
model="paraformer-zh",
|
|
vad_model="fsmn-vad",
|
|
punc_model="ct-punc",
|
|
)
|
|
|
|
print("\n✅ FunASR 模型加载完成！")
|
|
|
|
# 测试文件
|
|
uploads_dir = 'uploads'
|
|
if os.path.exists(uploads_dir):
|
|
files = [f for f in os.listdir(uploads_dir) if f.endswith(('.webm', '.wav', '.mp3'))]
|
|
if files:
|
|
test_file = os.path.join(uploads_dir, files[0])
|
|
print(f"\n🎵 测试文件: {test_file}")
|
|
|
|
# 转换为 wav
|
|
wav_file = test_file + '.wav'
|
|
subprocess.run(['ffmpeg', '-y', '-i', test_file, '-ar', '16000', '-ac', '1', wav_file],
|
|
capture_output=True)
|
|
|
|
print("📝 开始识别...")
|
|
res = model.generate(input=wav_file, batch_size_s=300)
|
|
|
|
text = res[0]["text"]
|
|
print(f"\n✅ 识别结果:")
|
|
print(f" {text}")
|
|
|
|
if os.path.exists(wav_file):
|
|
os.remove(wav_file)
|
|
else:
|
|
print("\n⚠️ 没有测试音频")
|
|
else:
|
|
print("\n⚠️ uploads 目录不存在")
|