Spaces:
Runtime error
Runtime error
| import edge_tts | |
| import asyncio | |
| import librosa | |
| import soundfile | |
| import io | |
| from inference.infer_tool import Svc | |
| TEXT = "私はヘルタ。今は忙しいから、リモート人形のオート返答機能に任せる。こんにちは、こんにちは、ごきげんよう、良い日になりますように。それじゃ" | |
| VOICE = "ja-JP-NanamiNeural" | |
| OUTPUT_FILE = "test.mp3" | |
| asyncio.run(edge_tts.Communicate(TEXT, VOICE).save(OUTPUT_FILE)) | |
| audio, sr = librosa.load(OUTPUT_FILE, sr=16000, mono=True) | |
| raw_path = io.BytesIO() | |
| soundfile.write(raw_path, audio, 16000, format="wav") | |
| raw_path.seek(0) | |
| print('checkpoint 1') | |
| model = Svc(fr"Herta-Svc/G_10000.pth", f"Herta-Svc/config.json", device = 'cpu') | |
| print('checkpoint 2') | |
| out_audio, out_sr = model.infer('speaker0', 0, raw_path, | |
| auto_predict_f0 = True, | |
| ) | |
| print('checkpoint 3') | |
| soundfile.write('out_audio.wav', out_audio.cpu().numpy(), 44100) | |
| print("done") |