Files
push_fenxiangfengshi/local_tts.py
T

70 lines
2.3 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""本地 TTS 语音播报(IndexTTS2),仅 Windows 本地模式使用。
服务器/容器不会启用(无 torch、无声卡),由 fengxiang.tts.Speaker 在非 win32 平台直接跳过。
路径可通过环境变量覆盖,不再写死 E:\\:
FX_INDEXTTS_HOME IndexTTS2 工程目录
FX_INDEXTTS_VOICE 参考音色 wav 路径
"""
from __future__ import annotations
import os
import sys
import tempfile
import time
_INDEXTTS_HOME = os.environ.get("FX_INDEXTTS_HOME", r"E:\2025Code\python\IndexTT\index-tts")
_INDEXTTS_MODEL_DIR = os.path.join(_INDEXTTS_HOME, "checkpoints")
_INDEXTTS_CFG = os.path.join(_INDEXTTS_MODEL_DIR, "config.yaml")
_DEFAULT_VOICE_PROMPT = os.environ.get(
"FX_INDEXTTS_VOICE", os.path.join(_INDEXTTS_HOME, "examples", "voice_01.wav")
)
_tts_instance = None
def _get_tts():
"""延迟初始化 IndexTTS2(首次加载模型 1-2 分钟)。"""
global _tts_instance
if _tts_instance is None:
if not os.path.isdir(_INDEXTTS_HOME):
raise FileNotFoundError(
f"未找到 IndexTTS2 目录: {_INDEXTTS_HOME}(可用 FX_INDEXTTS_HOME 指定)"
)
sys.path.insert(0, _INDEXTTS_HOME)
from indextts.infer_v2 import IndexTTS2
print("[TTS] 正在加载 IndexTTS2 模型(首次较慢)...")
_tts_instance = IndexTTS2(
cfg_path=_INDEXTTS_CFG,
model_dir=_INDEXTTS_MODEL_DIR,
use_fp16=True,
use_cuda_kernel=False,
use_deepspeed=False,
)
print("[TTS] 模型加载完成")
return _tts_instance
def speak(text: str, voice_prompt: str | None = None) -> str:
"""合成并播放,返回生成的 wav 路径。"""
if sys.platform != "win32":
raise RuntimeError("本地 TTS 仅支持 Windows(依赖 winsound)")
import winsound
voice_prompt = voice_prompt or _DEFAULT_VOICE_PROMPT
if not os.path.exists(voice_prompt):
raise FileNotFoundError(f"参考音色不存在: {voice_prompt}")
tts = _get_tts()
out = os.path.join(tempfile.gettempdir(), f"fx_tts_{int(time.time() * 1000)}.wav")
tts.infer(spk_audio_prompt=voice_prompt, text=text, output_path=out, verbose=False)
winsound.PlaySound(out, winsound.SND_FILENAME)
return out
if __name__ == "__main__":
speak("本地语音播报测试成功")