Hass/xiaozhi-esp32-server
1
0
Fork 0
mirror of https://github.com/xinnan-tech/xiaozhi-esp32-server.git synced 2026-07-30 05:13:59 +08:00
Code Issues Packages Projects Releases Wiki Activity

mergin main，速度提升一下

Browse Source
This commit is contained in:
lizhongxiang
2025-03-20 09:39:25 +08:00
parent a067a6ba87 4aefbe5aa8
commit b153bc11f2
96 changed files with 3438 additions and 1830 deletions
Download Patch File Download Diff File Expand all files Collapse all files
main/xiaozhi-server/core/providers/tts/base.py
+10 -9
View File
@@ -14,7 +14,7 @@ logger = setup_logging()
class TTSProviderBase(ABC):
def __init__(self, config, delete_audio_file):
self.delete_audio_file = delete_audio_file
self.output_file = config.get("output_file")
self.output_file = config.get("output_dir")
@abstractmethod
def generate_filename(self):
@@ -35,7 +35,7 @@ class TTSProviderBase(ABC):
return tmp_file
except Exception as e:
logger.bind(tag=TAG).info(f": {e}")
logger.bind(tag=TAG).info(f"Failed to generate TTS file: {e}")
return None
def to_tts_stream(self, text, queue: queue.Queue, text_index=0):
@@ -52,19 +52,20 @@ class TTSProviderBase(ABC):
async def text_to_speak_stream(self, text, queue: queue.Queue, text_index=0):
raise Exception("该TTS还没有实现stream模式")
def wav_to_opus_data(self, wav_file_path):
# 使用pydub加载PCM文件
def audio_to_opus_data(self, audio_file_path):
"""音频文件转换为Opus编码"""
# 获取文件后缀名
file_type = os.path.splitext(wav_file_path)[1]
file_type = os.path.splitext(audio_file_path)[1]
if file_type:
file_type = file_type.lstrip('.')
audio = AudioSegment.from_file(wav_file_path, format=file_type)
audio = AudioSegment.from_file(audio_file_path, format=file_type)
# 转换为单声道/16kHz采样率/16位小端编码（确保与编码器匹配）
audio = audio.set_channels(1).set_frame_rate(16000).set_sample_width(2)
# 音频时长(秒)
duration = len(audio) / 1000.0
# 转换为单声道和16kHz采样率（确保与编码器匹配）
audio = audio.set_channels(1).set_frame_rate(16000)
# 获取原始PCM数据（16位小端）
raw_data = audio.raw_data
main/xiaozhi-server/core/providers/tts/custom.py
+35
View File
@@ -0,0 +1,35 @@
import os
import uuid
import requests
from config.logger import setup_logging
from datetime import datetime
from core.providers.tts.base import TTSProviderBase
TAG = __name__
logger = setup_logging()
class TTSProvider(TTSProviderBase):
def __init__(self, config, delete_audio_file):
super().__init__(config, delete_audio_file)
self.url = config.get("url")
self.headers = config.get("headers", {})
self.params = config.get("params")
self.format = config.get("format", "wav")
self.output_file = config.get("output_dir", "tmp/")
def generate_filename(self):
return os.path.join(self.output_file, f"tts-{datetime.now().date()}@{uuid.uuid4().hex}.{self.format}")
async def text_to_speak(self, text, output_file):
request_params = {}
for k, v in self.params.items():
if isinstance(v, str) and "{prompt_text}" in v:
v = v.replace("{prompt_text}", text)
request_params[k] = v
resp = requests.get(self.url, params=request_params, headers=self.headers)
if resp.status_code == 200:
with open(output_file, "wb") as file:
file.write(resp.content)
else:
logger.bind(tag=TAG).error(f"Custom TTS请求失败: {resp.status_code} - {resp.text}")
main/xiaozhi-server/core/providers/tts/fishspeech.py
+13 -11
View File
@@ -177,18 +177,8 @@ class TTSProvider(TTSProviderBase):
async def text_to_speak_stream(self, text, queue: queue.Queue, text_index=0):
try:
# Prepare reference data
byte_audios = [audio_to_bytes(ref_audio) for ref_audio in self.reference_audio]
ref_texts = [read_ref_text(ref_text) for ref_text in self.reference_text]
data = {
"text": text,
"references": [
ServeReferenceAudio(
audio=audio if audio else b"", text=text
)
for text, audio in zip(ref_texts, byte_audios)
],
"reference_id": self.reference_id,
"normalize": self.normalize,
"format": self.format,
@@ -202,6 +192,18 @@ class TTSProvider(TTSProviderBase):
"seed": self.seed,
}
# Prepare reference data
if self.reference_audio and self.reference_text:
byte_audios = [audio_to_bytes(ref_audio) for ref_audio in self.reference_audio]
ref_texts = [read_ref_text(ref_text) for ref_text in self.reference_text]
data["references"] = [
ServeReferenceAudio(
audio=audio if audio else b"", text=text
)
for text, audio in zip(ref_texts, byte_audios)
],
data["reference_id"] = None
pydantic_data = ServeTTSRequest(**data)
audio_buff = None
chunk_total = b''
@@ -224,7 +226,7 @@ class TTSProvider(TTSProviderBase):
if len(chunk_total) % 2 == 0 and chunk_total[-2:] == b'\x00\x00':
audio = self._get_audio_from_tts(chunk_total)
audio_raw = audio_raw + audio.raw_data
#长度凑够2贞开始发送，60ms*4=240ms
# 长度凑够2贞开始发送，60ms*4=240ms
if len(audio_raw) >= 7680:
duration = 60 * len(audio_raw) // 1920
if (len(audio_raw) % 1920) > 0:
main/xiaozhi-server/core/providers/tts/gpt_sovits_v3.py
+19 -17
View File
@@ -12,17 +12,18 @@ class TTSProvider(TTSProviderBase):
def __init__(self, config, delete_audio_file):
super().__init__(config, delete_audio_file)
self.url = config.get("url")
self.text_lang = config.get("text_lang", "audo")
self.ref_audio_path = config.get("ref_audio_path")
self.prompt_lang = config.get("prompt_lang")
self.refer_wav_path = config.get("refer_wav_path")
self.prompt_text = config.get("prompt_text")
self.top_k = config.get("top_k", 5)
self.top_p = config.get("top_p", 1)
self.temperature = config.get("temperature", 1)
self.sample_steps = config.get("sample_steps", 16)
self.media_type = config.get("media_type", "wav")
self.streaming_mode = config.get("streaming_mode", False)
self.threshold = config.get("threshold", 30)
self.prompt_language = config.get("prompt_language")
self.text_language = config.get("text_language", "audo")
self.top_k = config.get("top_k", 15)
self.top_p = config.get("top_p", 1.0)
self.temperature = config.get("temperature", 1.0)
self.cut_punc = config.get("cut_punc","")
self.speed = config.get("speed", 1.0)
self.inp_refs = config.get("inp_refs",[])
self.sample_steps = config.get("sample_steps",32)
self.if_sr = config.get("if_sr",False)
def generate_filename(self, extension=".wav"):
@@ -30,18 +31,19 @@ class TTSProvider(TTSProviderBase):
async def text_to_speak(self, text, output_file):
request_params = {
"text": text,
"text_lang": self.text_lang,
"ref_audio_path": self.ref_audio_path,
"prompt_lang": self.prompt_lang,
"refer_wav_path": self.refer_wav_path,
"prompt_text": self.prompt_text,
"prompt_language": self.prompt_language,
"text": text,
"text_language": self.text_language,
"top_k": self.top_k,
"top_p": self.top_p,
"temperature": self.temperature,
"cut_punc": self.cut_punc,
"speed": self.speed,
"inp_refs": self.inp_refs,
"sample_steps": self.sample_steps,
"media_type": self.media_type,
"streaming_mode": self.streaming_mode,
"threshold": self.threshold,
"if_sr": self.if_sr,
}
resp = requests.get(self.url, params=request_params)
main/xiaozhi-server/core/providers/tts/openai.py
+1 -1
View File
@@ -14,7 +14,7 @@ class TTSProvider(TTSProviderBase):
self.voice = config.get("voice", "alloy")
self.response_format = "wav"
self.speed = config.get("speed", 1.0)
self.output_file = config.get("output_file", "tmp/")
self.output_file = config.get("output_dir", "tmp/")
check_model_key("TTS", self.api_key)
def generate_filename(self, extension=".wav"):
main/xiaozhi-server/core/providers/tts/ttson.py
+1 -1
View File
@@ -17,7 +17,7 @@ class TTSProvider(TTSProviderBase):
self.volume_change_dB = config.get("volume_change_dB", 0)
self.speed_factor = config.get("speed_factor", 1)
self.stream = config.get("stream", False)
self.output_file = config.get("output_file")
self.output_file = config.get("output_dir")
self.pitch_factor = config.get("pitch_factor", 0)
self.format = config.get("format", "mp3")
self.emotion = config.get("emotion", 1)