Integrate eou and vad
This commit is contained in:
Xin Wang
126
models/config.py
Normal file
126
models/config.py
Normal file
@@ -0,0 +1,126 @@
|
||||
"""Configuration models for call options."""
|
||||
|
||||
from typing import Optional, Dict, Any, List
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
|
||||
class VADOption(BaseModel):
|
||||
"""Voice Activity Detection configuration."""
|
||||
|
||||
type: str = Field(default="silero", description="VAD algorithm type (silero, webrtc)")
|
||||
samplerate: int = Field(default=16000, description="Audio sample rate for VAD")
|
||||
speech_padding: int = Field(default=250, description="Speech padding in milliseconds")
|
||||
silence_padding: int = Field(default=100, description="Silence padding in milliseconds")
|
||||
ratio: float = Field(default=0.5, description="Voice detection ratio threshold")
|
||||
voice_threshold: float = Field(default=0.5, description="Voice energy threshold")
|
||||
max_buffer_duration_secs: int = Field(default=50, description="Maximum buffer duration in seconds")
|
||||
silence_timeout: Optional[int] = Field(default=None, description="Silence timeout in milliseconds")
|
||||
endpoint: Optional[str] = Field(default=None, description="Custom VAD service endpoint")
|
||||
secret_key: Optional[str] = Field(default=None, description="VAD service secret key")
|
||||
secret_id: Optional[str] = Field(default=None, description="VAD service secret ID")
|
||||
|
||||
|
||||
class ASROption(BaseModel):
|
||||
"""Automatic Speech Recognition configuration."""
|
||||
|
||||
provider: str = Field(..., description="ASR provider (tencent, aliyun, openai, etc.)")
|
||||
language: Optional[str] = Field(default=None, description="Language code (zh-CN, en-US)")
|
||||
app_id: Optional[str] = Field(default=None, description="Application ID")
|
||||
secret_id: Optional[str] = Field(default=None, description="Secret ID for authentication")
|
||||
secret_key: Optional[str] = Field(default=None, description="Secret key for authentication")
|
||||
model_type: Optional[str] = Field(default=None, description="ASR model type (16k_zh, 8k_en)")
|
||||
buffer_size: Optional[int] = Field(default=None, description="Audio buffer size in bytes")
|
||||
samplerate: Optional[int] = Field(default=None, description="Audio sample rate")
|
||||
endpoint: Optional[str] = Field(default=None, description="Custom ASR service endpoint")
|
||||
extra: Optional[Dict[str, Any]] = Field(default=None, description="Additional parameters")
|
||||
start_when_answer: bool = Field(default=False, description="Start ASR when call is answered")
|
||||
|
||||
|
||||
class TTSOption(BaseModel):
|
||||
"""Text-to-Speech configuration."""
|
||||
|
||||
samplerate: Optional[int] = Field(default=None, description="TTS output sample rate")
|
||||
provider: str = Field(default="msedge", description="TTS provider (tencent, aliyun, deepgram, msedge)")
|
||||
speed: float = Field(default=1.0, description="Speech speed multiplier")
|
||||
app_id: Optional[str] = Field(default=None, description="Application ID")
|
||||
secret_id: Optional[str] = Field(default=None, description="Secret ID for authentication")
|
||||
secret_key: Optional[str] = Field(default=None, description="Secret key for authentication")
|
||||
volume: Optional[int] = Field(default=None, description="Speech volume level (1-10)")
|
||||
speaker: Optional[str] = Field(default=None, description="Voice speaker name")
|
||||
codec: Optional[str] = Field(default=None, description="Audio codec")
|
||||
subtitle: bool = Field(default=False, description="Enable subtitle generation")
|
||||
emotion: Optional[str] = Field(default=None, description="Speech emotion")
|
||||
endpoint: Optional[str] = Field(default=None, description="Custom TTS service endpoint")
|
||||
extra: Optional[Dict[str, Any]] = Field(default=None, description="Additional parameters")
|
||||
max_concurrent_tasks: Optional[int] = Field(default=None, description="Max concurrent tasks")
|
||||
|
||||
|
||||
class RecorderOption(BaseModel):
|
||||
"""Call recording configuration."""
|
||||
|
||||
recorder_file: str = Field(..., description="Path to recording file")
|
||||
samplerate: int = Field(default=16000, description="Recording sample rate")
|
||||
ptime: int = Field(default=200, description="Packet time in milliseconds")
|
||||
|
||||
|
||||
class MediaPassOption(BaseModel):
|
||||
"""Media pass-through configuration for external audio processing."""
|
||||
|
||||
url: str = Field(..., description="WebSocket URL for media streaming")
|
||||
input_sample_rate: int = Field(default=16000, description="Sample rate of audio received from WebSocket")
|
||||
output_sample_rate: int = Field(default=16000, description="Sample rate of audio sent to WebSocket")
|
||||
packet_size: int = Field(default=2560, description="Packet size in bytes")
|
||||
ptime: Optional[int] = Field(default=None, description="Buffered playback period in milliseconds")
|
||||
|
||||
|
||||
class SipOption(BaseModel):
|
||||
"""SIP protocol configuration."""
|
||||
|
||||
username: Optional[str] = Field(default=None, description="SIP username")
|
||||
password: Optional[str] = Field(default=None, description="SIP password")
|
||||
realm: Optional[str] = Field(default=None, description="SIP realm/domain")
|
||||
headers: Optional[Dict[str, str]] = Field(default=None, description="Additional SIP headers")
|
||||
|
||||
|
||||
class HandlerRule(BaseModel):
|
||||
"""Handler routing rule."""
|
||||
|
||||
caller: Optional[str] = Field(default=None, description="Caller pattern (regex)")
|
||||
callee: Optional[str] = Field(default=None, description="Callee pattern (regex)")
|
||||
playbook: Optional[str] = Field(default=None, description="Playbook file path")
|
||||
webhook: Optional[str] = Field(default=None, description="Webhook URL")
|
||||
|
||||
|
||||
class CallOption(BaseModel):
|
||||
"""Comprehensive call configuration options."""
|
||||
|
||||
# Basic options
|
||||
denoise: bool = Field(default=False, description="Enable noise reduction")
|
||||
offer: Optional[str] = Field(default=None, description="SDP offer string")
|
||||
callee: Optional[str] = Field(default=None, description="Callee SIP URI or phone number")
|
||||
caller: Optional[str] = Field(default=None, description="Caller SIP URI or phone number")
|
||||
|
||||
# Audio codec
|
||||
codec: str = Field(default="pcm", description="Audio codec (pcm, pcma, pcmu, g722)")
|
||||
|
||||
# Component configurations
|
||||
recorder: Optional[RecorderOption] = Field(default=None, description="Call recording config")
|
||||
asr: Optional[ASROption] = Field(default=None, description="ASR configuration")
|
||||
vad: Optional[VADOption] = Field(default=None, description="VAD configuration")
|
||||
tts: Optional[TTSOption] = Field(default=None, description="TTS configuration")
|
||||
media_pass: Optional[MediaPassOption] = Field(default=None, description="Media pass-through config")
|
||||
sip: Optional[SipOption] = Field(default=None, description="SIP configuration")
|
||||
|
||||
# Timeouts and networking
|
||||
handshake_timeout: Optional[int] = Field(default=None, description="Handshake timeout in seconds")
|
||||
enable_ipv6: bool = Field(default=False, description="Enable IPv6 support")
|
||||
inactivity_timeout: Optional[int] = Field(default=None, description="Inactivity timeout in seconds")
|
||||
|
||||
# EOU configuration
|
||||
eou: Optional[Dict[str, Any]] = Field(default=None, description="End of utterance detection config")
|
||||
|
||||
# Extra parameters
|
||||
extra: Optional[Dict[str, Any]] = Field(default=None, description="Additional custom parameters")
|
||||
|
||||
class Config:
|
||||
populate_by_name = True
|
||||
Reference in New Issue
Block a user