Spaces:
Sleeping
Sleeping
File size: 6,422 Bytes
5065491 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 |
"""
Enhanced Audio Services for Conversational Voice Bot
Based on friend's implementation with Murf TTS and Groq ASR
"""
from groq import AsyncGroq
from config import GROQ_API_KEY, MURF_API_KEY
import asyncio
import logging
import re
logger = logging.getLogger(__name__)
# Initialize Groq client
groq_client = AsyncGroq(api_key=GROQ_API_KEY)
# Initialize Murf client
try:
from murf import AsyncMurf
murf_client = AsyncMurf(api_key=MURF_API_KEY)
MURF_AVAILABLE = True
except ImportError:
logger.warning("Murf package not available. Install with: pip install murf")
MURF_AVAILABLE = False
def clean_markdown_for_tts(text: str) -> str:
"""
Clean markdown and formatting from text to make it suitable for TTS
"""
if not text:
return ""
# Remove markdown links [text](url) -> text
text = re.sub(r'\[([^\]]+)\]\([^\)]+\)', r'\1', text)
# Remove markdown emphasis **bold** and *italic* -> text
text = re.sub(r'\*\*([^\*]+)\*\*', r'\1', text)
text = re.sub(r'\*([^\*]+)\*', r'\1', text)
# Remove markdown headers ### ->
text = re.sub(r'^#+\s*', '', text, flags=re.MULTILINE)
# Remove code blocks ```code``` -> code
text = re.sub(r'```[^`]*```', '', text)
text = re.sub(r'`([^`]+)`', r'\1', text)
# Remove HTML tags
text = re.sub(r'<[^>]+>', '', text)
# Clean up bullet points and lists
text = re.sub(r'^\s*[-\*\+]\s*', '', text, flags=re.MULTILINE)
text = re.sub(r'^\s*\d+\.\s*', '', text, flags=re.MULTILINE)
# Remove extra whitespace and line breaks
text = re.sub(r'\n\s*\n', '. ', text)
text = re.sub(r'\s+', ' ', text)
# Remove special characters that TTS might struggle with
text = re.sub(r'[#\*_~`]', '', text)
return text.strip()
async def groq_asr_bytes(audio_bytes: bytes, model: str = "whisper-large-v3", language: str = "en") -> str:
"""
Transcribe audio using Groq Whisper ASR
Enhanced version similar to friend's implementation
"""
try:
logger.info(f"π€ Transcribing audio with Groq ASR (model: {model}, language: {language})")
# Groq client is async, so we can use it directly
response = await groq_client.audio.transcriptions.create(
model=model,
file=("audio.wav", audio_bytes, "audio/wav"),
response_format="text",
language=language,
temperature=0.0 # For more consistent results
)
transcription = response.strip() if response else ""
logger.info(f"π― Transcription result: {transcription}")
return transcription
except Exception as e:
logger.error(f"β Groq ASR failed: {e}")
return ""
async def murf_tts(text: str, voice_id: str = "en-IN-isha", format: str = "MP3") -> bytes:
"""
Convert text to speech using Murf TTS
Enhanced version similar to friend's implementation
"""
if not MURF_AVAILABLE:
logger.error("β Murf TTS not available")
return b""
if not text or not text.strip():
logger.warning("β οΈ Empty text provided to TTS")
return b""
try:
# Clean text for TTS
clean_text = clean_markdown_for_tts(text)
if not clean_text:
logger.warning("β οΈ Text became empty after cleaning")
return b""
logger.info(f"π Generating speech with Murf TTS (voice: {voice_id})")
logger.debug(f"TTS text: {clean_text}")
# Generate speech using async streaming
response = murf_client.text_to_speech.stream(
text=clean_text,
voice_id=voice_id,
format=format,
sample_rate=44100.0
)
# Collect all chunks
chunks = [chunk async for chunk in response]
full_audio = b''.join(chunks)
logger.info(f"β
Generated {len(full_audio)} bytes of audio")
return full_audio
except Exception as e:
logger.error(f"β Murf TTS failed: {e}")
return b""
async def edge_tts_fallback(text: str, voice: str = "en-IN-Neerja") -> bytes:
"""
Fallback TTS using edge-tts if Murf is not available
"""
try:
import edge_tts
clean_text = clean_markdown_for_tts(text)
if not clean_text:
return b""
logger.info(f"π Using Edge TTS fallback (voice: {voice})")
communicate = edge_tts.Communicate(clean_text, voice)
audio_chunks = []
async for chunk in communicate.stream():
if chunk["type"] == "audio":
audio_chunks.append(chunk["data"])
audio_data = b"".join(audio_chunks)
logger.info(f"β
Generated {len(audio_data)} bytes of audio with Edge TTS")
return audio_data
except ImportError:
logger.error("β Edge TTS not available. Install with: pip install edge-tts")
return b""
except Exception as e:
logger.error(f"β Edge TTS failed: {e}")
return b""
class ConversationalAudioService:
"""
Main audio service class for conversational voice bot
"""
def __init__(self):
self.groq_client = groq_client
self.murf_available = MURF_AVAILABLE
self.default_voice = "en-IN-isha" # Indian English voice
async def transcribe_audio(self, audio_bytes: bytes, language: str = "en") -> str:
"""Transcribe audio to text using Groq ASR"""
return await groq_asr_bytes(audio_bytes, language=language)
async def synthesize_speech(self, text: str, voice_id: str = None) -> bytes:
"""Convert text to speech using best available TTS"""
voice = voice_id or self.default_voice
if self.murf_available:
# Try Murf TTS first
audio = await murf_tts(text, voice_id=voice)
if audio:
return audio
# Fallback to Edge TTS
return await edge_tts_fallback(text, voice="en-IN-Neerja")
def set_default_voice(self, voice_id: str):
"""Set default voice for TTS"""
self.default_voice = voice_id
logger.info(f"π΅ Default voice set to: {voice_id}")
# Global audio service instance
conversational_audio_service = ConversationalAudioService() |