Michael Hu
default to parakeet model as it is much faster
31708ca
"""
Speech Recognition Module
Supports multiple ASR models including Whisper and Parakeet
Handles audio preprocessing and transcription
"""
import logging
import numpy as np
import os
from abc import ABC, abstractmethod
logger = logging.getLogger(__name__)
import torch
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor
from pydub import AudioSegment
import soundfile as sf
class ASRModel(ABC):
"""Base class for ASR models"""
@abstractmethod
def load_model(self):
"""Load the ASR model"""
pass
@abstractmethod
def transcribe(self, audio_path):
"""Transcribe audio to text"""
pass
def preprocess_audio(self, audio_path):
"""Convert audio to required format"""
logger.info("Converting audio format")
audio = AudioSegment.from_file(audio_path)
processed_audio = audio.set_frame_rate(16000).set_channels(1)
wav_path = audio_path.replace(".mp3", ".wav") if audio_path.endswith(".mp3") else audio_path
if not wav_path.endswith(".wav"):
wav_path = f"{os.path.splitext(wav_path)[0]}.wav"
processed_audio.export(wav_path, format="wav")
logger.info(f"Audio converted to: {wav_path}")
return wav_path
class WhisperModel(ASRModel):
"""Whisper ASR model implementation"""
def __init__(self):
self.model = None
self.processor = None
self.device = "cuda" if torch.cuda.is_available() else "cpu"
def load_model(self):
"""Load Whisper model"""
logger.info("Loading Whisper model")
logger.info(f"Using device: {self.device}")
self.model = AutoModelForSpeechSeq2Seq.from_pretrained(
"openai/whisper-large-v3",
torch_dtype=torch.float32,
low_cpu_mem_usage=True,
use_safetensors=True
).to(self.device)
self.processor = AutoProcessor.from_pretrained("openai/whisper-large-v3")
logger.info("Whisper model loaded successfully")
def transcribe(self, audio_path):
"""Transcribe audio using Whisper"""
if self.model is None or self.processor is None:
self.load_model()
wav_path = self.preprocess_audio(audio_path)
# Processing
logger.info("Processing audio input")
logger.debug("Loading audio data")
audio_data, sample_rate = sf.read(wav_path)
audio_data = audio_data.astype(np.float32)
# Increase chunk length and stride for longer transcriptions
inputs = self.processor(
audio_data,
sampling_rate=16000,
return_tensors="pt",
# Increase chunk length to handle longer segments
chunk_length_s=60,
stride_length_s=10
).to(self.device)
# Transcription
logger.info("Generating transcription")
with torch.no_grad():
# Add max_length parameter to allow for longer outputs
outputs = self.model.generate(
**inputs,
language="en",
task="transcribe",
max_length=448, # Explicitly set max output length
no_repeat_ngram_size=3 # Prevent repetition in output
)
result = self.processor.batch_decode(outputs, skip_special_tokens=True)[0]
logger.info(f"Transcription completed successfully")
return result
class ParakeetModel(ASRModel):
"""Parakeet ASR model implementation"""
def __init__(self):
self.model = None
def load_model(self):
"""Load Parakeet model"""
try:
import nemo.collections.asr as nemo_asr
logger.info("Loading Parakeet model")
self.model = nemo_asr.models.ASRModel.from_pretrained(model_name="nvidia/parakeet-tdt-0.6b-v2")
logger.info("Parakeet model loaded successfully")
except ImportError:
logger.error("Failed to import nemo_toolkit. Please install with: pip install -U 'nemo_toolkit[asr]'")
raise
def transcribe(self, audio_path):
"""Transcribe audio using Parakeet"""
if self.model is None:
self.load_model()
wav_path = self.preprocess_audio(audio_path)
# Transcription
logger.info("Generating transcription with Parakeet")
output = self.model.transcribe([wav_path])
result = output[0].text
logger.info(f"Transcription completed successfully")
return result
class ASRFactory:
"""Factory for creating ASR model instances"""
@staticmethod
def get_model(model_name="parakeet"):
"""
Get ASR model by name
Args:
model_name: Name of the model to use (whisper or parakeet)
Returns:
ASR model instance
"""
if model_name.lower() == "whisper":
return WhisperModel()
elif model_name.lower() == "parakeet":
return ParakeetModel()
else:
logger.warning(f"Unknown model: {model_name}, falling back to Whisper")
return WhisperModel()
def transcribe_audio(audio_path, model_name="parakeet"):
"""
Convert audio file to text using specified ASR model
Args:
audio_path: Path to input audio file
model_name: Name of the ASR model to use (whisper or parakeet)
Returns:
Transcribed English text
"""
logger.info(f"Starting transcription for: {audio_path} using {model_name} model")
try:
# Get the appropriate model
asr_model = ASRFactory.get_model(model_name)
# Transcribe audio
result = asr_model.transcribe(audio_path)
logger.info(f"transcription: %s" % result)
return result
except Exception as e:
logger.error(f"Transcription failed: {str(e)}", exc_info=True)
raise