# Core ML frameworks
torch
torchaudio
transformers

# Audio processing
librosa
soundfile
pydub
audioread

# Speech models and tools
speechbrain
resemblyzer
openai-whisper
silero-vad
demucs

# PyAnnote for diarization (requires HF token with model access)
pyannote.audio>=3.1.0

# Quality metrics
pesq
pystoi

# YouTube downloading
yt-dlp

# Data processing
numpy
scipy
pandas
scikit-learn
openpyxl

# GCP integration
google-cloud-storage

# Utilities
tqdm
pyyaml
python-dotenv
boto3

# Hugging Face dataset
datasets
pyarrow

# Additional audio utilities
webrtcvad
noisereduce
panns-inference