requirements.txt

# Hugging Face core libraries
transformers
huggingface_hub
accelerate
fastvideo

# Qwen Omni helper (handles multimodal processing)
qwen-omni-utils[decord]

# PyTorch (for model execution)
torch
torchvision
torchaudio

# UI with chat interface
gradio

# Audio file writing (used for speech output)
soundfile

# Video/audio decoding
ffmpeg-python
decord
moviepy

# Utility libraries
numpy