mpeg-diffusion / requirements.txt
multimodalart's picture
multimodalart HF Staff
Upload folder using huggingface_hub
6febba8 verified
Raw History Blame Contribute Delete
551 Bytes
# Vendored lot/ + third_party/musubi_tuner subset pulls these on the load path:
# - av (minimax_h3/media.py + dataset/*, module-level)
# - opencv (dataset/media_utils.py via cv2_compat — headless, no GUI deps)
# - accelerate (init_empty_weights in model.py / video_vae.py)
# - imageio(+ffmpeg) for two_track_h3.save_clip's mp4 writer
# gradio, spaces, huggingface_hub and torch are preinstalled on ZeroGPU — not pinned here.
accelerate
safetensors>=0.8.0
numpy
pillow
tqdm
av
opencv-python-headless
imageio
imageio-ffmpeg