openvid-wantrack-processed

OpenVid-WanTrack — preprocessed training data Preprocessed (FastVideo parquet) data used to train the TrackWan point-track-conditioned video model. Derived from OpenVid; each row is one 121-frame clip with everything the trainer consumes, so it is self-contained (no re-encoding needed). Contents (parquet columns) Each field is stored as bytes / shape / dtype (numpy buffers): vae latent — Wan VAE latent of the clip first frame latent — I2V conditioning… See the full description on the dataset page:

Source: Hugging Face Hub (noctuashap/openvid-wantrack-processed). Metadata imported from the dataset’s Hub tags.