Implementation of Trajectory Transformer with attention caching and batched beam search
Primary LanguagePythonMIT LicenseMIT