Bibliography (5):

  1. Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention

  2. https://huggingface.co/datasets/google/fleurs

  3. https://github.com/facebookresearch/seamless_communication