Gated Linear Attention Transformers with Hardware-Efficient Trainingarxiv.org2 points·dataminer··0 commentsOpen articleSaveView on HN