Skip to yearly menu bar Skip to main content


Transformer-VQ: Linear-Time Transformers via Vector Quantization

Lucas D. Lingle

Halle B #42
[ ] [ Project Page ]
Thu 9 May 7:30 a.m. PDT — 9:30 a.m. PDT


We introduce Transformer-VQ, a decoder-only transformer computing softmax-based dense self-attention in linear time. Transformer-VQ's efficient attention is enabled by vector-quantized keys and a novel caching mechanism. In our large-scale experiments, Transformer-VQ is shown highly competitive in quality, obtaining 0.99 bpb on Enwik8, 26.6 ppl on PG-19, and 3.16 bpb on ImageNet64. In addition, the optimized implementation of Transformer-VQ is over 3x faster than a comparable quadratic-time transformer at sequence length 8k, is over 12x faster at 32k, and can scale to 131k with similar throughput. Code available: \url{}

Chat is not available.