Transformer Quality in Linear Time
read the original abstract
We revisit the design choices in Transformers, and propose methods to address their weaknesses in handling long sequences. First, we propose a simple layer named gated attention unit, which allows the use of a weaker single-head attention with minimal quality loss. We then propose a linear approximation method complementary to this new layer, which is accelerator-friendly and highly competitive in quality. The resulting model, named FLASH, matches the perplexity of improved Transformers over both short (512) and long (8K) context lengths, achieving training speedups of up to 4.9$\times$ on Wiki-40B and 12.1$\times$ on PG-19 for auto-regressive language modeling, and 4.8$\times$ on C4 for masked language modeling.
This paper has not been read by Pith yet.
Forward citations
Cited by 3 Pith papers
-
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
FlashAttention reduces GPU high-bandwidth memory accesses in self-attention via tiling, delivering exact attention with lower IO complexity, 2-3x wall-clock speedups on models like GPT-2, and the ability to train on s...
-
Sessa: Selective State Space Attention
Sessa integrates attention within recurrent paths to achieve power-law memory tails and flexible non-decaying selective retrieval, outperforming baselines on long-context tasks.
-
StateX: Enhancing RNN Recall via Post-training State Expansion
StateX post-trains RNNs to expand recurrent state size, improving recall and in-context learning with negligible parameter growth.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.