DISTFLASHATTN: Distributed Memory-efficient Attention for Long-context LLMs Training

摘要

FlashAttention effectively reduces the quadratic peak memory usage to linear in training transformer-based large language models (LLMs) on a single GPU. In this paper, we introduce DistFlashAttention, a distributed memory-efficient attention mechanism optimized for long-context LLMs training. We propose three key techniques: token-level workload balancing, overlapping key-value communication, and a rematerialization-aware gradient checkpointing algorithm. We evaluate DistFlashAttention on Llama-7B and variants with sequence lengths from 32K to 512K. DistFlashAttention achieves 8x longer sequences, 4.45 - 5.64x speedup compared to Ring Self-Attention, 2-8x longer sequences, 1.24- 2.01x speedup compared to Megatron-LM with FlashAttention. It achieves 1.67x and 1.26-1.88x speedup compared to recent Ring Attention and DeepSpeed-Ulysses. Codes are available at https://github.com/RulinShao/LightSeq.