Skip to content
Trending storyDeveloping

PaddlePaddle releases FlashMask attention library for long-context LLM training

1 article1 sourcesince Oct 16Last article 1h ago ·

Overview

AISummary of 1 article

PaddlePaddle announces FlashMask, an attention library for LLM training that supports complex attention masks.

PaddlePaddle says it runs up to 1.55x faster than FlashAttention-4. The library adds context parallelism with load balancing and computation–communication overlap, and it is supported in PaddlePaddle and PyTorch through CuTe DSL on SM90 and SM100 GPUs. The team says FlashMask-3 has been accepted at NeurIPS 2026.

Written by AI from the articles below · updated Oct 16, 5:25 AM ET

Check the sources:

Latest developmentOct 16, 5:11 AM ET
FlashMask library speeds LLM long-context training up to 1.55x over FlashAttention-4

Article timeline

The articles in this story. Times are ET.

Oct 16
  1. PaddlePaddleOfficial
    FlashMask library speeds LLM long-context training up to 1.55x over FlashAttention-4

    AIPaddlePaddle announces FlashMask, an attention library for LLM training that supports complex attention masks and runs up to 1.55x faster than FlashAttention-4. It adds context parallelism with load balancing and computation–communication overlap, with PaddlePaddle and PyTorch support via CuTe DSL on SM90 and SM100 GPUs. The team says FlashMask-3 has been accepted at NeurIPS 2026.

    Image from @PaddlePaddle's post

Heat trend

Not enough continuous observations to show a trend yet.