1. 28 Aug, 2023 1 commit
    • dan_the_3rd's avatar
      [ft_attention] Fix for seqlen=8136 (#488) · c3f2a632
      dan_the_3rd authored
      When seqlen=8136, `smem_sz = 48840`, and apparently starting the kernel returns an `invalid argument` CUDA error.
      
      `48840 < 48 * 1024` but apparently it's still above the limit somehow..?
      Tested on A100
      c3f2a632
  2. 02 Jul, 2023 1 commit
  3. 21 Apr, 2023 1 commit
  4. 04 Jan, 2023 1 commit