Morph MORPH
SIGN IN SIGN UP
tejas / flash-attention

Fast and memory-efficient exact attention

0 0 209 Python

ACTIVITY

Commits Reviews
week month year
COMMITS
20
in the last year
CONTRIBUTORS
4
active
STARS
0
total
FORKS
0
total
TOP CONTRIBUTORS
T
Tri Dao
16 commits
R
Reuben Stern
2 commits
K
Kevin Wang
1 commit
J
Johnny
1 commit
RECENT COMMITS
T
[Cute,Bwd,Sm100] Implement cluster
Tri Dao
c59ecd8
T
[Cute,Bwd,Sm100] Remove delay_tma_store option
Tri Dao
285bf12
T
[Cute] Change utils.view_transpose back
Tri Dao
6c56a0c
T
[Cute,Bwd,Sm100] Reduce sync
Tri Dao
2c7177d
T
[Cute,Sm100] acc_tmem_addr is Int32 instead of constexpr
Tri Dao
e873ad0
T
[Cute,Bwd,Sm100] Tune registers
Tri Dao
b634499
T
[Cute] Add store_shared_remote_fp32x4 util function
Tri Dao
510fe92
T
[Cute,Bwd] Enable bwd benchmarks
Tri Dao
79b9030
T
[Cute,Bwd,Sm100] Enable bwd tests
Tri Dao
662cf9c
T
[Cute,Bwd,Sm100] Causal mask
Tri Dao
93a0afe
T
[Cute,Bwd,Sm100] Simplify layouts in compute_loop
Tri Dao
6eb7c80
T
[Cute,Bwd,Sm100] Make postprocessing work, add interface
Tri Dao
b3f437f
T
[Cute,Sm100] In gemm ptx, add to base smem_address instead
Tri Dao
7ef1a6f
T
[Cute,Fwd,Sm100] Fix interface w score mod to get it to run
Tri Dao
9450df6
K
Fix FA3 segfault with custom CUDA streams in ABI stable build (#1957)
Kevin Wang
3effce8
R
[CuTe DSL] Update "buffers" name to "aux_tensors"; fix flex bugs (#1961)
Reuben Stern
e4d25a4
T
[Cute,Bwd,Sm100] More cleanup
Tri Dao
1b8e1e6
T
[Cute,Bwd,Sm100] Use CopyBulkG2SOp copy op instead of calling ptx
Tri Dao
9dbed03
J
cutlass v4.3.0 (#1952)
Johnny
16c7f0f
R
Block Sparsity and Flex Attention mask mod support (#1942)
Reuben Stern
143b0ba