 sclarkson
|
1feb711f46
Fix compilation with clang on ARM64 (#1285)
|
1 周之前 |
 Kai Londenberg
|
b443207c1f
Paged Attention support for FA3 (#1268)
|
1 月之前 |
 jayhshah
|
a5a75274bc
FA3 kvcache + split kv + gqa parallelization (#1236)
|
2 月之前 |
 Ying Zhang
|
dff976a84a
fixes
|
3 月之前 |
|
Ying Zhang
|
7b4e68e04f
hopper local attention
|
3 月之前 |
|
Ying Zhang
|
db80387343
Add seqused_q in fwd / bwd and seqused_k in bwd.
|
3 月之前 |
|
jayhshah
|
c92ca63268
FA3 FP8 qkv descales + restore max offset for h128 causal + added sync for producer WG (#1173)
|
3 月之前 |
 Tri Dao
|
5d5bfbb619
Remove contiguous checks
|
4 月之前 |
|
Tri Dao
|
bafe253042
[FA3] Bwd
|
4 月之前 |
|
jayhshah
|
5018ac6ac5
Fp8 kernel with "in-kernel" transpose of V in producer (#1100)
|
4 月之前 |
|
Tri Dao
|
3aae9c18c1
Revert "Changes For FP8 (#1075)"
|
4 月之前 |
 ganeshcolfax
|
1899c970c8
Changes For FP8 (#1075)
|
4 月之前 |
 Ying Zhang
|
dfe1a59e4b
Add var-seq-len to FA3 fp16 / bf16 fwd (#1072)
|
4 月之前 |
|
Tri Dao
|
74b0761ff7
[FA3] BF16 forward
|
5 月之前 |
|
Tri Dao
|
7f67966cc7
FA3 initial code release
|
5 月之前 |
