 Tri Dao
|
7a802796e1
Big refactor and update
|
1 долоо хоног өмнө |
 jayhshah
|
a5a75274bc
FA3 kvcache + split kv + gqa parallelization (#1236)
|
3 сар өмнө |
 Ying Zhang
|
7b4e68e04f
hopper local attention
|
4 сар өмнө |
|
jayhshah
|
c92ca63268
FA3 FP8 qkv descales + restore max offset for h128 causal + added sync for producer WG (#1173)
|
4 сар өмнө |
|
Ying Zhang
|
a3a257c71d
Fix out-of-bound writes for var-seq-len zero-length KVs
|
5 сар өмнө |
|
jayhshah
|
5018ac6ac5
Fp8 kernel with "in-kernel" transpose of V in producer (#1100)
|
5 сар өмнө |
 Ying Zhang
|
dfe1a59e4b
Add var-seq-len to FA3 fp16 / bf16 fwd (#1072)
|
5 сар өмнө |
|
Tri Dao
|
74b0761ff7
[FA3] BF16 forward
|
6 сар өмнө |
|
Tri Dao
|
7f67966cc7
FA3 initial code release
|
6 сар өмнө |
