 AlpinDale
|
40f63268ee
disable new layernorm kernels for CUDA < 12.0
|
vor 9 Monaten |
 AlpinDale
|
d68fad5a79
feat: add optimized layernorm kernels (#398)
|
vor 9 Monaten |
|
AlpinDale
|
8fa608aeb7
feat: replace Ray with NCCL for control plane comms (#221)
|
vor 1 Jahr |
|
AlpinDale
|
b9b295d74e
chore: backlogs 1 (#191)
|
vor 1 Jahr |
|
AlpinDale
|
7612f33afd
feat: fused add RMSNorm kernels (#125)
|
vor 1 Jahr |
|
AlpinDale
|
3d72f05c7b
feat: flattened 1D tensor -> 2D tensor (#85)
|
vor 1 Jahr |
|
AlpinDale
|
32844c1522
add GELU kernels and remove compile bloat
|
vor 1 Jahr |
|
AlpinDale
|
081545bde6
fix: various CUDA kernel tweaks
|
vor 1 Jahr |
|
AlpinDale
|
b8f4337c5b
chore: various fixes
|
vor 1 Jahr |
|
AlpinDale
|
0ec53128b6
feat: add layernorm kernels
|
vor 1 Jahr |
