Cuda kernels created with kernels library
attention-int8 : A high-performance INT8 fused attention kernel for diffusion transformers.
ParagEkbote/model-kernels
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Cuda kernels created with kernels library
attention-int8 : A high-performance INT8 fused attention kernel for diffusion transformers.