.. |
broadcast_load_epilogue_c2x.hpp
|
54f4f1e7f3
allow the cutlass kernels to take scales that reside on the GPU
|
7 сар өмнө |
broadcast_load_epilogue_c3x.hpp
|
54f4f1e7f3
allow the cutlass kernels to take scales that reside on the GPU
|
7 сар өмнө |
common.hpp
|
2313c97e3d
add cutlass w8a8 kernels (#556)
|
7 сар өмнө |
scaled_mm_dq_c2x.cu
|
54f4f1e7f3
allow the cutlass kernels to take scales that reside on the GPU
|
7 сар өмнө |
scaled_mm_dq_c3x.cu
|
f2c6791527
feat: update cutlass fp8 configs
|
7 сар өмнө |
scaled_mm_dq_entry.cu
|
67084aca5b
do not build cutlass kernels if cuda version is too low
|
7 сар өмнө |