mirror of
https://github.com/zenlm/engine.git
synced 2026-10-30 15:00:05 +00:00
Guoqing BaoandGitHub
d38a7e1984
* Split Marlin and Paged Attention kernels for faster build * Typo fix
11 lines
477 B
Plaintext
11 lines
477 B
Plaintext
#include "marlin_kernel.cuh"
|
|
|
|
extern "C" void marlin_gptq_4bit_f16(const void *A, const void *B, void *scales,
|
|
void *zeros, void *C, int prob_m,
|
|
int prob_k, int prob_n, void *workspace,
|
|
int groupsize, int64_t stream) {
|
|
marlin_matmul<half, ScalarTypeID::kU4B8, false, 4>(
|
|
A, B, scales, zeros, C, prob_m, prob_k, prob_n, workspace, groupsize,
|
|
stream);
|
|
}
|