[Kernel] Expand FP8 support to Ampere GPUs using FP8 Marlin (#5975)
This commit is contained in:
@@ -93,6 +93,11 @@ torch::Tensor gptq_marlin_repack(torch::Tensor& b_q_weight, torch::Tensor& perm,
|
||||
int64_t size_k, int64_t size_n,
|
||||
int64_t num_bits);
|
||||
|
||||
torch::Tensor fp8_marlin_gemm(torch::Tensor& a, torch::Tensor& b_q_weight,
|
||||
torch::Tensor& b_scales, torch::Tensor& workspace,
|
||||
int64_t num_bits, int64_t size_m, int64_t size_n,
|
||||
int64_t size_k);
|
||||
|
||||
bool cutlass_scaled_mm_supports_fp8(int64_t cuda_device_capability);
|
||||
|
||||
void cutlass_scaled_mm(torch::Tensor& out, torch::Tensor const& a,
|
||||
|
||||
Reference in New Issue
Block a user