13 lines
532 B
C
13 lines
532 B
C
#pragma once
|
|
|
|
#include <torch/all.h>
|
|
|
|
torch::Tensor marlin_gemm_moe(
|
|
const torch::Tensor& a, const torch::Tensor& b_q_weights,
|
|
const torch::Tensor& sorted_ids, const torch::Tensor& topk_weights,
|
|
const torch::Tensor& topk_ids, const torch::Tensor& b_scales,
|
|
const torch::Tensor& g_idx, const torch::Tensor& perm,
|
|
torch::Tensor& workspace, int64_t size_m, int64_t size_n, int64_t size_k,
|
|
bool is_k_full, int64_t num_experts, int64_t topk, int64_t moe_block_size,
|
|
bool replicate_input, bool apply_weights);
|