mirror of
https://github.com/vllm-project/vllm.git
synced 2025-10-20 14:53:52 +08:00
8 lines
224 B
C
8 lines
224 B
C
#pragma once
|
|
|
|
#include <torch/extension.h>
|
|
|
|
void topk_softmax(torch::Tensor& topk_weights, torch::Tensor& topk_indices,
|
|
torch::Tensor& token_expert_indices,
|
|
torch::Tensor& gating_output);
|