CUDA: fix bug in topk-moe softmax (#16711)

This commit is contained in:
Aman Gupta
2025-10-22 12:33:08 +08:00
committed by GitHub
parent 03792ad936
commit 9285325ce0

View File

@@ -141,7 +141,7 @@ __launch_bounds__(4 * WARP_SIZE, 1) __global__ void topk_moe_cuda(const float *
wt_sum = warp_reduce_sum(wt_sum);
const float inv_sum = 1.0f / wt_sum;
for (int i = threadIdx.x; i < n_expert_used; i += WARP_SIZE) {
for (int i = 0; i < experts_per_thread; i++) {
output_weights[i] *= inv_sum;
}
}