Skip to content

Commit

Permalink
fix: remove redundant load (#495)
Browse files Browse the repository at this point in the history
  • Loading branch information
Bruce-Lee-LY authored Sep 11, 2024
1 parent f2ca781 commit 2de16b0
Showing 1 changed file with 0 additions and 1 deletion.
1 change: 0 additions & 1 deletion include/flashinfer/norm.cuh
Original file line number Diff line number Diff line change
Expand Up @@ -178,7 +178,6 @@ __global__ void FusedAddRMSNormKernel(T* __restrict__ input, T* __restrict__ res
weight_vec.fill(0.f);
residual_vec.fill(0.f);
if ((i * num_threads + thread_id) * VEC_SIZE < d) {
input_vec.load(input + bx * d + i * num_threads * VEC_SIZE + thread_id * VEC_SIZE);
weight_vec.load(weight + i * num_threads * VEC_SIZE + thread_id * VEC_SIZE);
residual_vec.load(residual + bx * d + i * num_threads * VEC_SIZE + thread_id * VEC_SIZE);
}
Expand Down

0 comments on commit 2de16b0

Please sign in to comment.