aparler/c-llm
1
0
Fork 0
mirror of https://github.com/karpathy/llm.c.git synced 2026-07-28 20:35:09 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix error in small typos in matmul_forward.cu

Browse source
This commit is contained in:
lancer 2024-04-10 13:35:20 -07:00
parent ef7a4867b1
commit 68d2ce0601
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
dev/cuda/matmul_forward.cu
View file

@ -239,12 +239,12 @@ int main(int argc, char **argv) {
printf("%f %f\n", out[i], out_gpu[i]);
}
// ensure correctness for all elements
if (i >= 5 && fabs(out[i] - out_gpu[i]) > 1e-4) {
if (fabs(out[i] - out_gpu[i]) > 1e-4) {
printf("Mismatch at %d: %f vs %f\n", i, out[i], out_gpu[i]);
exit(1);
}
}
printf("Results match at block_size=256!\n");
printf("Results match at block_size=1024!\n");
// time the kernel at different block sizes
int sqrt_block_sizes[] = {4, 8, 16, 32};