Skip to content

benchmark-matmult : do not use integer abs() on a float #3277

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Sep 20, 2023
Merged

benchmark-matmult : do not use integer abs() on a float #3277

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b530a55
benchmark-matmult : do not use integer abs() on a float
cebtenzzre Sep 20, 2023
c9d5b54
benchmark-matmult : fix missing 'static' specifiers
cebtenzzre Sep 20, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions examples/benchmark/benchmark-matmult.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@
#pragma warning(disable: 4244 4267) // possible loss of data
#endif

void ggml_graph_compute_helper(std::vector<uint8_t> & buf, ggml_cgraph * graph, int n_threads) {
static void ggml_graph_compute_helper(std::vector<uint8_t> & buf, ggml_cgraph * graph, int n_threads) {
struct ggml_cplan plan = ggml_graph_plan(graph, n_threads);

if (plan.work_size > 0) {
Expand All @@ -32,7 +32,7 @@ void ggml_graph_compute_helper(std::vector<uint8_t> & buf, ggml_cgraph * graph,
ggml_graph_compute(graph, &plan);
}

float tensor_sum_elements(const ggml_tensor * tensor) {
static float tensor_sum_elements(const ggml_tensor * tensor) {
double sum = 0;
if (tensor->type == GGML_TYPE_F32) {
for (int j = 0; j < tensor->ne[1]; j++) {
Expand All @@ -44,7 +44,7 @@ float tensor_sum_elements(const ggml_tensor * tensor) {
return sum;
}

void tensor_dump(const ggml_tensor * tensor, const char * name) {
static void tensor_dump(const ggml_tensor * tensor, const char * name) {
printf("%15s: type = %i (%5s) ne = %5" PRIi64 " x %5" PRIi64 " x %5" PRIi64 ", nb = (%5zi, %5zi, %5zi) - ", name,
tensor->type, ggml_type_name(tensor->type),
tensor->ne[0], tensor->ne[1], tensor->ne[2], tensor->nb[0], tensor->nb[1], tensor->nb[2]);
Expand All @@ -59,7 +59,7 @@ struct benchmark_params_struct {
int32_t n_iterations = 10;
};

void print_usage(int /*argc*/, char ** argv, struct benchmark_params_struct params) {
static void print_usage(int /*argc*/, char ** argv, struct benchmark_params_struct params) {
fprintf(stderr, "usage: %s [options]\n", argv[0]);
fprintf(stderr, "\n");
fprintf(stderr, "options:\n");
Expand Down Expand Up @@ -253,7 +253,7 @@ int main(int argc, char ** argv) {
// Check that the matrix multiplication result is in the right ballpark
// We cannot use the exact value from the F32 multiplication because the quantizuation will be slightly different
float sum_of_Q4_result = tensor_sum_elements(gf31.nodes[0]);
float delta = abs(sum_of_Q4_result - sum_of_F32_reference);
float delta = std::abs(sum_of_Q4_result - sum_of_F32_reference);
float allowed_delta = (sum_of_F32_reference) / 1000 / 1000; // Let's accept an epsilon of 10^-6

if (delta > allowed_delta) {
Expand Down