vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ggml : minor naming changes (#8433)

Browse source 
* ggml : minor naming changes

ggml-ci

* ggml : use PRId64 [no ci]

* ggml : revert FA K/Q names
This commit is contained in:
Georgi Gerganov 2024-07-12 10:46:02 +03:00 • committed by GitHub
parent b549a1bbef
commit 370b1f7e7a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
10 changed files with 192 additions and 189 deletions
Download patch file Download diff file Expand all files Collapse all files

4
tests/test-double-float.cpp
View file

@ -14,7 +14,7 @@
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wdouble-promotion"
// ggml.c::quantize_row_q4_0_reference
// ggml.c::quantize_row_q4_0_ref
inline static uint8_t round_orig(float v0) { return ((int8_t) (round(v0))) + 8; }
// ggml.c::ggml_silu_f32
@ -24,7 +24,7 @@ inline static float silu_orig(float x) {
#pragma GCC diagnostic pop
// ggml.c::quantize_row_q4_0_reference
// ggml.c::quantize_row_q4_0_ref
inline static uint8_t round_float(float v0) { return (int8_t)roundf(v0) + 8; }
// ggml.c::ggml_silu_f32

2
tests/test-quantize-fns.cpp
View file

@ -60,7 +60,7 @@ static float reference_quantization_error(ggml_type_traits_t & qfns, size_t test
qfns.from_float(test_data, tmp_q.data(), test_size);
qfns.to_float(tmp_q.data(), tmp_out.data(), test_size);
qfns.from_float_reference(test_data, tmp_q.data(), test_size);
qfns.from_float_ref(test_data, tmp_q.data(), test_size);
qfns.to_float(tmp_q.data(), tmp_out_ref.data(), test_size);
return array_rmse(tmp_out.data(), tmp_out_ref.data(), test_size);

2
tests/test-quantize-perf.cpp
View file

@ -285,7 +285,7 @@ int main(int argc, char * argv[]) {
for (size_t size : params.test_sizes) {
printf(" %zu values (%.2f MB)\n", size, 4*size/(float)(1024*1024));
auto quantize_fn = [&](void) -> float {
qfns.from_float_reference(test_data1, test_q1, size);
qfns.from_float_ref(test_data1, test_q1, size);
return test_q1[0];
};
size_t quantized_size = ggml_row_size(type, size);