happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fill zeros for experts with zero counts to preserve shape

This commit is contained in:
Ed Addario 2025-10-28 18:35:17 +00:00
parent 637e674da6
commit 683ef8dfb7
No known key found for this signature in database
GPG Key ID: E7875815A3230993
1 changed files with 12 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
tools/imatrix/imatrix.cpp
View File

@ -132,25 +132,34 @@ static std::vector<float> compute_tensor_averages(const Stats & tstats) {
const size_t n_mat = tstats.counts.size();
const size_t len = !tstats.activations.empty() ? tstats.activations.size() : tstats.values.size();
if (len == 0 || len % n_mat != 0) return {};
if (len == 0 || n_mat == 0 || len % n_mat != 0) { return {}; }
const size_t row = len / n_mat;
std::vector<float> vec;
vec.reserve(len);
if (tstats.activations.empty()) {
// Use mean of squares; fill zeros for experts with zero counts to preserve shape
for (size_t m = 0; m < n_mat; ++m) {
const float c = (float)tstats.counts[m];
if (c <= 0) return {};
const size_t off = m * row;
if (c <= 0.0f) {
vec.insert(vec.end(), row, 0.0f);
continue;
}
for (size_t j = 0; j < row; ++j) {
vec.push_back(tstats.values[off + j] / c);
}
}
} else {
// Use mean; fill zeros for experts with zero counts to preserve shape
for (size_t m = 0; m < n_mat; ++m) {
const float c = (float)tstats.counts[m];
if (c <= 0) return {};
const size_t off = m * row;
if (c <= 0.0f) {
vec.insert(vec.end(), row, 0.0f);
continue;
}
for (size_t j = 0; j < row; ++j) {
vec.push_back(tstats.activations[off + j] / c);
}