CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

Another bucket sort (#5109)

Browse Source 
* Initial bucket sort

* Bucket sort: slightly better version

* Bucket sort: another minor improvement

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2024-01-26 09:14:39 +02:00
committed by GitHub
parent fe54033b69
commit 1182cf4d4f
1 changed files with 50 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
llama.cpp
View File

@@ -7956,10 +7956,57 @@ void llama_sample_top_k(struct llama_context * ctx, llama_token_data_array * can
auto comp = [](const llama_token_data & a, const llama_token_data & b) { auto comp = [](const llama_token_data & a, const llama_token_data & b) {
return a.logit > b.logit; return a.logit > b.logit;
}; };
if (k == (int) candidates->size) { if (k <= 128) {
std::sort(candidates->data, candidates->data + candidates->size, comp);
} else {
std::partial_sort(candidates->data, candidates->data + k, candidates->data + candidates->size, comp); std::partial_sort(candidates->data, candidates->data + k, candidates->data + candidates->size, comp);
} else {
constexpr int nbuckets = 128;
constexpr float bucket_low = -10.0f;
constexpr float bucket_high = 10.0f;
constexpr float bucket_scale = nbuckets/(bucket_high - bucket_low);
constexpr float bucker_inter = -bucket_low * bucket_scale;
std::vector<int> bucket_idx(candidates->size);
std::vector<int> histo(nbuckets, 0);
for (int i = 0; i < (int)candidates->size; ++i) {
const float val = candidates->data[i].logit;
int ib = int(bucket_scale * val + bucker_inter); //nbuckets * (val - bucket_low) / (bucket_high - bucket_low);
ib = std::max(0, std::min(nbuckets-1, ib));
bucket_idx[i] = ib;
++histo[ib];
}
int nhave = 0;
int ib = nbuckets - 1;
for ( ; ib >= 0; --ib) {
nhave += histo[ib];
if (nhave >= k) break;
}
std::vector<llama_token_data> tmp_tokens(nhave);
auto ptr = tmp_tokens.data();
std::vector<llama_token_data*> bucket_ptrs;
bucket_ptrs.reserve(nbuckets - ib);
for (int j = nbuckets - 1; j >= ib; --j) {
bucket_ptrs.push_back(ptr);
ptr += histo[j];
}
for (int i = 0; i < (int)candidates->size; ++i) {
int j = bucket_idx[i];
if (j >= ib) {
*bucket_ptrs[nbuckets-1-j]++ = candidates->data[i];
}
}
ptr = tmp_tokens.data();
int ndone = 0;
for (int j = nbuckets-1; j > ib; --j) {
std::sort(ptr, ptr + histo[j], comp);
ptr += histo[j];
ndone += histo[j];
}
std::partial_sort(ptr, ptr + k - ndone, ptr + histo[ib], comp);
std::memcpy(candidates->data, tmp_tokens.data(), k*sizeof(llama_token_data));
} }
candidates->sorted = true; candidates->sorted = true;
} }