tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-27 12:05:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : normalize embeddings (#5956)

Browse Source 
* output normalize embedding in '/v1/embeddings'

* common : reuse llama_embd_normalize

* common : better normalize impl

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
SeungWon Jeong
2024-03-09 21:27:58 +09:00
committed by GitHub
parent 2c4f566c88
commit fb215c3832
4 changed files with 30 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
common/common.cpp
View File

@ -1852,3 +1852,18 @@ void dump_kv_cache_view_seqs(const llama_kv_cache_view & view, int row_size) {
printf("\n=== Done dumping\n");
}
void llama_embd_normalize(const float * inp, float * out, int n) {
double sum = 0.0;
for (int i = 0; i < n; i++) {
sum += inp[i] * inp[i];
}
sum = sqrt(sum);
const float norm = sum > 0.0 ? 1.0f / sum : 0.0f;
for (int i = 0; i < n; i++) {
out[i] = inp[i] * norm;
}
}