diff options
| author | SeungWon Jeong <65549245+redlion0929@users.noreply.github.com> | 2024-03-09 21:27:58 +0900 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-03-09 14:27:58 +0200 |
| commit | fb215c3832236fec7380c4fb618bd7154cb196ef (patch) | |
| tree | 1c2d0eb8fce7d2c1f70024b1f2c7b4b35baaa029 /common/common.h | |
| parent | 2c4f566c88322ebf2f9bd11b01b5ebdaa0130b89 (diff) | |
server : normalize embeddings (#5956)
* output normalize embedding in '/v1/embeddings'
* common : reuse llama_embd_normalize
* common : better normalize impl
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Diffstat (limited to 'common/common.h')
| -rw-r--r-- | common/common.h | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/common/common.h b/common/common.h index 977ce419..f8d82b87 100644 --- a/common/common.h +++ b/common/common.h @@ -260,3 +260,10 @@ void dump_kv_cache_view(const llama_kv_cache_view & view, int row_size = 80); // Dump the KV cache view showing individual sequences in each cell (long output). void dump_kv_cache_view_seqs(const llama_kv_cache_view & view, int row_size = 40); + +// +// Embedding utils +// + +void llama_embd_normalize(const float * inp, float * out, int n); + |