tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-29 12:35:16 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama : update llama_kv_self API

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov
2025-01-14 16:47:34 +02:00
parent fd05ab87aa
commit 17b363afd3
30 changed files with 387 additions and 205 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
examples/embedding/embedding.cpp
View File

@ -34,11 +34,10 @@ static void batch_add_seq(llama_batch & batch, const std::vector<int32_t> & toke
static void batch_decode(llama_context * ctx, llama_batch & batch, float * output, int n_seq, int n_embd, int embd_norm) {
const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
const llama_model * model = llama_get_model(ctx);
llama_kv_cache * kv = llama_get_kv_cache(ctx);
const struct llama_model * model = llama_get_model(ctx);
// clear previous kv_cache values (irrelevant for embeddings)
llama_kv_cache_clear(kv);
llama_kv_self_clear(ctx);
// run model
LOG_INF("%s: n_tokens = %d, n_seq = %d\n", __func__, batch.n_tokens, n_seq);