mikhail/llama.cpp-mtp-turboquant
1
0
Fork 0
Code Issues Pull Requests Actions 27 Packages Projects Releases Wiki Activity

kv-cache : support V-less cache (#19067)

Browse Source 
* kv-cache : support V-less cache

* cuda : better check for V_is_K_view

* cuda : improve V_is_K_view check

* graph : add comments

* hparams : refactor
This commit is contained in:
Georgi Gerganov
2026-01-25 15:48:56 +02:00
committed by GitHub
parent 70d860824a
commit d9c6ce46f7
11 changed files with 246 additions and 53 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-model-saver.cpp
+2 -2
View File
@@ -146,8 +146,8 @@ void llama_model_saver::add_kv_from_model() {
add_kv(LLM_KV_VOCAB_SIZE, vocab.n_tokens());
add_kv(LLM_KV_CONTEXT_LENGTH, hparams.n_ctx_train);
add_kv(LLM_KV_EMBEDDING_LENGTH, hparams.n_embd);
if (hparams.n_embd_out > 0) {
add_kv(LLM_KV_EMBEDDING_LENGTH_OUT, hparams.n_embd_out);
if (hparams.n_embd_out_impl > 0) {
add_kv(LLM_KV_EMBEDDING_LENGTH_OUT, hparams.n_embd_out_impl);
}
add_kv(LLM_KV_BLOCK_COUNT, hparams.n_layer);
add_kv(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_layer_dense_lead);