nicolas
/
ik_llama_opt
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

GLM DSA: reduce indexer cache size (#2093)

This commit is contained in:
Kawrakow 2026-07-07 09:35:42 +02:00 committed by GitHub
parent ba62eaffe7
commit 05cba319e0
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/llama.cpp
View File

@ -1030,7 +1030,7 @@ static bool llama_kv_cache_init(
cache.k_l.push_back(kv);
// DSA lightning-indexer key cache (MQA, single head). Store the Hadamard-rotated
// indexer keys in F16 so a decoded token can score against ALL past keys.
if (has_dsa_indexer && model.layers[i].indexer_attn_k && !is_mtp_tail_layer) {
if (has_dsa_indexer && model.layers[i].indexer_attn_k && hparams.indexer_is_full[i] && !is_mtp_tail_layer) {
ggml_tensor * kr = ggml_new_tensor_2d(ctx, idx_type_k, hparams.indexer_head_size, kv_size);
ggml_format_name(kr, "cache_kr_l%d", i);
cache.kr_l[i] = kr;