CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

server : fix cache reuse logic (#12161)

Browse Source 
The first kv shift offsets the positions of all tokens after head_c.
When using llama_kv_cache_seq_rm next, using head_c will remove the valid tokens because their positions have already been offset.
This commit is contained in:
Clauszy
2025-03-05 15:25:45 +08:00
committed by GitHub
parent a057897ad4
commit 06a92a193a
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/server/server.cpp
View File

@@ -3003,7 +3003,7 @@ struct server_context {
const int64_t kv_shift = (int64_t) head_p - (int64_t) head_c;
llama_kv_cache_seq_rm (ctx, slot.id, head_p, head_c);
llama_kv_cache_seq_add(ctx, slot.id, head_c, -1, kv_shift);
llama_kv_cache_seq_add(ctx, slot.id, head_c, head_c + n_match, kv_shift);
for (size_t i = 0; i < n_match; i++) {
slot.cache_tokens[head_p + i] = slot.cache_tokens[head_c + i];