ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-04 21:40:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix kv cache save and load for GLM model (#965)

Browse Source 
Co-authored-by: firecoperana <firecoperana>
This commit is contained in:
firecoperana
2025-11-15 15:04:16 +00:00
committed by GitHub
parent 5ec0def0ef
commit b40d11b22d
2 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/server/server.cpp
View File

@@ -1837,7 +1837,7 @@ struct server_context {
LLAMA_LOG_INFO("prompt cache save took %.2f ms\n", (ggml_time_us() - t_start) / 1000.0);
}
// has prompts saved earlier to load
if (!prompt_cache->states.empty()) {
if (prompt_cache && !prompt_cache->states.empty()) {
const int64_t t_start = ggml_time_us();
ret->server_cached_prompt.tokens = server_tokens(tokens.get_text_tokens(), false); // copy cache tokens
ret->prompt_load(*prompt_cache, task.tokens);