vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : llama_kv_cache_clear zeroes data + fix save-load seq

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-04-25 19:37:27 +03:00
parent ac1c6d91de
commit c225609f10
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
2 changed files with 103 additions and 46 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llama.h
View file

@ -526,7 +526,7 @@ extern "C" {
// Returns the number of used KV cells (i.e. have at least one sequence assigned to them)
LLAMA_API int32_t llama_get_kv_cache_used_cells(const struct llama_context * ctx);
// Clear the KV cache
// Clear the KV cache - both cell info is erased and KV data is zeroed
LLAMA_API void llama_kv_cache_clear(
struct llama_context * ctx);