vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-14 19:09:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

llama : enable LLAMA_SET_ROWS=1 by default (#14959)

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2025-08-02 17:14:21 +03:00 • committed by GitHub
parent 15e92fd337
commit a4569c41fd
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
4 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/llama-context.cpp
View file

@ -105,7 +105,7 @@ llama_context::llama_context(
{
const char * LLAMA_SET_ROWS = getenv("LLAMA_SET_ROWS");
supports_set_rows = LLAMA_SET_ROWS ? (atoi(LLAMA_SET_ROWS) != 0) : false;
supports_set_rows = LLAMA_SET_ROWS ? (atoi(LLAMA_SET_ROWS) != 0) : supports_set_rows;
if (!supports_set_rows && !cparams.kv_unified) {
LLAMA_LOG_WARN("%s: non-unified KV cache requires ggml_set_rows() - forcing unified KV cache\n", __func__);

2
src/llama-context.h
View file

@ -289,7 +289,7 @@ private:
// env: LLAMA_SET_ROWS (temporary)
// ref: https://github.com/ggml-org/llama.cpp/pull/14285
bool supports_set_rows = false;
bool supports_set_rows = true;
// env: LLAMA_GRAPH_REUSE_DISABLE
bool graph_reuse_disable = false;

2
src/llama-kv-cache-unified.cpp
View file

@ -193,7 +193,7 @@ llama_kv_cache_unified::llama_kv_cache_unified(
debug = LLAMA_KV_CACHE_DEBUG ? atoi(LLAMA_KV_CACHE_DEBUG) : 0;
const char * LLAMA_SET_ROWS = getenv("LLAMA_SET_ROWS");
supports_set_rows = LLAMA_SET_ROWS ? atoi(LLAMA_SET_ROWS) != 0 : 0;
supports_set_rows = LLAMA_SET_ROWS ? atoi(LLAMA_SET_ROWS) != 0 : supports_set_rows;
if (!supports_set_rows) {
// ref: https://github.com/ggml-org/llama.cpp/pull/14363

2
src/llama-kv-cache-unified.h
View file

@ -230,7 +230,7 @@ private:
// env: LLAMA_SET_ROWS (temporary)
// ref: https://github.com/ggml-org/llama.cpp/pull/14285
bool supports_set_rows = false;
bool supports_set_rows = true;
const llama_swa_type swa_type = LLAMA_SWA_TYPE_NONE;