vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 17:14:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge commit 'f7625019c5' into concedo_experimental

Browse source 
# Conflicts:
#	.github/ISSUE_TEMPLATE/bug.md
#	.github/workflows/build.yml
#	CMakeLists.txt
#	Makefile
#	README.md
#	tests/test-backend-ops.cpp
#	tests/test-opt.cpp
#	tests/test-quantize-fns.cpp
This commit is contained in:
Concedo 2024-02-26 10:44:23 +08:00
commit 3ccaf8e09a
37 changed files with 3514 additions and 698 deletions
Download patch file Download diff file Expand all files Collapse all files

14
examples/llama-bench/llama-bench.cpp
View file

@ -158,9 +158,9 @@ static const char * output_format_str(output_formats format) {
static const char * split_mode_str(llama_split_mode mode) {
switch (mode) {
case LLAMA_SPLIT_NONE: return "none";
case LLAMA_SPLIT_LAYER: return "layer";
case LLAMA_SPLIT_ROW: return "row";
case LLAMA_SPLIT_MODE_NONE: return "none";
case LLAMA_SPLIT_MODE_LAYER: return "layer";
case LLAMA_SPLIT_MODE_ROW: return "row";
default: GGML_ASSERT(!"invalid split mode");
}
}
@ -194,7 +194,7 @@ static const cmd_params cmd_params_defaults = {
/* type_v */ {GGML_TYPE_F16},
/* n_threads */ {get_num_physical_cores()},
/* n_gpu_layers */ {99},
/* split_mode */ {LLAMA_SPLIT_LAYER},
/* split_mode */ {LLAMA_SPLIT_MODE_LAYER},
/* main_gpu */ {0},
/* no_kv_offload */ {false},
/* mul_mat_q */ {true},
@ -359,11 +359,11 @@ static cmd_params parse_cmd_params(int argc, char ** argv) {
for (const auto & m : p) {
llama_split_mode mode;
if (m == "none") {
mode = LLAMA_SPLIT_NONE;
mode = LLAMA_SPLIT_MODE_NONE;
} else if (m == "layer") {
mode = LLAMA_SPLIT_LAYER;
mode = LLAMA_SPLIT_MODE_LAYER;
} else if (m == "row") {
mode = LLAMA_SPLIT_ROW;
mode = LLAMA_SPLIT_MODE_ROW;
} else {
invalid_param = true;
break;