vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-05-12 05:52:26 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

Merge commit 'b677721819' into concedo_experimental

Browse source 
# Conflicts:
#	CONTRIBUTING.md
#	common/chat.cpp
#	docs/ops.md
#	docs/ops/CPU.csv
#	docs/ops/CUDA.csv
#	docs/ops/OpenCL.csv
#	ggml/src/ggml-cann/aclnn_ops.cpp
#	ggml/src/ggml-cann/common.h
#	ggml/src/ggml-cann/ggml-cann.cpp
#	ggml/src/ggml-sycl/softmax.cpp
#	grammars/README.md
#	src/CMakeLists.txt
#	tests/test-backend-ops.cpp
#	tests/test-chat.cpp
#	tests/test-grammar-integration.cpp
#	tests/test-grammar-parser.cpp
#	tests/test-llama-grammar.cpp
#	tools/mtmd/CMakeLists.txt
This commit is contained in:
Concedo 2025-12-11 23:33:19 +08:00
commit 34d243bf3c
24 changed files with 1434 additions and 136 deletions
Download patch file Download diff file Expand all files Collapse all files

33
src/llama-model.cpp
View file

@ -58,7 +58,7 @@
#include "models/gemma-embedding.cpp"
#include "models/gemma.cpp"
#include "models/gemma2-iswa.cpp"
#include "models/gemma3-iswa.cpp"
#include "models/gemma3.cpp"
#include "models/gemma3n-iswa.cpp"
#include "models/glm4-moe.cpp"
#include "models/glm4.cpp"
@ -225,6 +225,7 @@ const char * llm_type_name(llm_type type) {
case LLM_TYPE_16B_A1B: return "16B.A1B";
case LLM_TYPE_21B_A3B: return "21B.A3B";
case LLM_TYPE_30B_A3B: return "30B.A3B";
case LLM_TYPE_80B_A3B: return "80B.A3B";
case LLM_TYPE_100B_A6B: return "100B.A6B";
case LLM_TYPE_106B_A12B: return "106B.A12B";
case LLM_TYPE_230B_A10B: return "230B.A10B";
@ -1369,18 +1370,25 @@ void llama_model::load_hparams(llama_model_loader & ml) {
} break;
case LLM_ARCH_GEMMA3:
{
hparams.swa_type = LLAMA_SWA_TYPE_STANDARD;
hparams.set_swa_pattern(6);
const bool found_swa = ml.get_key(LLM_KV_ATTENTION_SLIDING_WINDOW, hparams.n_swa, false);
if (found_swa && hparams.n_swa > 0) {
hparams.swa_type = LLAMA_SWA_TYPE_STANDARD;
hparams.set_swa_pattern(6);
hparams.rope_freq_base_train_swa = 10000.0f;
hparams.rope_freq_scale_train_swa = 1.0f;
hparams.rope_freq_base_train_swa = 10000.0f;
hparams.rope_freq_scale_train_swa = 1.0f;
} else {
hparams.swa_type = LLAMA_SWA_TYPE_NONE;
}
ml.get_key(LLM_KV_ATTENTION_SLIDING_WINDOW, hparams.n_swa);
hparams.f_final_logit_softcapping = 0.0f;
ml.get_key(LLM_KV_FINAL_LOGIT_SOFTCAPPING, hparams.f_final_logit_softcapping, false);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
switch (hparams.n_layer) {
case 18: type = LLM_TYPE_270M; break;
case 26: type = LLM_TYPE_1B; break;
case 32: type = LLM_TYPE_8B; break; // Rnj-1
case 34: type = LLM_TYPE_4B; break;
case 48: type = LLM_TYPE_12B; break;
case 62: type = LLM_TYPE_27B; break;
@ -1704,8 +1712,9 @@ void llama_model::load_hparams(llama_model_loader & ml) {
ml.get_key(LLM_KV_EXPERT_SHARED_COUNT, hparams.n_expert_shared);
ml.get_key(LLM_KV_EXPERT_WEIGHTS_SCALE, hparams.expert_weights_scale);
switch (hparams.n_layer) {
case 28: type = LLM_TYPE_20B; break;
switch (hparams.n_ff_exp) {
case 1408: type = LLM_TYPE_16B; break;
case 1792: type = LLM_TYPE_20B; break;
default: type = LLM_TYPE_UNKNOWN;
}
} break;
@ -2354,7 +2363,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
}
switch (hparams.n_layer) {
case 80: type = LLM_TYPE_80B_A3B; break;
case 48: type = LLM_TYPE_80B_A3B; break;
default: type = LLM_TYPE_UNKNOWN;
}
} break;
@ -7466,7 +7475,11 @@ ggml_cgraph * llama_model::build_graph(const llm_graph_params & params) const {
} break;
case LLM_ARCH_GEMMA3:
{
llm = std::make_unique<llm_build_gemma3_iswa>(*this, params);
if (hparams.swa_type == LLAMA_SWA_TYPE_STANDARD) {
llm = std::make_unique<llm_build_gemma3<true>>(*this, params);
} else {
llm = std::make_unique<llm_build_gemma3<false>>(*this, params);
}
} break;
case LLM_ARCH_GEMMA3N:
{