Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-19 09:15:18 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit '128d522c04286e019666bd6ee4d18e3fbf8772e2' into concedo_experimental

Browse Source 
# Conflicts:
#	.github/workflows/build.yml
#	.github/workflows/release.yml
#	ggml/src/ggml-vulkan/ggml-vulkan.cpp
#	tests/test-alloc.cpp
#	tests/test-chat.cpp
This commit is contained in:
Concedo
2025-10-04 23:51:22 +08:00
parent b8680cd0c5 128d522c04
commit 1d728bbc89
18 changed files with 381 additions and 226 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-model.cpp
+11
View File
@@ -7905,6 +7905,8 @@ struct llm_build_bert : public llm_graph_context {
}
if (model.layers[il].attn_q_norm) {
Qcur = ggml_reshape_2d(ctx0, Qcur, n_embd_head*n_head, n_tokens);
Qcur = build_norm(Qcur,
model.layers[il].attn_q_norm,
model.layers[il].attn_q_norm_b,
@@ -7914,6 +7916,8 @@ struct llm_build_bert : public llm_graph_context {
}
if (model.layers[il].attn_k_norm) {
Kcur = ggml_reshape_2d(ctx0, Kcur, n_embd_head*n_head_kv, n_tokens);
Kcur = build_norm(Kcur,
model.layers[il].attn_k_norm,
model.layers[il].attn_k_norm_b,
@@ -8296,6 +8300,9 @@ struct llm_build_mpt : public llm_graph_context {
// Q/K Layernorm
if (model.layers[il].attn_q_norm) {
Qcur = ggml_reshape_2d(ctx0, Qcur, n_embd_head*n_head, n_tokens);
Kcur = ggml_reshape_2d(ctx0, Kcur, n_embd_head*n_head_kv, n_tokens);
Qcur = build_norm(Qcur,
model.layers[il].attn_q_norm,
model.layers[il].attn_q_norm_b,
@@ -20206,6 +20213,10 @@ bool llama_model_is_recurrent(const llama_model * model) {
return llm_arch_is_recurrent(model->arch);
}
bool llama_model_is_hybrid(const llama_model * model) {
return llm_arch_is_hybrid(model->arch);
}
bool llama_model_is_diffusion(const llama_model * model) {
return llm_arch_is_diffusion(model->arch);
}