ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-02 12:39:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

Don't split the output tensor

Browse Source
This commit is contained in:
Kawrakow
2025-12-05 13:16:11 +00:00
parent 9264abfbaf
commit ed8a3d8e3d
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/llama-load-tensors.cpp
View File

@@ -388,7 +388,7 @@ void create_tensors_helper::create_std_ffn(int i, const LLM_TN & tn, llama_layer
bool create_tensors_helper::create_llama_tensors(const LLM_TN & tn) {
LOADING_PRELUDE
create_embd_output(tn, n_embd, n_vocab, true, true);
create_embd_output(tn, n_embd, n_vocab, true, false); //true);
for (int i = 0; i < n_layer; ++i) {
ggml_context * ctx_layer = ctx_for_layer(i);
@@ -1843,7 +1843,7 @@ bool create_tensors_helper::create_glm4_moe_tensors(const LLM_TN & tn) {
GGML_ASSERT(hparams.n_expert > 0 && "n_expert must be > 0 for GLM4_MOE MoE layers");
GGML_ASSERT(hparams.n_expert_used > 0 && "n_expert_used must be > 0 for GLM4_MOE MoE layers");
create_embd_output(tn, n_embd, n_vocab, true, true);
create_embd_output(tn, n_embd, n_vocab, true, false); //true);
for (int i = 0; i < n_layer; ++i) {
ggml_context * ctx_layer = ctx_for_layer(i);