vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-10 17:34:49 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

remove token_embd from metal mem

Browse source
This commit is contained in:
Lizonghang 2025-01-23 16:58:25 +04:00
parent 92f00303d5
commit fb05f80f89
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/llama.cpp
View file

@ -5260,7 +5260,7 @@ struct llama_model_loader {
try { try {
const char * tname = ggml_get_name(tensor); const char * tname = ggml_get_name(tensor);
if (keep_only_inp_out && !( if (keep_only_inp_out && !(
strcmp(tname, "token_embd.weight") == 0 || // strcmp(tname, "token_embd.weight") == 0 || // lookup table is used so we do not need to keep it in metal memory
strcmp(tname, "output_norm.weight") == 0 || strcmp(tname, "output_norm.weight") == 0 ||
strcmp(tname, "output.weight") == 0)) { strcmp(tname, "output.weight") == 0)) {
continue; continue;