wylab/llama.cpp
1
0
Fork 1
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-06-09 07:16:44 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix(mtmd): handle Gemma 4 audio projector embedding size (#24091)

Browse Source 
* mtmd: handle Gemma 4 audio projector embedding size

* rm projection_dim from clip_n_mmproj_embd

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
This commit is contained in:
Andrei
2026-06-04 02:51:23 -07:00
committed by GitHub
parent 6ddc9430b1
commit e3ba22d6cc
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/mtmd/clip.cpp
+2 -2
View File
@@ -4347,6 +4347,8 @@ int clip_n_mmproj_embd(const struct clip_ctx * ctx) {
return ctx->model.mm_input_proj_w->ne[0];
case PROJECTOR_TYPE_GEMMA4V:
case PROJECTOR_TYPE_GEMMA4UV:
case PROJECTOR_TYPE_GEMMA4A:
case PROJECTOR_TYPE_GEMMA4UA:
return ctx->model.mm_input_proj_w->ne[1];
case PROJECTOR_TYPE_IDEFICS3:
return ctx->model.mm_fc_w->ne[1];
@@ -4381,8 +4383,6 @@ int clip_n_mmproj_embd(const struct clip_ctx * ctx) {
return ctx->model.mm_fc_w->ne[1];
case PROJECTOR_TYPE_LFM2A:
return ctx->model.position_embeddings->ne[0];
case PROJECTOR_TYPE_GEMMA4UA:
return ctx->model.hparams.projection_dim;
case PROJECTOR_TYPE_GRANITE_SPEECH:
return ctx->model.qf_proj_linear_w->ne[1];
case PROJECTOR_TYPE_GLM4V: