Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

mtp: support for gemma-4 E2B and E4B assistants (#24282)

Browse Source 
* models: update converter to support smaller assistants

* models: add masked_embd tensors to gemma4-assist arch

* gemma-4: remove temp debug for conversion

* gemma-4-mtp: filter out masked_embedding tensors during conversion
This commit is contained in:
Max Krasnyansky
2026-06-08 13:48:52 -07:00
committed by GitHub
parent 42a0afd594
commit 7d2b45b4f7
6 changed files with 34 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/models/gemma4-assistant.cpp
+3
View File
@@ -39,6 +39,9 @@ void llama_model_gemma4_assistant::load_arch_tensors(llama_model_loader &) {
output_norm = create_tensor(tn(LLM_TENSOR_OUTPUT_NORM, "weight"), { n_embd }, 0);
create_tensor(tn(LLM_TENSOR_MASKED_EMBD_CENTROIDS, "weight"), {}, TENSOR_NOT_REQUIRED);
create_tensor(tn(LLM_TENSOR_MASKED_EMBD_ORDERING), {}, TENSOR_NOT_REQUIRED);
const int64_t n_embd_backbone = hparams.n_embd_inp();
nextn_proj_post = create_tensor(tn(LLM_TENSOR_NEXTN_PROJ_POST, "weight"), { n_embd, n_embd_backbone }, 0);