thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

model : add LFM2-ColBert-350M (#18607)

Browse source 
* model : add LFM2-ColBert-350M

* llama_model_n_embd_out() - returns `hparams.n_embd_out` if set and fallbacks to `hparams.n_embd`
This commit is contained in:
Tarek Dakhran 2026-01-05 19:52:56 +01:00 • committed by GitHub
parent df17a4c94f
commit 73d284a250
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
16 changed files with 118 additions and 60 deletions
Download patch file Download diff file Expand all files Collapse all files

8
src/llama-model.cpp
View file

@ -507,6 +507,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
ml.get_key(LLM_KV_CONTEXT_LENGTH, hparams.n_ctx_train);
ml.get_key(LLM_KV_EMBEDDING_LENGTH, hparams.n_embd);
ml.get_key(LLM_KV_EMBEDDING_LENGTH_OUT, hparams.n_embd_out, false);
ml.get_key(LLM_KV_BLOCK_COUNT, hparams.n_layer);
ml.get_key(LLM_KV_EXPERT_COUNT, hparams.n_expert, false);
ml.get_key(LLM_KV_EXPERT_USED_COUNT, hparams.n_expert_used, false);
@ -6469,6 +6470,9 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
layer.shortconv.out_proj = create_tensor(tn(LLM_TENSOR_SHORTCONV_OUTPROJ, "weight", i), {n_embd, n_embd}, 0);
}
}
// for LFM2-ColBert-350M
dense_2_out_layers = create_tensor(tn(LLM_TENSOR_DENSE_2_OUT, "weight"), {n_embd, hparams.get_n_embd_out()}, TENSOR_NOT_REQUIRED);
} break;
case LLM_ARCH_SMALLTHINKER:
{
@ -8003,6 +8007,10 @@ int32_t llama_model_n_embd_inp(const llama_model * model) {
return model->hparams.n_embd_inp();
}
int32_t llama_model_n_embd_out(const llama_model * model) {
return model->hparams.get_n_embd_out();
}
int32_t llama_model_n_layer(const llama_model * model) {
return model->hparams.n_layer;
}