vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-05-09 19:46:11 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

model: try to improve Qwen3 Next (#18683)

Browse source 
* qwen3next: simplify qkvz projection

* use ggml_swiglu_split

* revert swiglu_split, but remove redundant repeat()

* fix missing reshape

* rm 2 redundant transposes

* move mul_mat(k,q) to outside of chunking

* rm redundant cont

* improve g_cs_chunk

* add comments about no cont

* use std::pair instead of ggml_concat

* vectorize key_gdiff calculation

* rm unused tensor

* avoid ggml_concat inside loop

* bring back ggml_concat as it may not work on other backend

* nits
This commit is contained in:
Xuan-Son Nguyen 2026-01-11 12:53:33 +01:00 • committed by GitHub
parent 79456a690a
commit 506bb6e010
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 245 additions and 187 deletions
Download patch file Download diff file Expand all files Collapse all files

5
src/llama-model.cpp
View file

@ -6763,7 +6763,10 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
} else {
// Linear attention (gated delta net) specific tensors
// Create tensors with calculated dimensions
layer.ssm_in = create_tensor(tn(LLM_TENSOR_SSM_IN, "weight", i), { n_embd, qkvz_dim }, 0);
// note: ssm_in is used by legacy GGUF
layer.ssm_in = create_tensor(tn(LLM_TENSOR_SSM_IN, "weight", i), { n_embd, qkvz_dim }, TENSOR_NOT_REQUIRED);
layer.wqkv = create_tensor(tn(LLM_TENSOR_ATTN_QKV, "weight", i), { n_embd, key_dim * 2 + value_dim }, TENSOR_NOT_REQUIRED);
layer.wqkv_gate = create_tensor(tn(LLM_TENSOR_ATTN_GATE, "weight", i), { n_embd, value_dim }, TENSOR_NOT_REQUIRED);
layer.ssm_conv1d = create_tensor(tn(LLM_TENSOR_SSM_CONV1D, "weight", i), { hparams.ssm_d_conv, conv_dim }, 0);
layer.ssm_dt = create_tensor(tn(LLM_TENSOR_SSM_DT, "bias", i), { hparams.ssm_dt_rank }, 0);
layer.ssm_a = create_tensor(tn(LLM_TENSOR_SSM_A_NOSCAN, i), { hparams.ssm_dt_rank }, 0);