vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-07-10 01:18:32 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

model : register t_layer_inp for qwen3next (#25141)

Browse source 
* Fix input assignment in layer processing loop

Fix DFLASH for qwen-coder-next

* add line break

Added tensor for attention normalization in Qwen3 model.
This commit is contained in:
Jürgen Schmied 2026-06-30 17:57:14 +02:00 • committed by GitHub
parent 799fcc04a5
commit 4f31eedb0c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 2 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/models/qwen3next.cpp
View file

@ -121,6 +121,8 @@ llama_model_qwen3next::graph::graph(const llama_model & model, const llm_graph_p
ggml_tensor * inp_out_ids = build_inp_out_ids();
for (int il = 0; il < n_layer; ++il) {
res->t_layer_inp[il] = inpL;
ggml_tensor * inpSA = inpL;
cur = build_norm(inpL, model.layers[il].attn_norm, nullptr, LLM_NORM_RMS, il);