]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 799fcc0)
model : register t_layer_inp for qwen3next (#25141)
authorJürgen Schmied <redacted>
Tue, 30 Jun 2026 15:57:14 +0000 (17:57 +0200)
committerGitHub <redacted>
Tue, 30 Jun 2026 15:57:14 +0000 (17:57 +0200)
commit4f31eedb0ccf546b7e8d6bb243b170f12522f54d
tree39fd03672b587ef09becf117f8378cbadaba5175tree
parent799fcc04a5bbe2823e698b9e524f5a6cd0c594b6commit | diff
model : register t_layer_inp for qwen3next (#25141)

* Fix input assignment in layer processing loop

Fix DFLASH for qwen-coder-next

* add line break

Added tensor for attention normalization in Qwen3 model.
src/models/qwen3next.cpp diff | blob | history
Packaging of ggml-org/llama.cpp