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

models : optimize qwen3next graph (#19375)

Browse Source 
* models : optimizing qwen3next graph

* cont

* wip

* wip

* wip

* wip

* wip

* wip

* wip

* wip

* wip

* wip

* cont : remove redundant q, g chunking

* minor

* minor

* avoid passing masks around

* avoid concats during chunking

* naming + shapes

* update names and use prefix to disable CUDA graphs
This commit is contained in:
Georgi Gerganov
2026-02-14 12:57:36 +02:00
committed by GitHub
parent b7742cf321
commit 1725e316c1
4 changed files with 263 additions and 300 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-metal/ggml-metal-common.cpp
+1
View File
@@ -273,6 +273,7 @@ static std::vector<int> ggml_metal_graph_optimize_reorder(const std::vector<node
case GGML_OP_DIAG:
case GGML_OP_MUL:
case GGML_OP_ADD:
case GGML_OP_SUB:
case GGML_OP_DIV:
case GGML_OP_GLU:
case GGML_OP_SCALE: