CERBERUS/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-10-20 18:41:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
ollama/mlxrunner/nn
History
Exact
Daniel Hiltgen c0f8da35b2
mlx: speed up Qwen 3.8 prompt processing (#18550) 
* mlx: speed up Qwen 3.8 prompt processing

Use MLX's gated-delta kernel for long scans and fold dense MLP global scales into SwiGLU.

* address comments
2026-09-22 17:53:23 -07:00
..
conv.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
embedding.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
embedding_test.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
linear.go mlx: speed up Qwen 3.8 prompt processing (#18550) 2026-09-22 17:53:23 -07:00
linear_test.go mlx: speed up Qwen 3.8 prompt processing (#18550) 2026-09-22 17:53:23 -07:00
multilinear.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
norm.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
norm_test.go mlxrunner: lay out nn one layer kind per file 2026-09-16 14:06:08 -07:00
recurrent.go
recurrent_test.go
rope.go
sdpa.go
sdpa_test.go