ollama/create
Daniel Hiltgen cecac8bf1b
mlxrunner: add Inkling 2 Mini (#19799)
* mlxrunner: add Inkling 2 Mini

* mlx: fuse Inkling 2 Mini residual mixing on Metal

Fuse RMS normalization and query scoring in one kernel. Apply the
softmax weights in a second kernel to avoid storing per-query products.

* create: expose Inkling 2 Mini context length in model metadata

Copy model_max_length to max_position_embeddings during import so
ollama show reports the checkpoint's context length. Reject configs
that specify conflicting limits.

* mlxrunner: reduce Metal dispatches in Inkling 2 Mini decode

Reuse the prefill mask shader to compute relative position bias in one
dispatch for single-query attention. On M3 Ultra with Metal 3, compute
the loop controller's gain and forget projections in one matrix multiply.
2026-11-20 18:38:44 -08:00
..
autotrust_decision_test.go Add AutoTrust decision model to MLX (#19399) 2026-11-04 09:47:37 -08:00
blockfp8.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
blockfp8_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
classify.go mlxrunner: add Inkling 2 Mini (#19799) 2026-11-20 18:38:44 -08:00
classify_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
clef.go mlx: add PPLX Decider v1.2 support (#19072) 2026-10-17 14:45:55 -07:00
clef_test.go mlx: System one support (#18701) 2026-10-03 13:17:21 -07:00
cohere2moe.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
create.go Add AutoTrust decision model to MLX (#19399) 2026-11-04 09:47:37 -08:00
create_test.go New model... (#19030) 2026-10-15 07:28:26 -07:00
d3.go mlx: add d3 decision model support (#19196) 2026-10-22 08:50:14 -07:00
d3_test.go mlx: add d3 decision model support (#19196) 2026-10-22 08:50:14 -07:00
draft.go mlx: System one support (#18701) 2026-10-03 13:17:21 -07:00
draft_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
filetype.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
filetype_test.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
gemma4.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
gemma4_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
gemma4embedding.go model: add multimodal embeddings (#18820) 2026-10-06 08:08:38 -07:00
gemma4embedding_test.go model: add multimodal embeddings (#18820) 2026-10-06 08:08:38 -07:00
glimmer.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
glimmer_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
inkling2.go mlxrunner: add Inkling 2 Mini (#19799) 2026-11-20 18:38:44 -08:00
inkling_test.go mlxrunner: add Inkling 2 Mini (#19799) 2026-11-20 18:38:44 -08:00
inventory.go Add AutoTrust decision model to MLX (#19399) 2026-11-04 09:47:37 -08:00
inventory_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
kev.go Add Kev decision scoring to MLX (#18976) 2026-10-12 21:50:41 -07:00
kev_test.go Add Kev decision scoring to MLX (#18976) 2026-10-12 21:50:41 -07:00
laguna.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
laguna_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
laya.go mlx: System one support (#18701) 2026-10-03 13:17:21 -07:00
laya_test.go mlx: System one support (#18701) 2026-10-03 13:17:21 -07:00
lfm3.go llm: LFM3 GGUF fixes and llama-server speech (#19344) 2026-10-27 18:56:22 -07:00
lfm3_test.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
lfm3speech.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
lfm3speech_test.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
manifest.go Merge remote-tracking branch 'upstream/main' into release_v0.40.0 2026-09-29 11:15:31 -07:00
manifest_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
metadata.go Add ripple preview support to the MLX runner (#19592) 2026-11-07 16:36:53 -08:00
metadata_test.go Add ripple preview support to the MLX runner (#19592) 2026-11-07 16:36:53 -08:00
mlxthread.go mlx: reject unsupported CUDA devices (#19521) 2026-11-04 09:57:25 -08:00
nemotron_h.go New model... (#19030) 2026-10-15 07:28:26 -07:00
nemotron_h_test.go New model... (#19030) 2026-10-15 07:28:26 -07:00
phonon.go voice: add voice catalog for Phonon (#19242) 2026-10-23 16:15:08 -07:00
phonon_test.go voice: add voice catalog for Phonon (#19242) 2026-10-23 16:15:08 -07:00
pipeline.go mlxrunner: add Inkling 2 Mini (#19799) 2026-11-20 18:38:44 -08:00
pipeline_test.go mlx: add LFM3 and d1 models (#19333) 2026-10-27 08:47:18 -07:00
plan.go mlx: add PPLX Decider v1.2 support (#19072) 2026-10-17 14:45:55 -07:00
plan_test.go mlx: add Flint model support (#19181) 2026-10-21 07:45:50 -07:00
pplx.go mlx: add d3 decision model support (#19196) 2026-10-22 08:50:14 -07:00
pplx_test.go mlx: add d3 decision model support (#19196) 2026-10-22 08:50:14 -07:00
prequant.go mlxrunner: add Inkling 2 Mini (#19799) 2026-11-20 18:38:44 -08:00
prequant_test.go mlx: add Flint model support (#19181) 2026-10-21 07:45:50 -07:00
quantize.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
quantpolicy.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
qwen4_exp.go MLX: make Qwen3.8 Flash Next memory efficient (#18078) 2026-11-06 13:39:21 -08:00
qwen4_exp_test.go MLX: make Qwen3.8 Flash Next memory efficient (#18078) 2026-11-06 13:39:21 -08:00
qwen35.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
qwen35_decision.go Add Kev decision scoring to MLX (#18976) 2026-10-12 21:50:41 -07:00
qwen35_decision_test.go Add AutoTrust decision model to MLX (#19399) 2026-11-04 09:47:37 -08:00
transform.go mlx: add PPLX Decider v1.2 support (#19072) 2026-10-17 14:45:55 -07:00
verify.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
verify_test.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
writer.go mlx, mlxrunner: move the MLX engine out of x/ 2026-09-16 14:06:08 -07:00
writer_test.go mlx: add Flint model support (#19181) 2026-10-21 07:45:50 -07:00