🔄 In a Training Loop

Barney Greenway

McG-221

·

AI & ML interests

LLMs on Apple Silicon

Recent Activity

liked a model 1 day ago

Jackrong/Qwopus3.8-27B-Flash-GGUF

liked a model 1 day ago

mradermacher/Gemma-4-MoonGem-31B-GGUF

liked a model 1 day ago

Ateron/Gemma-4-MoonGem-31B

View all activity

Organizations

upvoted a collection 17 days ago

Unsloth Dynamic 3.0 Quants

Introducing Dynamic V3.0 quants, our new SOTA quantization methodology. • 1 item • Updated 11 days ago • 63

upvoted a collection 21 days ago

Qwen3.8

4 items • Updated 24 days ago • 493

upvoted a collection 22 days ago

Qwen3.8

Qwen3.8 models converted to MLX format for Apple silicon. • 12 items • Updated 22 days ago • 20

upvoted a collection about 1 month ago

Recent medium models

57 items • Updated 6 days ago • 9

upvoted 2 collections about 2 months ago

Bonsai 27B

10 items • Updated Jul 15 • 232

NVIDIA NVFP4 → MLX

NVIDIA post-trained NVFP4 checkpoints (ModelOpt PTQ) transcoded to MLX for Apple Silicon / mlx-node, keeping NVIDIA's per-layer bit allocation. • 1 item • Updated Jul 21 • 2

upvoted 2 collections 2 months ago

Ornith-1.0

Ornith-1.0 is a family of open-source LLMs specialized for agentic coding. • 8 items • Updated 18 days ago • 394

Ornith 1.0

MLX versions of Ornith 1.0 • 6 items • Updated Jun 27 • 4

upvoted 4 collections 3 months ago

Gemma 4 QAT

Gemma 4 QAT (Quantization-Aware Training) for 3x less memory use and near original accuracy. • 16 items • Updated 11 days ago • 120

Qwen 3.x MTP

MLX MTP drafter checkpoints for Qwen 3.x speculative decoding with mlx-vlm. • 12 items • Updated Jun 1 • 9

Gemma 4 QAT Q4_0

19 items • Updated Jul 21 • 161

Gemma 4 QAT

40 items • Updated Jun 5 • 13

upvoted 6 collections 4 months ago

Gemma-4-31B-IT-unsloth-mlx

Gemma-4-31B-IT (dense vision-language) quantized for Apple Silicon (MLX) — Unsloth Dynamic 2.0 with AWQ imatrix pre-scaling. • 9 items • Updated May 16 • 1

Gemma-4-26B-IT-unsloth-mlx

Gemma-4-26B-A4B-IT MoE quantized for Apple Silicon (MLX) — Unsloth Dynamic 2.0 with AWQ imatrix pre-scaling. • 8 items • Updated May 16 • 1

Magistry

Q1 2026 merges based on Mistral 24B • 2 items • Updated Mar 22 • 2

Qwen-3.6-unsloth-mlx

AWQ-style pre-scaling using Unsloth's imatrix calibration data, then 3-6-bit affine quantization with the Unsloth mixed-precision recipe via MLX • 18 items • Updated May 15 • 19

gemma-4-abliterated

21 items • Updated Jul 11 • 35

Uncensored Gemma 4 MLX

10 items • Updated Apr 16 • 2

upvoted 2 collections 5 months ago

Qwen3.6

4 items • Updated Apr 22 • 495

Qwen3.5

Qwen3.5 is Qwen's new model family including Qwen3.5 Small: 0.8B, 2B, 4B, 9B and Qwen3.5 Medium: 35B-A3B, 27B, 122B-A10B and 397B-A17B. • 25 items • Updated 11 days ago • 167