Brooooooklyn/Qwen3.6-27B-NVFP4-mlx
Text Generation • 9B • Updated • 902
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.
Note NVIDIA Qwen3.6-27B-NVFP4 → MLX (nvfp4 MLP + mxfp8 attn, bf16 MTP head).
-
Brooooooklyn/Gemma-4-26B-A4B-NVFP4-mlx
Text Generation • 9B • Updated • 367 • 1