augible
/

qwen3.5-2b-4bit-text-mlx

4-bit precision

Model card Files Files and versions

Qwen3.5-2B — 4-bit MLX, text-only layout

The text decoder of Qwen/Qwen3.5-2B (Apache-2.0), quantized to 4-bit with mlx-lm (group 64, affine) and re-keyed to the qwen3_5_text layout (model.* tree) so mlx-swift-lm's Qwen35TextModel + PEFT adapter loader consume it directly. The chat template is pinned to non-thinking rendering. Base weights only — no fine-tuning of any kind.

Downloads last month: 94

Safetensors

Model size

0.3B params

Tensor type

BF16

·

U32

·

F32

·

MLX

Hardware compatibility

Log In to add your hardware

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for augible/qwen3.5-2b-4bit-text-mlx

Base model

Qwen/Qwen3.5-2B-Base

Finetuned

Qwen/Qwen3.5-2B

Quantized

(150)

this model