Hugging Face's logo

ToPo-ToPo
/
Qwen3.6-27B-mlx-4bit

Image-Text-to-Text
MLX
Safetensors
qwen3_5
vision
conversational
4-bit precision
Model card Files
xet
 Community

ToPo-ToPo/Qwen3.6-27B-mlx-4bit

MLX 4bit conversion of Qwen/Qwen3.6-27B for Apple Silicon (mlx-vlm).

Provenance (self-converted from official weights)

  • Source: Qwen/Qwen3.6-27B (license: apache-2.0)
  • Tool: mlx-vlm 0.6.3 — mlx_vlm.convert --hf-path Qwen/Qwen3.6-27B --mlx-path . -q --q-bits 4 --q-group-size 64
  • Effective: 4.695 bits/weight
  • Validation: reproduced geometrically exact CAD output in an agentic CAD+FEM pipeline (volumes match the reference mlx-community conversion).

Usage 

from mlx_vlm import load, generate
model, processor = load("ToPo-ToPo/Qwen3.6-27B-mlx-4bit")
Downloads last month
29
Safetensors
Model size
5B params
Tensor type
BF16
·
U32
·
MLX
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
Image-Text-to-Text
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ToPo-ToPo/Qwen3.6-27B-mlx-4bit

Base model

Qwen/Qwen3.6-27B
Quantized
(533)
this model

Collection including ToPo-ToPo/Qwen3.6-27B-mlx-4bit