Downloads · 30 days
33
25% of all-time downloads
Ornith-1.5-35B-A3B-MTP-2bit-MLX
lukaskremla/Ornith-1.5-35B-A3B-MTP-2bit-MLX
Ornith-1.5-35B-A3B-MTP-2bit-MLX is a text generation model from lukaskremla. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as mit.
This repository contains a 2-bit affine quantization of the native Multi-Token Prediction (MTP) tensors extracted from ornith-ai/Ornith-1.5-35B-A3B and converted into the standalone MLX drafter format expected by mlx-…
All-time downloads
131
Public
Parameters
845M
284 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors264 MB · 91%
Parameter types
How the weights are stored.
U32845M · 100%
Try a prompt
Model card
From the Hugging Face model README
lukaskremla/Ornith-1.5-35B-A3B-MTP-2bit-MLX
This repository contains a 2-bit affine quantization of the native Multi-Token Prediction (MTP) tensors extracted from ornith-ai/Ornith-1.5-35B-A3B and converted into the standalone MLX drafter format expected by mlx-vlm.
This is not a standalone language model. It must be loaded as the draft model alongside a compatible Ornith-1.5-35B-A3B target model.
This 2-bit drafter can be used with the 2-bit, 3-bit, 4-bit, 5-bit, 6-bit and 8-bit MLX target models. Quantizations don't need to match.
You can download the vision capable or text-only Ornith 1.5 35B-A3B weights in a variety of MLX quantizations from this collection Ornith 1.5 35B-A3B MLX-Quants (Vision, Text-Only & MTP).
Hugging Face might render incorrect parameter counts for this model, it is a common display bug for MLX quants.
Credits
Converted to MLX format from ornith-ai/Ornith-1.5-35B-A3B using mlx-vlm version 0.6.17.
- Task
- Text Generation
- Library
- mlx
- Type
- qwen3_5_mtp
- License
- mit
- Created
- Sep 1, 2026
- Updated
- Sep 1, 2026