Downloads · 30 days
19
20% of all-time downloads
MossFormer2-SE-4bit
starkdmi/MossFormer2-SE-4bit
MossFormer2-SE-4bit is a machine learning model from starkdmi. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for mlx. The card lists the license as apache-2.0.
48kHz speech enhancement model converted to MLX format (4-bit quantized).
All-time downloads
94
Public
Parameters
55.3M
67.3 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors67.3 MB · 100%
Parameter types
How the weights are stored.
U3245.6M · 83%
Model card
From the Hugging Face model README
MossFormer2 SE 4-bit (MLX)
48kHz speech enhancement model converted to MLX format (4-bit quantized).
Original Model
alibabasglab/MossFormer2_SE_48K
Usage
from mlx_audio.sts.models.mossformer2_se import MossFormer2SEModel
model = MossFormer2SEModel.from_pretrained("starkdmi/MossFormer2-SE-4bit")
enhanced = model.enhance("noisy.wav")
Precision Variants
- MossFormer2-SE (fp32, 211MB)
- MossFormer2-SE-fp16 (fp16, 106MB)
- MossFormer2-SE-8bit (int8, 86MB)
- MossFormer2-SE-6bit (int6, 75MB)
- MossFormer2-SE-4bit (int4, 64MB) ← this
Performance
~30x real-time on Apple M4
- Library
- mlx
- Type
- mossformer2_se
- License
- apache-2.0
- Created
- Jan 1, 2026
- Updated
- Jan 1, 2026