MossFormer2 SE (MLX)

48kHz speech enhancement model converted to MLX format.

Original Model

alibabasglab/MossFormer2_SE_48K

Usage

from mlx_audio.sts.models.mossformer2_se import MossFormer2SEModel

model = MossFormer2SEModel.from_pretrained("starkdmi/MossFormer2-SE")
enhanced = model.enhance("noisy.wav")

Precision Variants

Performance

~30x real-time on Apple M4

Downloads last month
3,991
Safetensors
Model size
55.3M params
Tensor type
F32
Β·
MLX
Hardware compatibility
Log In to add your hardware

Quantized

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Space using starkdmi/MossFormer2-SE 1