Skip to content
ModelListedExternal
Download

Mlx MOSS Transcribe Diarize 4bit

4bit MLX weight-only quantized variant of vanch007/mlx-MOSS-Transcribe-Diarize.

KindText modelBase modelvanch007/mlx-MOSS-Transcribe-DiarizeVersionmodel

Model source

Kind
Text model
File size
916 MB
Base model
vanch007/mlx-MOSS-Transcribe-Diarize
Version
vmodel
Updated
Sep 28, 2026
Downloads
306
Source
Hugging Face

Source description

mlx-MOSS-Transcribe-Diarize 4bit

4bit MLX weight-only quantized variant of vanch007/mlx-MOSS-Transcribe-Diarize.

Quantization uses MLX affine quantization with bits=4 and group_size=64 on the text backbone, excluding the Whisper audio encoder and VQ adaptor.

python -m moss_transcribe_diarize.mlx.cli /path/to/input.wav \
  --model vanch007/mlx-MOSS-Transcribe-Diarize-4bit \
  --out-dir runs/mlx_4bit_example

Sources

1 source
Hugging FacePrimaryVersion d42a296ee807e933ddd7588e2041dbbc84aff85d
huggingface.co/vanch007/mlx-MOSS-Transcribe-Diarize-4bit/tree/d42a296ee807e933ddd7588e2041dbbc84aff85d

Verified Sep 22

Model artifacts

1 artifact

model.safetensors

safetensors · 916 MB · SHA-256 0483be31b7ad…a3c4 · Hugging Face

Download file

Source excerpts

2 excerpts

--- license: apache-2.0 library_name: mlx tags: - mlx - int4 - automatic-speech-recognition - speaker-diarization base_model: - vanch007/mlx-MOSS-Transcribe-Diarize pipeline_tag: automatic-speech-recognition --- # mlx-MOSS-Transcribe-Diarize 4bit 4bit MLX weight-only quantized variant of `vanch007/mlx-MOSS-Transcribe-Diarize`. Quantization uses MLX affine quantization with `bits=4` and `group_size=64` on the text backbone, excluding the Whisper audio encoder and VQ adaptor. ```bash python -m moss_transcribe_diarize.mlx.cli /path/to/input.wav \ --model vanch007/mlx-MOSS-Transcribe-Diarize-4bit \ --out-dir runs/mlx_4bit_example ```

Source context: 306 downloads · 0 likes · Pipeline automatic-speech-recognition · Library mlx · Repo vanch007/mlx-MOSS-Transcribe-Diarize-4bit