Instructions to use bearcove/zipa-small-crctc-ns-no-diacritics-700k-mlx-q8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use bearcove/zipa-small-crctc-ns-no-diacritics-700k-mlx-q8 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir zipa-small-crctc-ns-no-diacritics-700k-mlx-q8 bearcove/zipa-small-crctc-ns-no-diacritics-700k-mlx-q8
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
ZIPA Small CR-CTC NS No-Diacritics 700k (MLX Q8)
This repository contains an MLX-native Q8 checkpoint bundle for the ZIPA small CR-CTC non-streaming no-diacritics model.
What This Is
- Base model family:
anyspeech/zipa-small-crctc-ns-no-diacritics-700k - Published repo target:
bearcove/zipa-small-crctc-ns-no-diacritics-700k-mlx-q8 - Quantization target: MLX Q8
- Intended consumer: the Bee repository
- Reference implementation:
rust/bee-zipa-mlxinside Bee
This is a Bee model bundle. It is not currently intended as a standalone general-purpose MLX package with its own separately published loader crate.
Files
model.safetensors: quantized MLX checkpoint in thezipa-mlx-quantized-v1formattokens.txt: CTC vocabularyconfig.json: model architecture and quantization metadata
Quantization Scheme
- Linear layers are quantized to 8-bit where MLX supports the target group size
- Small incompatible projections remain dense inside the checkpoint
- Norms, bypass scales, convolution weights, and downsample weights remain dense
- Group size:
64
Usage From Bee
Example from a Bee checkout:
cargo run -q -p bee-zipa-mlx --bin zipa-infer -- \
--quantized-checkpoint /path/to/model.safetensors \
/path/to/audio.wav
Notes
- This bundle was generated from local dense ZIPA reference artifacts and quantized with
bee-zipa-mlx - The artifact layout is currently project-specific to Bee
- The bundle is currently marked
MITbased on the upstream ZIPA repository license and the upstream release statement that code and pretrained models would be released under permissive licenses. This should still be confirmed directly with upstream model authors.
- Downloads last month
- 26
Model size
18.7M params
Tensor type
F32
·
U32 ·
Hardware compatibility
Log In to add your hardware
Quantized