You need to agree to share your Hugging Face information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Access to this model is by request and is governed by the Hebrish terms of use (https://hebrish.app/terms). Non-commercial use is free; commercial use needs a commercial licence from access-requests@hebrish.app. Please also complete the attestation form: https://hebrish.app/legal/hebrish-terms-of-use.pdf

Jachnun PoC-3 (MLX): English-Hebrew code-switching

MLX weights for Apple Silicon of danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching, a fine-tune of nvidia/parakeet-tdt-0.6b-v2 for English speech with Hebrew words dropped in ("I have to go to Bituach Leumi"). Hebrew words come out in Latin romanization. Model details, training data and evaluation are on the main card. This is PoC-3 (published as v3, run r3-en30), a proof-of-concept checkpoint. It is not the current release, Jachnun v1.0, which has not been converted to MLX yet.

Website: hebrish.app: get the Hebrish keyboard for Android and the desktop apps, see the results, and find every model, dataset and demo in one place. Android updates through F-Droid: add the Hebrish F-Droid repository (https://fdroid.hebrish.app/repo; steps).

Access and terms of use. Use of this model and every converted or quantised build of it (GGUF, ONNX, sherpa-onnx, MLX) is governed by the Hebrish terms of use. Access is by request: use the button on this page, or write to access-requests@hebrish.app. Non-commercial use (personal, research, education, open-source projects) is free; please complete the attestation form. Commercial use, including telephony and call centres, needs a commercial licence: access-requests@hebrish.app. Versions released under CC-BY-4.0 (PoC-1 to PoC-5, v1.0) stay under that licence; the terms do not restrict rights it grants.

Usage

parakeet-mlx:

pip install parakeet-mlx
parakeet-mlx audio.wav --model Hebrish/Jachnun-MLX

from parakeet_mlx import from_pretrained
model = from_pretrained("Hebrish/Jachnun-MLX")
print(model.transcribe("audio.wav").text)

mlx-audio:

python -m mlx_audio.stt.generate --model Hebrish/Jachnun-MLX --audio audio.wav --output-path out --format txt

Results

Benchmark results for every checkpoint, PoC-3 included, are at hebrish.app/benchmarks. They were scored with the main model, not this MLX conversion; the check below compares the two. One speaker's voice: all training and test audio is the same speaker, so gains on other voices are untested.

Conversion and check

Converted from training run r3-en30 with the remap behind the stock mlx-community Parakeet repos (senstella's script), fp32 weights (parakeet-mlx casts to bf16 at load). Checked with parakeet-mlx 0.5.3 on MLX's Linux CPU backend, against the NeMo PyTorch output on held-out code-switching test clips:

	Identical to NeMo	WER vs NeMo	WER vs reference
NeMo (PyTorch)			4.4%
MLX bf16 (default)	7/10	1.5%	4.4%
MLX fp32	7/10	1.5%	4.4%

WER here is raw (cased and punctuated) on a small sample; the main card has the full evaluation. The conversion is a direct key remap (no retraining or quantization), following the script above.