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.
Log in or Sign Up to review the conditions and access this model content.
Sabich-Parakeet-110M v1.0: the small Sabich, for phones, watches and wearables
Sabich trained on NVIDIA's smallest Parakeet. Sabich is the kitchen-sink line of the Hebrish project: one speaker's English dictation with everything in it, that is English with Hebrew words in it (code-switching, in canonical romanization: teudat zehut, arnona, makolet), tech vocabulary, brand names, acronyms, numbers and currency in digits, Irish-English phrasing and everyday email and message dictation. The main Sabich is Parakeet TDT 0.6B v2; this variant puts the same training data on nvidia/parakeet-tdt_ctc-110m, about a fifth of the size, for small devices: low-end phones, watches and wearables.
Name. Each fine-tune line has a food code name for what it is trained on (Sabich: everything, the pita stuffed with a bit of everything). A variant on a different base model keeps the line name and adds the base: Sabich-<Base>-<Size>. Later retrains keep the name and bump the version; the training run id is r8-ctc110m-all.
A full fine-tune of nvidia/parakeet-tdt_ctc-110m (FastConformer encoder with a TDT decoder and an auxiliary CTC head, 114M parameters) on one
speaker's read-aloud phone recordings (an Irish English speaker living in Israel), run r8-ctc110m-all, trained
2026-10-21. Output is cased and punctuated; numbers are written in digits.
Other Hebrish models: Sabich v1.0 (Parakeet TDT 0.6B v2) · Sabich-Qwen3-ASR-1.7B v1.0 (desktop GPU) · Shakshuka v1.0 (streaming) · Jachnun v1.0 (code-switching only).
Access and terms of use. Governed by the Hebrish terms of use; access is by request (button on this page, or access-requests@hebrish.app). Non-commercial use is free; commercial use needs a commercial licence.
Results
Not benchmarked yet. This release was published straight after training, before any test-set evaluation, so this card has no accuracy numbers. Its validation WER at the chosen checkpoint was 9.43% (raw, on the speaker's validation takes, used only to pick the checkpoint); that is not a test score and is not comparable across base models. Expect it to be less accurate than the 0.6B Sabich: it is the choice when size and speed matter more than the last few points of accuracy.
Training data
Source: danielrosehill/Read-Aloud-Voice-Sentences at revision e918e381e831; the mix is rebuilt from it by prepare (blend all).
- Every one of the speaker's takes, code-switching and English: code-switching sentences, Hebrew names, Yiddish and religious terms, and every task subset (tech, brand names, acronyms, Irish English, currency, numerics, hard sentences, disfluency, singing, whispered, web addresses, spelled-out, Latin phrases and the rest) except
name-cipher, whose targets contradict the rest on purpose. - Mix: 592.4 min of the speaker's takes (382.9 min code-switching) plus 66.0 min of LibriSpeech-PC replay (cased, punctuated; CC-BY-4.0) = 658.4 min, 5,634 clips.
- Validation: 1 in 16 of the speaker's train takes by id hash, used only to pick the checkpoint.
Data categories
Every category in the training data, what it is for, one real prompt, and which model line trains on it (Sabich = everything blend, Jachnun = code-switching, Boxty = no Hebrew). Generated by scripts/build_training_categories.py in the training repo.
| Category | What it trains | Example (said → written) | Sabich | Jachnun | Boxty |
|---|---|---|---|---|---|
| Code-switching | Hebrew words inside English sentences, written in the canonical romanization (the core objective). | “Do you want amba on your sabich or just tahini?” | ✓ | ✓ | – |
| Israeli names and places | Israeli and Jewish first names, surnames and place names. | “Hi Omri, the tiyul to Ein Gedi is still on for Saturday morning if you want to join us.” | ✓ | ✓ | – |
| Yiddish | Yiddish words in English; always also code-switching. | “These rugelach are fresh and absolutely delicious.” | ✓ | ✓ | – |
| Jewish religious terms | Jewish religious and observance vocabulary; always also code-switching. | “Where exactly did the Beis Hamikdash stand relative to the Old City walls?” | ✓ | ✓ | – |
| Plain English | Everyday English, kept so the model does not forget ordinary speech. | “Do we have any rice left, or should I make pasta for dinner tonight instead?” | ✓ | ✓ (English share) | ✓ |
| Dictated email | Email register: greetings, requests, sign-offs. | “Hi Dana, just following up on my email from last week, did you get a chance to look at it?” | ✓ | ✓ (English share) | ✓ |
| Messages and notes | WhatsApp messages and notes to self. | “Are you still going to the gym tonight, or skipping again?” | ✓ | ✓ (English share) | ✓ |
| Homophones | Sentences whose context settles a homophone (their / there, steak / stake). | “Could we hire a van for Saturday instead of borrowing Pete's?” | ✓ | ✓ (English share) | ✓ |
| Numbers in digits | Numbers written in digits, not words (DeepSeek 4.1, 45 seconds). | “The battery dropped from 80% to 15% in under three hours.” | ✓ | ✓ (English share) | ✓ |
| Personal vocabulary | Daniel's own names, companies and project names. | “Daniel Rosehill writes the prompts, and Corn and Herman turn them into an episode.” | ✓ | ✓ (English share) | ✓ |
| Tech vocabulary | AI and software terms, product names, commands. | “A GitHub Gist is still the fastest way to share a one-off shell script with someone.” | ✓ | ✓ (English share) | ✓ |
| Talking to AI tools | Prompts and instructions dictated to AI assistants. | “Open the model picker and switch to Sonnet for the rest of this.” | ✓ | ✓ (English share) | ✓ |
| Linux admin | Linux commands, tools and system administration. | “Run a traceroute to the VPS from the home server and see where the latency jumps.” | ✓ | ✓ (English share) | ✓ |
| Smart home | Home Assistant and smart-home devices and commands. | “SmartThings was my very first hub, before I found Home Assistant.” | ✓ | ✓ (English share) | ✓ |
| Long dictation | Paragraph-length dictation, 50 to 110 words in one take. | “Quick update on the server situation. The WhatsApp bot stopped replying last night because uvicorn crashed after I updated Ubuntu. …” | ✓ | ✓ (English share) | ✓ |
| Field corrections | Free speech the stock model got wrong, corrected by hand (Correct mode). | ✓ | ✓ (English share) | ✓ | |
| Currency | Spoken amounts written as typed: fifty shekels becomes 50 NIS, twenty dollars $20. | The license is 99 dollars once, no subscription. → “The license is $99 once, no subscription.” | ✓ | ✓ (English share) | ✓ |
| Numbers (not money) | Temperatures, decimals, percentages, units, times, dates and ordinals in digits. | It was minus two degrees Celsius in Berlin this morning. → “It was -2 degrees Celsius in Berlin this morning.” | ✓ | – | ✓ |
| Web addresses | Addresses said aloud (dot, slash, colon) written as typed. | The meeting notes are at example dot com slash notes slash october dash twenty. → “The meeting notes are at example.com/notes/october-20.” | ✓ | – | ✓ |
| Spelling out | Letters spelled aloud written as the word they spell. | My surname is B R E N N A N. → “My surname is Brennan.” | ✓ | – | ✓ |
| Acronyms | Letter-by-letter acronyms (API, SSH) and word acronyms (NASA, JSON) in their standard form. | “The ROI on that campaign was higher than anyone expected.” | ✓ | – | ✓ |
| Expanded acronyms | Chat and business shorthand said as letters, written out (EOD becomes end of day). | The documentation is the weakest part of the project, IMO. → “The documentation is the weakest part of the project, in my opinion.” | ✓ | – | ✓ |
| Brand and product names | Company, product, AI model and open-source names in their official spelling. | “I asked Claude to review the pull request, and it caught a typo in the config.” | ✓ | – | ✓ |
| Procurement | Incoterms, POs, RFQs, lead times and payment terms. | “Ask them if they'll lower the MOQ for the first order and treat it as a trial.” | ✓ | – | ✓ |
| Idioms | Idioms ASR tends to spell as a sound-alike (free rein, toe the line). | “We gave the contractor free rein on the kitchen layout, and honestly it turned out better than my plan.” | ✓ | – | ✓ |
| Latin phrases | Latin phrases used in English (de facto, ad hominem, de minimis) in their standard spelling: lowercase, no hyphen. | “We ended up with a de facto ban on parking outside the school after the council's new notice.” | ✓ | – | ✓ |
| Hard sentences | Harvard sentences, tongue twisters, homographs, garden paths, rare words. | “The birch canoe slid on the smooth planks.” | ✓ | – | ✓ |
| Irish English | Irish English words, grammar and Irish-language names with fadas. | “The new lad at work is deadly, he sorted out the printer in 5 minutes.” | ✓ | – | ✓ |
| Disfluency reduction (experimental) | Read with uh, um and eh; the transcript stays clean. | Read with hesitations: uh, um, eh. → “Honestly, I never properly thanked her for helping us last week, and it's bugging me now.” | ✓ | – | ✓ |
| Repeated words (experimental) | Words repeated the way people restart; the transcript stays clean. | Read repeating a word now and then (the the report). → “Has anyone heard how she's doing today? She sounded really rough on the phone yesterday and I'm getting worried.” | ✓ | – | ✓ |
| False starts (experimental) | Abandoned starts and doubled words; the transcript drops them. | It's not that I — I just don't think we have time for it this week. → “I just don't think we have time for it this week.” | ✓ | – | ✓ |
| Self-corrections (experimental) | A spoken correction (Tuesday, no, Wednesday); the transcript keeps only the corrected version. | Leave the box by the front door, no, by the back door, please. → “Leave the box by the back door, please.” | ✓ | – | ✓ |
| Asides (experimental) | A remark to someone else in the room mid-dictation; it stays out of the transcript. | We need to buy a new washing machine — Ezra, not now, sweetie — before the old one breaks down completely, ideally under $600. → “We need to buy a new washing machine before the old one breaks down completely, ideally under $600.” | ✓ | – | ✓ |
| Paragraphs (experimental) | Whole passages in one take, for paragraph inference later. | “Things to sort out with the internet this week. The router keeps dropping the connection for a minute or so in the evenings. …” | ✓ | – | ✓ |
| Whispered (experimental) | Whispered speech (robustness). | Whispered. → “Honestly, I think the kids are getting too much homework for their age, and nobody at school seems to listen.” | ✓ | – | ✓ |
| Singing (experimental) | Sung sentences (robustness). | Sung, any tune. → “Do you think it's rude to give someone a gift card, or is it actually more useful than guessing what they like?” | ✓ | – | ✓ |
| Weird voices (experimental) | Put-on voices: deep, squeaky, robot, accents (robustness). | Read in a put-on voice. → “Do you think it's rude if we don't reply to the wedding invitation until we know whether the kids can come?” | – | – | – |
| Name cipher (PII test) (experimental) | Experiment: real names read, swapped by rule in the transcript (Hannah becomes Sarah). | Can you tell Daniel the meeting moved to 3 PM? → “Can you tell Dave the meeting moved to 3 PM?” | – | – | – |
✓ = in the training recipe. “English share” = competes for Jachnun's 30% monolingual-English share. Experimental sets are recorded but off in the capture app by default. This is the current recipe; a published run's exact composition is in its blend file.
Training
NeMo 2.7.3 on one L40S (Modal), bf16. AdamW, lr 3e-05 (cosine, 10% warmup), 20 epochs,
batch 16, seed 7. Best checkpoint best-epoch=17-val_wer=0.0943.ckpt,
39.8 min of training.
Decoding uses the TDT branch (the default). The CTC head was trained jointly, as in the base model's hybrid loss.
Usage
import nemo.collections.asr as nemo_asr
from huggingface_hub import hf_hub_download
path = hf_hub_download("Hebrish/Sabich-Parakeet-110M", "Sabich-Parakeet-110M.nemo")
model = nemo_asr.models.ASRModel.restore_from(path)
print(model.transcribe(["clip.wav"])[0].text) # 16 kHz mono
Use the TDT branch (default). For CTC decoding: model.change_decoding_strategy(decoder_type="ctc") (not checked on this fine-tune).
Builds: Hebrish/Sabich-Parakeet-110M-sherpa-onnx-int8 (sherpa-onnx int8 transducer, offline) and Hebrish/Sabich-Parakeet-110M-GGUF (transcribe.cpp: Q8_0, Q4_K_M, F16).
Limitations
- One speaker. It is adapted to one voice, accent and phone; others will see less benefit.
- Not benchmarked: no test-set numbers yet.
- About a fifth of the size of the 0.6B Sabich; expect lower accuracy, especially on rare Hebrew terms.
- Disfluent and sung takes were trained on with clean targets, so the model may drop some fillers and repeated words.
Licence
CC-BY-4.0, from the base model's licence, with use governed by the Hebrish terms above. Base model by NVIDIA.
- Downloads last month
- -