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@hebrish.app.

Meorav v2.2

A single-speaker text-to-speech fine-tune of Qwen3-TTS-12Hz-1.7B-Base on one voice: Daniel Rosehill, Irish-accented English with Hebrew words mixed in ("Hebrish"). Meorav is Hebrew for "mixed", as code-switching is. First released in Hebrish Release v2 (2026-10-22), alongside Sabich v2.0 and Shakshuka v2.0.

v2.2 (2026-11-09) keeps v2.1's training recipe and changes the data: 3,995 takes (about 5.5 hours) instead of 1,103 (88 min). Besides the read TTS takes, it adds 2,924 of Daniel's speech-to-text recording takes, screened to code-switching sentences only (no tech vocabulary, no digits, no noisy rooms; sound events tagged with PANNs, quality scored with DNSMOS). Every take was enhanced with NVIDIA NemotronLabs RE-USE (OpenMDW-1.1), run as the audio-cpp/RE-USE-GGUF build through audio.cpp, in place of v2.1's DeepFilterNet 3. Sidon v0.1 was also tried: both removed a crying baby completely without the robotic tinge of older denoisers; RE-USE was chosen on Daniel's listening and a spectrogram comparison (2026-11-08). Takes longer than 15 s were left out of training.

Access and terms of use. Governed by the Hebrish terms of use; access is by request (button on this page, or access@hebrish.app). Non-commercial use is free; commercial use needs a commercial licence.

It has not been benchmarked: no listening score or metric is reported here.

Listen

Before and after, on the same text. Before: the untuned base model, Qwen3-TTS-12Hz-1.7B-Base, cloning the voice from one reference recording. After: Meorav v2.2. Nothing here was picked by ear. Training takes were at most 15 s, so the longer passages (more than three sentences) are generated sentence by sentence and joined.

Text	Before	After
Yalla, let's go, we're already late for the brit mila.
We're doing a quiet shabbat this week, just challah, some chamin and an early night.
I spent the whole morning at Misrad Hapnim, and I still don't have the teudat zehut.
Everyone here says everything will happen acharei hachagim. I used to find it frustrating, but honestly, after a few years you start saying it yourself.
On Friday morning I went down to the shuk for the usual run. The guy at the bourekas stall remembered me, gave me an extra one, and told me to say shabbat shalom to the kids.
The arnona bill came in twice this month, so I rang the iriya. After forty minutes on hold they told me, beseder, it's a mistake, just ignore the second one.
When I started ulpan, the morah told us not to worry about the grammar yet. Just listen, she said, and the words will come. She was right about most of it, although I still mix up my verb patterns when I'm tired. These days I can argue with the makolet guy about the price of cottage cheese, which feels like real progress.
Our shul does a kiddush after davening every shabbat, and it has become the real centre of the week. You hear about every simcha, every new oleh, and who is looking for a chavruta. The gabbai keeps a list of who is bringing what, and somehow there is always too much kugel. If you are new to the neighbourhood, honestly, just turn up, nobody will let you stand on your own for long.
Last chol hamoed we did a tiyul up in the Galil with the whole mishpacha. We stopped for shakshuka at a little place near the Kinneret, then walked along the Jordan for an hour. My father-in-law kept telling everyone it was nothing compared to the old days, before the road was paved. By the time we got back to Yerushalayim it was nearly midnight, and the kids were fast asleep in the back.
When we made aliyah, the first few months were a blur of paperwork. There was the misrad haklita, the kupat cholim, Bituach Leumi, and a bank that wanted three copies of everything. Our neighbour, a sabra who had seen it all before, sat us down with a cup of coffee and said, savlanut, it all gets done eventually. She was right. By our first Pesach in the country we had a lease, a car, and a teudat oleh that was already falling apart from use. Looking back now, the bureaucracy is not what I remember. What I remember is the people who helped, without being asked, and the way the whole street seemed to come out for Yom Haatzmaut.

More on the public samples page, Hebrish/Meorav-Samples: every line from the release run next to Daniel's real recordings, and longer passages generated in one pass.

Listen: what Hebrish is and how it's made

A spoken introduction to the project by the person who makes it: why Hebrish exists, why stock speech-to-text gets Hebrew words wrong, how the fine-tunes are trained, the model lines (Sabich, Jachnun, Shakshuka, Meorav), how the training data is recorded, and the plans for the apps.


Listen	hebrish.app/updates: player, chapters, edited transcript, MP3 download
Author	Daniel Rosehill
Recorded	2026-10-22
Runtime	1:07:03

Chapters (each opens the player at that point): 0:00 Why this voice note · 1:36 Coming to speech to text · 6:36 Why Hebrew words go wrong · 9:23 How the fine-tune works · 12:27 Why Parakeet, and the model plan · 17:44 Disfluencies, shekels, paragraphs · 25:40 Tech vocabulary and Meorav · 29:07 The kitchen sink: Sabich and Jachnun · 33:35 Licensing, accents, noisy recordings · 39:10 One recorder for ASR and TTS · 41:04 Desktop shortcuts and the macro pad · 43:34 Android: Bluetooth and one button · 45:35 Streaming and Shakshuka · 50:28 Long-form transcription · 56:00 Output modes · 63:01 Report a word, and more voices

Training


Data	`Hebrish/Hebrish-TTS-Voice-Recordings`, train split (1,103 read takes) and `from-stt` config (2,924 screened code-switching takes): 3,995 takes of at most 15 s, about 328 min after trimming, 24 kHz; NVIDIA NemotronLabs RE-USE (GGUF, audio.cpp), then a podcast voice chain (EQ, de-ess, 2:1 compression, -20 LUFS)
Method	Full SFT, upstream `finetuning/sft_12hz.py` with fixes (label shift, sub-talker, `text_projection` applied in training), fp32 master weights with bf16 autocast
Schedule	LR 7e-6, 3 epochs, batch 2 x grad. accumulation 4, final checkpoint
Hardware	1 x L40S (Modal)

Numbers are spelled out in the training text, so write them out in prompts too ("four hundred and twelve").

Version history

Version	Date	Data	Repo
Meorav v2.2	2026-11-09	3,995 takes, about 328 min, RE-USE-enhanced	this repo
Meorav v2.1	2026-10-22	1,103 takes, 88.2 min, DeepFilterNet-cleaned	this repo, revision `41ada105`
Meorav v2.0	2026-10-22	1,103 takes, 87.8 min	this repo, revision `600a2820`
Meorav PoC-1	2026-10-22	657 takes, 50 min	`Hebrish/TTS-V1`

Usage

import torch, soundfile as sf
from qwen_tts import Qwen3TTSModel

tts = Qwen3TTSModel.from_pretrained("Hebrish/Meorav", device_map="cuda:0", dtype=torch.bfloat16)
wavs, sr = tts.generate_custom_voice(text="Yalla, I'll ring you after the match.", speaker="daniel", language="English")
sf.write("out.wav", wavs[0], sr)

Requires qwen-tts (tested with 0.1.1). The repo is gated: log in with a token that has access.

Licence

CC BY-NC 4.0, matching the dataset. The base model is Apache 2.0.

Base model and licence notices

This model is a fine-tune of Qwen/Qwen3-TTS-12Hz-1.7B-Base, Copyright 2026 Alibaba Cloud, licensed under Apache-2.0 (text in LICENSE-Qwen-Apache-2.0). Hebrish modified the base model by fine-tuning it.

The licence stated at the top of this card covers Hebrish's contribution; the base model's licence keeps applying to the base model, and nothing here restricts what that licence allows for it. Terms of use: https://hebrish.app/terms.

Hebrish is not affiliated with or endorsed by Alibaba Cloud.