CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.speech / whisper-large-v3-turbo
From: ivrit-ai <huggingface.co/ivrit-ai>Newsgroups: alt.freeweights.speechSubject:

whisper-large-v3-turbo

Date: 2026-11-15Commit: f33172aSigned: 2DC28CC7B26A283B ✓

Automatic speech recognition model with 809M parameters. Safetensors weights, 3.24 GB in all. Published by ivrit-ai under Apache-2.0.

809M params3.24GBwhisper-large-v3-turbosafetensorspermissive
FactsParams809MSize3.24 GBFormatssafetensorsTaskautomatic speech recognitionLicenseApache-2.0From sourcepermissive
$ fw get ivrit-ai/whisper-large-v3-turbo
# manifest sha256 32118dbd1d20… · 3.24 GB · 13 files
$ fw verify ./whisper-large-v3-turbo ivrit-ai/whisper-large-v3-turbo
From sourceHosted (signed)ivrit-ai/whisper-large-v3-turboLast check: 2026-11-15 02:42 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Fine-tune of whisper-large-v3-turbo.

Works with

Run24Fine-tune7Convert8Scan2
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
3.24 GB
Weights
3.24 GB (safetensors)
Parameters
809M
Formats
safetensors
Task
automatic speech recognition
License
Apache-2.0 (permissive)
Files
13, each with a sha256
Publisher
ivrit-ai
Added
2026-11-15
Section
alt.freeweights.speech
Lineage

Original release from the publisher.

Base model: openai/whisper-large-v3-turbo

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 02:42 UTCpassed

Model card

From the publisher at commit f33172a. Shown as plain text with links; images open at the publisher.

Model Card for Model ID#

This model is a Hebrew finetune (continued training) of the OpenAI Whisper Large v3 Turbo model.

Model Details#

Model Description#
  • Developed by: ivrit-ai
  • Language(s) (NLP): Hebrew
  • License: Apache-2.0
  • Finetuned from model openai/whisper-large-v3-turbo
  • Training Date Apr 2025

Bias, Risks, and Limitations#

Language detection capability of this model has been degraded during training - it is intended for mostly-hebrew audio transcription. Language token should be explicitly set to Hebrew.

Additionally, the tanslation task was not trained and also degraded. This model would not be able to translate in any reasonable capacity.

How to Get Started with the Model#

Please follow the original model card for usage details - replacing with this model name. You can also fine other weight formats ad quantizations on the ivrit ai HF page.

We created some simple example scripts using this model and weights for other inference runtimes. Find those in the "examples" folder within the training GitHub repo.

Training Details#

Training Data#

This model was trained on the following datasets:

Training Procedure#

This model was trained in two main phases:

  • Knesset based pre-training - over all ~4700h of data - 3 epochs, ~48h run
  • Mixed post-training over all crowd-transcribe-v5 (300h), crowd-recital-whisper-training (50h) and highest-quality filtered knessets data (150h) - 2 epochs
  • Interleaving of datasets with sampling probs: (0.9, 0.025, 0.075) respectively
  • Note that crowd-transcribe-v5 has about 5x shorter samples on average thus the over-sampling.

This model is a weighted-average of the 2 lowest eval loss checkpoints (From around the end of epoch 2) from two seprate runs with the same setup. Training code can be found on the ivrit-ai Github here

Preprocessing#

The "Crowd Recital" and "Knesset" datasets contain timestamps and previous text following the Whisper expected inputs. Timestamps were used from 40% of samples from those datasets, and 50% of the previous text was used.

The "Crowd Transcribe" datasets has no timestamps or previous text and this preprocessing only included melspec feature extraction and text encoding.

Preprocessing code can be found within the training code repository.

Datasets were interleaved with 0.15:0.8:0.05 ratio (knesset:crowd-transcribe:crowd-recital).

Training Hyperparameters#
  • Training regime: bf16 mixed precision with sdpa
  • Learning Rate: 1e-5, Linear decay, 800 steps warmup for 3 epochs
  • Batch Size: 32
Training Hardware / Duration#
  • GPU Type: 8 x Nvidia A40 machine
  • Duration: ~55h run across both phases

Evaluation#

Please refer to the ivrit-ai/hebrew-transcription-leaderboard

Files

13 files · 3.24 GBWeights 3.24 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 13 files with their sha256
FileSizesha256
.gitattributes1.52 KB11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361
README.md3.98 KBf5cbe1baff06950a306a24f98658a89a795077309bab65376572a160e6189c7a
added_tokens.json34.6 KB3c51f66c4c21f9e126970078f11ae77a78c74aee8df606ee9daba86e467108e0
config.json1.23 KB22df89eb892c95148f6beaf3eb6ed2a70cf496953923e8e1667730d06efb3a69
generation_config.json2.94 KB753fd467b8fe1aa2d728f081d0031fbc68c387543e6f72bfd243640d02d93618
merges.txt494 KB2df2990a395e35e8dfbc7511e08c12d56018d8d04691e0133e5d63b21e154dc6
model.safetensors3.24 GB600f71638a2c8e674a9bead39ddc16553efc969541e0b74e74e25572b58e1a58
normalizer.json52.7 KBbf1c507dc8724ca9cf9903640dacfb69dae2f00edee4f21ceba106a7392f26dd
preprocessor_config.json357 B654cf18d3e163b948ceaf9766da56ce0b52de265d58673cf61c9376f126bd499
special_tokens_map.json2.19 KBbaea4ea09372eb4fca86b4e4346139fd73cb807d5087e9de0948e971739c3e74
tokenizer.json2.71 MB297b13372ac43916285644fb9687add3cc62ee2a1adb60da3dc25cc94c1871fd
tokenizer_config.json283 KB3c75940dfce3a294fca7041a5faff011677f1b68fa85e47511bb8cf6dccaded6
vocab.json1.04 MBe2aa043ef015641d363d8288e7c241c85e36a5c761fb303598e0710233344387
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
f33172a8c3 pinned now2026-11-15 02:42 UTC13 files · 3.24 GB

Use it

$ fw get ivrit-ai/whisper-large-v3-turbo
# manifest sha256 32118dbd1d20… · 3.24 GB · 13 files
$ fw verify ./whisper-large-v3-turbo ivrit-ai/whisper-large-v3-turbo
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/ivrit-ai/whisper-large-v3-turbo.json --dir ./whisper-large-v3-turbo $ python3 fw.py verify ./whisper-large-v3-turbo ./ivrit-ai__whisper-large-v3-turbo.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download ivrit-ai/whisper-large-v3-turbo --revision f33172a8c3c6efbc040a7200e00835257cac0447 --local-dir ./whisper-large-v3-turbo
transformersSpeech recognition with the transformers pipeline.
import os os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com" from transformers import pipeline asr = pipeline("automatic-speech-recognition", model="ivrit-ai/whisper-large-v3-turbo", revision="f33172a8c3c6efbc040a7200e00835257cac0447") print(asr("audio.wav")["text"])
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
f5cbe1baff06950a306a24f98658a89a795077309bab65376572a160e6189c7a
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: ivrit-ai <huggingface.co/ivrit-ai>Newsgroups: alt.freeweights.speechSubject: whisper-large-v3-turboDate: 2026-11-15Commit: f33172aSigned: 2DC28CC7B26A283B ✓
Upstream
ivrit-ai/whisper-large-v3-turbo
Pinned commit
f33172a8c3c6efbc040a7200e00835257cac0447
Manifest
whisper-large-v3-turbo.json · .minisig
Manifest sha256
32118dbd1d200370a051dd5c9878cfcb896798895b26d9fe3436d68770172db5
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 02:42 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/ivrit-ai/whisper-large-v3-turbo.json $ curl -O https://freeweights.allternit.com/manifests/ivrit-ai/whisper-large-v3-turbo.json.minisig $ minisign -Vm whisper-large-v3-turbo.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

openai/whisper-large-v3-turbo · this model is a finetune of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download ivrit-ai/whisper-large-v3-turbo
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. whisper-large-v3-turbo.json. Provenance. Takedown history: none.
Source
huggingface.co/ivrit-ai/whisper-large-v3-turbo

Related models

All of alt.freeweights.speech →
From: MIT <huggingface.co/MIT>Subject: ast-finetuned-audioset-10-10-0.4593
From: onnx-community <huggingface.co/onnx-community>Subject: ast-finetuned-audioset-10-10-0.4593-ONNX
From: Edge0 <huggingface.co/Edge0>Subject: Audio8-TTS-Preview-0.6b
From: Edge0 <huggingface.co/Edge0>Subject: Audio8-TTS-Realistic-Air
From: facebook <huggingface.co/facebook>Subject: audiobox-aesthetics
From: mlx-community <huggingface.co/mlx-community>Subject: chatterbox-4bit