Downloads · 30 days
40
35% of all-time downloads
whisper-small-he
amitkot/whisper-small-he
whisper-small-he is a automatic speech recognition model from amitkot. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
Hebrew fine-tuned Whisper Small for automatic speech recognition.
All-time downloads
115
Public
Parameters
242M
967 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors967 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- whisper
- Access
- Public
- Created
- Mar 10, 2026
- Updated
- Mar 10, 2026
- SHA
- 023bbedf
Base models
Datasets
Model card
From the Hugging Face model README
whisper-small-he
Hebrew fine-tuned Whisper Small for automatic speech recognition.
Training
- Base model: openai/whisper-small
- Dataset: ivrit-ai/whisper-training (~400h Hebrew)
- Method: Supervised fine-tuning with
Seq2SeqTrainer - Steps: 5,000 (streaming, effective batch size 16)
- Hardware: Apple M4 (MPS), fp32
- Best eval WER: 0.368 (on 200-sample test split, step 4000)
Usage
from transformers import WhisperProcessor, WhisperForConditionalGeneration
processor = WhisperProcessor.from_pretrained("amitkot/whisper-small-he")
model = WhisperForConditionalGeneration.from_pretrained("amitkot/whisper-small-he")
model.generation_config.language = "he"
model.generation_config.task = "transcribe"
Training pipeline
Trained using whisper-acft-pipeline:
uv run python scripts/finetune.py --config configs/hebrew_small_finetune.yaml
See also
- amitkot/whisper-small-he-acft — ACFT-optimized version of this model for short audio (FUTO Keyboard)
- amitkot/whisper-tiny-he — Smaller/faster variant
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- License
- apache-2.0
- Languages
- he
- Created
- Mar 10, 2026
- Updated
- Mar 10, 2026