Downloads · 30 days
16
21% of all-time downloads
whisper-yoad-tiny-he-acft
amitkot/whisper-yoad-tiny-he-acft
whisper-yoad-tiny-he-acft is a automatic speech recognition model from amitkot. Use it when you need speech turned into text. It is set up for transformers.
ACFT (Audio Context Fine-Tuning) applied to yoad/whisper-tiny for Hebrew speech recognition.
All-time downloads
77
Public
Parameters
37.8M
302 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin151 MB · 49%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- Model type
- whisper
- Access
- Public
- Created
- Mar 28, 2026
- Updated
- Mar 28, 2026
- SHA
- de907bfb
Base models
Model card
From the Hugging Face model README
whisper-yoad-tiny-he-acft
ACFT (Audio Context Fine-Tuning) applied to yoad/whisper-tiny for Hebrew speech recognition.
ACFT aligns partial-context encoder representations with full-context ones, improving short-utterance inference (e.g., keyboard dictation).
Evaluation
WER on ivrit-ai/whisper-training test split (2000 samples, no normalization):
| Model | WER |
|---|---|
| yoad/whisper-tiny (base) | 0.3978 |
| yoad/whisper-tiny + ACFT (this model) | 0.3993 |
Training
- Method: ACFT (encoder MSE alignment)
- Dataset: google/fleurs he_il
- Epochs: 8
- Device: Apple MPS (M4 Pro)
Usage
from transformers import WhisperForConditionalGeneration, WhisperProcessor
model = WhisperForConditionalGeneration.from_pretrained("amitkot/whisper-yoad-tiny-he-acft")
processor = WhisperProcessor.from_pretrained("amitkot/whisper-yoad-tiny-he-acft")
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- Languages
- he
- Created
- Mar 28, 2026
- Updated
- Mar 28, 2026