Downloads · 30 days
301
96% of all-time downloads
svale-110M
3dio-ai/svale-110M
svale-110M is a automatic speech recognition model from 3dio-ai. Use it when you need speech turned into text. It is set up for nemo. The card lists the license as other.
Danish speech recognition for CPU. nvidia/parakeet-rnnt-110m-da-dk fine-tuned on 2,850 h of public Danish speech (CoRal-v3, FTSpeech, Common Voice, FLEURS, YODAS; train splits only). Lowercase, no punctuation. 16x rea…
All-time downloads
313
Public
Repo size
452 MB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.nemo452 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- nemo
- License
- other
- Access
- Public
- Created
- Sep 9, 2026
- Updated
- Sep 9, 2026
- SHA
- cf6b5d33
Base models
Model card
From the Hugging Face model README
svale-110M
Danish speech recognition for CPU. nvidia/parakeet-rnnt-110m-da-dk fine-tuned on 2,850 h of
public Danish speech (CoRal-v3, FTSpeech, Common Voice, FLEURS, YODAS; train splits only).
Lowercase, no punctuation. 16x realtime on 8 CPU threads.
WER, Danish ASR leaderboard normaliser:
| CoRal conversation | CoRal read-aloud | FTSpeech | Common Voice | FLEURS | mean |
|---|---|---|---|---|---|
| 26.74 | 11.52 | 7.33 | 9.52 | 9.63 | 12.95 |
Use
from huggingface_hub import hf_hub_download
import nemo.collections.asr as nemo_asr
model = nemo_asr.models.ASRModel.restore_from(hf_hub_download("3dio-ai/svale-110M", "svale-110M.nemo"))
print(model.transcribe(["audio.wav"]))
NeMo 2.1+, 16 kHz mono.
Licence
NVIDIA Open Model License. CoRal OpenRAIL-D use restrictions apply: no speech synthesis, no biometric identification.
- Task
- Automatic Speech Recognition
- Library
- nemo
- License
- other
- Languages
- da
- Created
- Sep 9, 2026
- Updated
- Sep 9, 2026