GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
LokaalHub
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
LokaalHub
Model Tree
openai/whisper-tiny
Input Modalities
Output Modalities
Supported Functionality
A single fine-tune on ~94.4h of Dutch Common Voice takes WER from ~44.03% (base Whisper-tiny) to ~22.34% (49.3% relative drop) on a held-out, speaker- and sentence-disjoint test split.
All systems scored on the same held-out panel through one shared text normalizer
(BasicTextNormalizer). RTF = CPU compute seconds per audio second (lower is faster).
| Model | params | size (fp32) | RTF (CPU) | cv17-test | fleurs-test | mean WER |
|---|---|---|---|---|---|---|
| LokaalHub/whisper-klein-nl (ours) | 58M | 230.7 MB | 0.163 | 29.16 | 38.99 | 34.08% |
| openai/whisper-tiny | 38M | 151.0 MB | 0.091 | 46.15 | 49.14 | 47.64% |
python
from transformers import pipelineasr = pipeline("automatic-speech-recognition", model="LokaalHub/whisper-klein-nl")asr("audio.wav", generate_kwargs={"language": "nl", "task": "transcribe"})Standard Hugging Face Seq2SeqTrainer fine-tune (bf16), built and verified by the
tiny-asr-loop pipeline.
Tiny-model fine-tune on read speech (Common Voice). The internal test split is small and speaker-disjoint — see the panel table for FLEURS / out-of-domain numbers.