GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
mahwizzzz
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
mahwizzzz
Model Tree
Input Modalities
Output Modalities
Supported Functionality
Evaluated on 1,000 test samples. Normalization: diacritics stripped, NFKC Unicode normalization applied.
| Model | WER ↓ | CER ↓ | RTF ↓ |
|---|---|---|---|
| voxbridge-37m (baseline) | 0.6076 | 0.2455 | 0.0268 |
| whispLM-600m | 0.5566 | 0.2805 | 0.1953 |
python
import torch, torchaudiofrom transformers import WhisperForConditionalGeneration, WhisperProcessor device = "cuda" if torch.cuda.is_available() else "cpu"processor = WhisperProcessor.from_pretrained("mahwizzzz/voxbridge-37m")model = WhisperForConditionalGeneration.from_pretrained("mahwizzzz/voxbridge-37m").to(device).eval() audio, sr = torchaudio.load("audio.wav")if sr != 16000: audio = torchaudio.functional.resample(audio, sr, 16000)inputs = processor(audio.squeeze().numpy(), sampling_rate=16000, return_tensors="pt").input_features.to(device) with torch.no_grad(): predicted_ids = model.generate(inputs)text = processor.batch_decode(predicted_ids, skip_special_tokens=True)[0]print(text)bibtex
@misc{mahwiz2026voxbridge, title={voxbridge-37m: Whisper-tiny Baseline for Urdu ASR}, author={Mahwiz Khalil}, year={2026}, publisher={HuggingFace}, url={https://huggingface.co/mahwizzzz/voxbridge-37m}}@misc{mahwiz2026voxbridge, title={voxbridge-37m: Whisper-tiny Baseline for Urdu ASR}, author={Mahwiz Khalil}, year={2026}, publisher={HuggingFace}, url={https://huggingface.co/mahwizzzz/voxbridge-37m}}