alphacep
/

vosk-model-small-streaming-uz

Automatic Speech Recognition

Eval Results (legacy)

Model card Files Files and versions

Vosk Uzbek Speech Recognition Model

Uzbek automatic speech recognition (ASR) model for sherpa-onnx. Architecture is non-streaming Zipformer2.

Version 0.60

Training Data

This model was trained on CommonVoice, ISSAI USC and Uzvoice

Downloads last month: -; Downloads are not tracked for this model. How to track

Datasets used to train alphacep/vosk-model-small-streaming-uz

Evaluation results

Test WER on FLEURS Uzbek
test set self-reported

17.670
Test WER on issai/Uzbek_Speech_Corpus
test set self-reported

15.420