metadata
license: mit
datasets:
- mozilla-foundation/common_voice_11_0
language:
- mk
pipeline_tag: automatic-speech-recognition
tags:
- ASR
- whisper
OpenAI whisper на Македонски јазик
Loss: 0.54 WER: 54
Fine tuning dataset
Common Voice Corpus 11.0
Initial observations
Althoug using an old dataset from 2022 the results show superior performance compared to Wav2Vec 2.0