metadata

language:
  - sr
  - hr
  - bs
tags:
  - audio
  - asr
  - automatic-speech-recognition
license: mit
datasets:
  - juzne-vesti-srpski
  - juznevesti-sr
  - google/fleurs
  - mozilla-foundation/common_voice_16_1
  - espnet/yodas

Model

Fine-tune of openAI's whisper-medium on the multiple datasets.

datasets:

juzne-vesti-srpski
juznevesti-sr
google/fleurs
mozilla-foundation/common_voice_16_1
espnet/yodas

It achieves the following results on the evaluation set:

Loss: 0.1358
Wer Ortho: 0.1814
Wer: 0.0901