Pretrained and finetuned wav2vec2.0 Base model on Flemish data.

Pre-training data: CGN (all components, VL) + VrijeWesten (2000h) + VRT (500h)
Finetuning data: CGN (all components, VL)