Word Error Rate in Wav2vec2 Fine Tuning

While fine-tuning wav2vec2 my model on training does not show the WER despite having the compute metrics Google Colab. Can someone help?