A fine-tuned Wav2Vec2 model for Japanese automatic speech recognition using the XLSR-53 multilingual pretrained checkpoint.