A fine-tuned Wav2Vec2 model for automatic speech recognition in Arabic, based on the XLSR-53 multilingual checkpoint.