--- dataset_info: features: - name: audio dtype: audio: sampling_rate: 16000 - name: transcription dtype: string splits: - name: train num_bytes: 2711921688.826 num_examples: 10287 download_size: 2711453905 dataset_size: 2711921688.826 configs: - config_name: default data_files: - split: train path: data/train-* license: mit task_categories: - automatic-speech-recognition - text-to-speech language: - ur pretty_name: Jalandhary size_categories: - 10K<n<100K --- Jalandhary dataset is created using whisper model for STT and TTS. Some audios are ommited due to issues while trimming them. If there are some isues in the dataset or audio not matching the text you can start a discussion or ping me to correcting it.