LibrispeechMaleFemale in WebDataset Format

This dataset is the LibriSpeech dataset with the splits {train-clean-100, dev-clean, test-clean}, formatted in the WebDataset format. WebDataset files are essentially tar archives, where each example in the dataset is represented by a pair of files: a WAV audio file and a corresponding JSON metadata file. The JSON file contains the class label and other relevant information for that particular audio sample.

$ tar tvf train-clean-100-000000.tar | head-r--r--r-- bigdata/bigdata 203670 2025-01-14 04:25 5561-39621-0030.flac-r--r--r-- bigdata/bigdata    219 2025-01-14 04:25 5561-39621-0030.json-r--r--r-- bigdata/bigdata 361438 2025-01-14 04:25 5561-39621-0043.flac-r--r--r-- bigdata/bigdata    311 2025-01-14 04:25 5561-39621-0043.json-r--r--r-- bigdata/bigdata  57964 2025-01-14 04:25 5561-39621-0000.flac-r--r--r-- bigdata/bigdata     64 2025-01-14 04:25 5561-39621-0000.json-r--r--r-- bigdata/bigdata 190735 2025-01-14 04:25 5561-39621-0028.flac-r--r--r-- bigdata/bigdata    190 2025-01-14 04:25 5561-39621-0028.json-r--r--r-- bigdata/bigdata  58973 2025-01-14 04:25 5561-39621-0037.flac-r--r--r-- bigdata/bigdata     86 2025-01-14 04:25 5561-39621-0037.json

$ cat 5561-39621-0030.json | jq .{  "speaker": 5561,  "gender": "F",  "trans": "THEN THE GENERAL SENT ME BACK THE LETTER BY AN AIDE DE CAMP INFORMING ME THAT IF I WERE FOUND THE NEXT DAY WITHIN THE CIRCUMSCRIPTION OF HIS COMMAND HE WOULD HAVE ME ARRESTED"}