--- license: cc-by-4.0 dataset_info: - config_name: assamese features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 443707022.795 num_examples: 3873 - name: train num_bytes: 45095490490.71 num_examples: 367970 download_size: 83768335117 dataset_size: 45539197513.505 - config_name: bengali features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 374970443.066 num_examples: 3082 - name: train num_bytes: 39982022188.005 num_examples: 256773 download_size: 65765282014 dataset_size: 40356992631.071 - config_name: bodo features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 299874784.5 num_examples: 3156 - name: train num_bytes: 29260849420.794 num_examples: 282922 download_size: 58106016446 dataset_size: 29560724205.294 - config_name: dogri features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 250247735.494 num_examples: 1659 - name: train num_bytes: 18381323089.355 num_examples: 136521 download_size: 37388588697 dataset_size: 18631570824.849 - config_name: gujarati features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 196975098.092 num_examples: 1468 - name: train num_bytes: 12247067300.56 num_examples: 93870 download_size: 26128896594 dataset_size: 12444042398.651999 - config_name: hindi features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 339997010.24 num_examples: 4180 - name: train num_bytes: 36711453158.328 num_examples: 333256 download_size: 70756999351 dataset_size: 37051450168.568 - config_name: kannada features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 326421454.786 num_examples: 2743 - name: train num_bytes: 29532680269.928 num_examples: 185716 download_size: 43860433686 dataset_size: 29859101724.714 - config_name: kashmiri features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 260980674.406 num_examples: 2301 - name: train num_bytes: 19894306917.96 num_examples: 160120 download_size: 36388001887 dataset_size: 20155287592.365997 - config_name: konkani features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 198640518.718 num_examples: 1694 - name: train num_bytes: 14369324982.814 num_examples: 127534 download_size: 29583030035 dataset_size: 14567965501.532 - config_name: maithili features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 316460820.95 num_examples: 2525 - name: train num_bytes: 33296659626.454 num_examples: 229667 download_size: 59277630201 dataset_size: 33613120447.404 - config_name: malayalam features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 270363094.712 num_examples: 3688 - name: train num_bytes: 30050461452.018 num_examples: 263487 download_size: 64548929998 dataset_size: 30320824546.730003 - config_name: manipuri features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 196330652.504 num_examples: 1599 - name: train num_bytes: 14311739665.408 num_examples: 112864 download_size: 26856055652 dataset_size: 14508070317.912 - config_name: marathi features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 337440763.08 num_examples: 2792 - name: train num_bytes: 29481182902.72 num_examples: 225035 download_size: 55203303433 dataset_size: 29818623665.800003 - config_name: nepali features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 263124099.442 num_examples: 2589 - name: train num_bytes: 41350688945.542 num_examples: 226166 download_size: 51549356838 dataset_size: 41613813044.984 - config_name: odia features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 252524784.605 num_examples: 3463 - name: train num_bytes: 31173625919.422 num_examples: 263059 download_size: 58746623937 dataset_size: 31426150704.027 - config_name: punjabi features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 317944755.456 num_examples: 2489 - name: train num_bytes: 30272563762.54 num_examples: 205884 download_size: 56569098622 dataset_size: 30590508517.996002 - config_name: sanskrit features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 207896091.008 num_examples: 1896 - name: train num_bytes: 17666288958.08 num_examples: 138680 download_size: 33930409107 dataset_size: 17874185049.088 - config_name: santali features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 241512352.824 num_examples: 2592 - name: train num_bytes: 21923662503.076 num_examples: 223108 download_size: 45446386060 dataset_size: 22165174855.9 - config_name: sindhi features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 272236826.91999996 num_examples: 2320 - name: train num_bytes: 17449969299.059 num_examples: 152341 download_size: 39766452875 dataset_size: 17722206125.978996 - config_name: tamil features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 457582273.726 num_examples: 4426 - name: train num_bytes: 44686288575.74 num_examples: 353805 download_size: 85187554733 dataset_size: 45143870849.465996 - config_name: telugu features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 210059233.673 num_examples: 2629 - name: train num_bytes: 22740057499.686 num_examples: 219702 download_size: 55269638271 dataset_size: 22950116733.359 - config_name: urdu features: - name: audio_filepath dtype: audio - name: text dtype: string - name: duration dtype: float64 - name: lang dtype: string - name: samples dtype: int64 - name: verbatim dtype: string - name: normalized dtype: string - name: speaker_id dtype: string - name: scenario dtype: string - name: task_name dtype: string - name: gender dtype: string - name: age_group dtype: string - name: job_type dtype: string - name: qualification dtype: string - name: area dtype: string - name: district dtype: string - name: state dtype: string - name: occupation dtype: string - name: verification_report dtype: string - name: unsanitized_verbatim dtype: string - name: unsanitized_normalized dtype: string splits: - name: valid num_bytes: 284618007.86 num_examples: 2588 - name: train num_bytes: 27766733703.64 num_examples: 186535 download_size: 49027126179 dataset_size: 28051351711.5 configs: - config_name: assamese data_files: - split: valid path: assamese/valid-* - split: train path: assamese/train-* - config_name: bengali data_files: - split: valid path: bengali/valid-* - split: train path: bengali/train-* - config_name: bodo data_files: - split: valid path: bodo/valid-* - split: train path: bodo/train-* - config_name: dogri data_files: - split: valid path: dogri/valid-* - split: train path: dogri/train-* - config_name: gujarati data_files: - split: valid path: gujarati/valid-* - split: train path: gujarati/train-* - config_name: hindi data_files: - split: valid path: hindi/valid-* - split: train path: hindi/train-* - config_name: kannada data_files: - split: valid path: kannada/valid-* - split: train path: kannada/train-* - config_name: kashmiri data_files: - split: valid path: kashmiri/valid-* - split: train path: kashmiri/train-* - config_name: konkani data_files: - split: valid path: konkani/valid-* - split: train path: konkani/train-* - config_name: maithili data_files: - split: valid path: maithili/valid-* - split: train path: maithili/train-* - config_name: malayalam data_files: - split: valid path: malayalam/valid-* - split: train path: malayalam/train-* - config_name: manipuri data_files: - split: valid path: manipuri/valid-* - split: train path: manipuri/train-* - config_name: marathi data_files: - split: valid path: marathi/valid-* - split: train path: marathi/train-* - config_name: nepali data_files: - split: valid path: nepali/valid-* - split: train path: nepali/train-* - config_name: odia data_files: - split: valid path: odia/valid-* - split: train path: odia/train-* - config_name: punjabi data_files: - split: valid path: punjabi/valid-* - split: train path: punjabi/train-* - config_name: sanskrit data_files: - split: valid path: sanskrit/valid-* - split: train path: sanskrit/train-* - config_name: santali data_files: - split: valid path: santali/valid-* - split: train path: santali/train-* - config_name: sindhi data_files: - split: valid path: sindhi/valid-* - split: train path: sindhi/train-* - config_name: tamil data_files: - split: valid path: tamil/valid-* - split: train path: tamil/train-* - config_name: telugu data_files: - split: valid path: telugu/valid-* - split: train path: telugu/train-* - config_name: urdu data_files: - split: valid path: urdu/valid-* - split: train path: urdu/train-* --- # IndicVoices: Towards building an Inclusive Multilingual Speech Dataset for Indian Languages
GitHub ArXiv CC BY 4.0
## Dataset Description - **Homepage:** [IndicVoices](https://huggingface.co/datasets/ai4bharat/IndicVoices) - **Repository:** [Github](https://github.com/AI4Bharat/IndicVoices) - **Paper:** [IndicVoices: Towards building an Inclusive Multilingual Speech Dataset for Indian Languages ](https://arxiv.org/abs/2403.01926) ## Overview INDICVOICES is a dataset of natural and spontaneous speech containing a total of 19550 hours of read (8%), extempore (76%) and conversational (15%) audio from 29K speakers covering 400+ Indian districts and 22 languages. Of these 19550 hours, 9200 hours have already been transcribed. Through this paper, we share our journey of capturing the cultural, linguistic and demographic diversity of India to create a one-of-its-kind inclusive and representative dataset. More specifically, we share an open-source blueprint for data collection at scale comprising of standardised protocols, centralised tools, a repository of engaging questions, prompts and conversation scenarios spanning multiple domains and topics of interest, quality control mechanisms, comprehensive transcription guidelines and transcription tools. We hope that this open source blueprint will serve as a comprehensive starter kit for data collection efforts in other multilingual regions of the world. Using INDICVOICES, we build IndicASR, the first ASR model to support all the 22 languages listed in the 8th schedule of the Constitution of India. This work is funded by Bhashini, MeitY and Nilekani Philanthropies ## Usage The [datasets](https://huggingface.co/docs/datasets) library enables you to load and preprocess the dataset directly in Python. Ensure you have an active HuggingFace access token (obtainable from [Hugging Face settings](https://huggingface.co/settings/tokens)) before proceeding. To load the dataset, run: ```python from datasets import load_dataset # Load the dataset from the HuggingFace Hub dataset = load_dataset("ai4bharat/IndicVoices","assamese",split="valid") # Check the dataset structure print(dataset) ``` You can also stream the dataset by enabling the `streaming=True` flag: ```python from datasets import load_dataset dataset = load_dataset("ai4bharat/IndicVoices","assamese",split="valid", streaming=True) print(next(iter(dataset))) ``` ## Citation If you use IndicVoices in your work, please cite us: ```bibtex @inproceedings{DBLP:conf/acl/JavedNGJBMSAFPR24, author = {Tahir Javed and Janki Nawale and Eldho Ittan George and Sakshi Joshi and Kaushal Santosh Bhogale and Deovrat Mehendale and Ishvinder Virender Sethi and Aparna Ananthanarayanan and Hafsah Faquih and Pratiti Palit and Sneha Ravishankar and Saranya Sukumaran and Tripura Panchagnula and Sunjay Murali and Kunal Sharad Gandhi and Ambujavalli R and Manickam K. M and C. Venkata Vaijayanthi and Krishnan Srinivasa Raghavan Karunganni and Pratyush Kumar and Mitesh M. Khapra}, title = {IndicVoices: Towards building an Inclusive Multilingual Speech Dataset for Indian Languages}, booktitle = {{ACL} (Findings)}, pages = {10740--10782}, publisher = {Association for Computational Linguistics}, year = {2024} } ``` ## License This dataset is released under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). ## Contact For any questions or feedback, please contact: - Tahir Javed (tahir@cse.iitm.ac.in) - Mitesh M. Khapra (miteshk@cse.iitm.ac.in)