Dataset · Hugging Face

ai4bharat/sangraha

Name: sangraha
Creator: ai4bharat
License: cc-by-4.0
Keywords: task_categories:text-generation, language:as, language:bn, language:gu, language:en, language:hi, language:kn, language:ks, language:ml, language:mr, language:ne, language:or, language:pa, language:sa, language:sd

Sangraha Sangraha is the largest high-quality, cleaned Indic language pretraining data containing 251B tokens summed up over 22 languages, extracted from curated sources, existing multilingual corpora and large scale translations. More information: For detailed information on the curation and cleaning process of Sangraha, please checkout our paper on Arxiv; Check out the scraping and cleaning pipelines used to curate Sangraha on GitHub; Getting Started For… See the full description on the dataset page: https://huggingface.co/datasets/ai4bharat/sangraha.

View on Hugging FaceReport an issue

Author / organization	ai4bharat
License (as declared)	cc-by-4.0
Downloads	40.4K (as reported by Hugging Face)
Likes	85
Created	5 March 2024
Last updated at source	5 March 2025

language-modelingcasual-lmllm

Embed this entry

<iframe src="https://silytics.si/embed/item/9fea38c3-ab16-4230-ba82-350b59b077e0" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

More from ai4bharat

ai4bharat/IndicVoicesDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 15 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.