cminst/stringatlas-multilingual-20b download history

cminst/stringatlas-multilingual-20b is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 115 times (10 in the last 7 days), and 115 times in total. It ranks #135,844 among datasets by monthly downloads.

String Atlas Multilingual 20B A public six-language pretraining corpus for String Atlas experiments. The corpus is balanced by tokens under multilingual_general_bpe32. Each language contributes approximately 3.333B anchor tokens. Files languages/<language>/documents.js

Open cminst/stringatlas-multilingual-20b on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.