yasutoshi-lab/pre-train download history

yasutoshi-lab/pre-train is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 73 times (73 in the last 7 days), and 73 times in total. It ranks #192,987 among datasets by monthly downloads.

yasutoshi-lab/pre-train A deduplicated pretraining corpus with four subsets. ja and en are built from Wikipedia and FineWeb. math and code are built from SwallowMath-v2 and SwallowCode-v2. subset split docs size (parquet) ja train 117,164,071 229.2 GB ja validation 10,000

Open yasutoshi-lab/pre-train on Hugging Face

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