ereverter/fineweb10B-tokenized-bert
收藏官方服务：
资源简介：
这是一个包含id、token数量和输入id序列的数据集，分为训练集和验证集。数据集规模较大，训练集有大约1472万个样本，验证集有大约15万个样本。
This dataset includes id, token count, and input id sequences, split into training and validation sets. The dataset is large in scale, with about 14.72 million samples in the training set and about 150,000 samples in the validation set.
提供机构：
ereverter
