xfxcwynlc/cosmopedia-qwen3-tokenized-packed
收藏Hugging Face2025-06-23 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/xfxcwynlc/cosmopedia-qwen3-tokenized-packed
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个特征：input_ids（int32类型），labels（int64类型），和attention_mask（int8类型）。数据集仅包含训练集，共有6700958个样本，数据集大小为356893023080字节。数据集的下载大小为107470905813字节。默认配置下，训练数据存储在data/train-*路径下。
The dataset includes three features: input_ids (int32 type), labels (int64 type), and attention_mask (int8 type). The dataset consists only of a training set with 6700958 examples, with a total size of 356893023080 bytes. The download size of the dataset is 107470905813 bytes. Under the default configuration, the training data is stored in the path data/train-*.
提供机构：
xfxcwynlc
