leonli66/nemotron-sft-pretrain-compression
收藏Hugging Face2025-10-24 更新2025-11-15 收录
下载链接:
https://hf-mirror.com/datasets/leonli66/nemotron-sft-pretrain-compression
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含预训练和微调的数据,每个样本由聊天格式的prompt和对应的target字段组成。数据集分为pretrain-sft和pretrain-sft-masked两种配置,均包含训练集数据,数据文件采用.parquet格式。
The dataset consists of pretraining and fine-tuning data, with each sample including a chat-format prompt and a corresponding target field. The dataset is divided into two configurations: pretrain-sft and pretrain-sft-masked, both containing training set data in .parquet format.
提供机构:
leonli66



