michaelbzhu/tulu-3-sft-tokenized-llama3.1
收藏Hugging Face2025-10-07 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/michaelbzhu/tulu-3-sft-tokenized-llama3.1
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个经过tokenization处理的训练集，包含了input_ids，attention_mask和labels三个特征，用于训练序列到序列的模型。数据集共有939,343个示例，总大小为约23GB，使用Apache-2.0许可证。
This dataset is a tokenized training set that includes input_ids, attention_mask, and labels features for training sequence-to-sequence models. The dataset contains 939,343 examples with a total size of approximately 23GB, licensed under Apache-2.0.
提供机构：
michaelbzhu
