jihuny/tldr_1.4b_10k_gopt_policy_norm_penul_mean_1e3_iter0

该数据集包含了一系列文本相关的特征，如提示文本(prompt)、选择的文本(chosen)、被拒绝的文本(rejected)等。每个文本都有相应的索引和分数。此外，还包含了文本的嵌入向量差异等信息。数据集分为训练集，共有10000个示例，整个数据集的大小为180,615,977字节。

The dataset consists of a set of text-related features such as prompt text, chosen text, and rejected text. Each text has its corresponding index and score. In addition, it includes information such as the difference in embedding vectors of the text. The dataset is split into a training set with a total of 10,000 examples, and the entire dataset size is 180,615,977 bytes.