jihuny/tldr_1.4b_10k_gopt_policy_norm_penul_last_1e3_iter0

该数据集包含多个特征字段，如提示文本(prompt)、索引(prompt_idx)、选择的文本(chosen)、被拒绝的文本(rejected)及其索引，以及相关的分数、对数概率和完成原因等。数据集被划分为训练集(train)，共有10000个示例，大小为182,894,568字节。数据集还包含了默认配置，指定了训练数据的文件路径。

The dataset includes multiple feature fields such as prompt text, prompt index, chosen text, rejected text and their indices, as well as related scores, log probabilities, and completion reasons. The dataset is split into a training set (train) with 10,000 examples and a size of 182,894,568 bytes. The dataset also contains a default configuration specifying the file path for the training data.