dogtooth/uf_Llama-3.1-Tulu-3-8B-SFT_2_Skywork-Reward-Gemma-2-27B-v0.2_weighted_sft

这个数据集包含了一个prompt字段和response字段（其中包括content和role子字段），以及三个分数字段：归一化分数(normalized_score)、原始分数(raw_score)和剪辑分数(clipped_score)。数据集分为训练集(train)，共有约122,270个示例，总文件大小为280,767,398字节。数据集的下载大小为104,310,251字节。

The dataset includes a prompt field and a response field (which contains subfields content and role), as well as three score fields: normalized score (normalized_score), raw score (raw_score), and clipped score (clipped_score). The dataset is split into a training set (train) with approximately 122,270 examples, with a total file size of 280,767,398 bytes. The download size of the dataset is 104,310,251 bytes.