Ayush-Singh/reward-bench-Qwen2.5-Coder-7B-yes-no

该数据集包含了编程语言相关的提示(prompt)、选择的代码(chosen)、选择代码对应的模型(chosen_model)、被拒绝的代码(rejected)、被拒绝代码对应的模型(rejected_model)、数据子集(subset)、唯一标识符(id)以及与选择和拒绝代码相关的概率(chosen_yes_prob, chosen_no_prob, rejected_yes_prob, rejected_no_prob)。数据集被分割为多个子集，每个子集对应不同的编程语言，例如Python、Java、C++、JavaScript、Go和Rust，每个子集包含相同数量的示例。数据集的总大小超过了1MB。

The dataset includes programming language-related prompts, chosen code snippets, models corresponding to the chosen code, rejected code snippets, models corresponding to the rejected code, dataset subsets, unique identifiers, and probabilities associated with choosing and rejecting code snippets. The dataset is split into multiple subsets, each corresponding to a different programming language, such as Python, Java, C++, JavaScript, Go, and Rust, with each subset containing the same number of examples. The total size of the dataset is over 1MB.