Ayush-Singh/reward-bench-starcoder2-7b-yes-no

该数据集包含了与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）。数据集分为多个子集，每个子集对应不同的编程语言，包括Python、Java、C++、JavaScript、Go和Rust。每个子集包含164个示例。

The dataset includes prompts related to programming languages, chosen code snippets (chosen), models corresponding to the chosen code (chosen_model), rejected code snippets (rejected), models corresponding to the rejected code (rejected_model), dataset subsets (subset), unique identifiers (id), and probabilities associated with choosing and rejecting code (chosen_yes_prob, chosen_no_prob, rejected_yes_prob, rejected_no_prob). The dataset is divided into multiple subsets, each corresponding to a different programming language, including Python, Java, C++, JavaScript, Go, and Rust. Each subset contains 164 examples.