tuenguyen/s1K_tokenized_gemini

该数据集包含了解决方案(solution)、问题(question)、类型(cot_type)、来源类型(source_type)、元数据(metadata)、两种思考轨迹(gemini_thinking_trajectory和deepseek_thinking_trajectory)、两种尝试(gemini_attempt和deepseek_attempt)、消息内容(messages)和长度(length)等字段。数据集分为训练集(train)和测试集(test)，训练集包含990个示例，测试集包含10个示例。

The dataset includes fields such as solution, question, cot_type, source_type, metadata, two types of thinking trajectories (gemini_thinking_trajectory and deepseek_thinking_trajectory), two types of attempts (gemini_attempt and deepseek_attempt), message content (messages), and length. The dataset is divided into a training set (train) and a test set (test), with the training set containing 990 examples and the test set containing 10 examples.