BanglaLLM/s1k-Bangla_tokenized-qwq32b

该数据集包含了解题方案(solution)、问题(question)、问题类型(cot_type)、来源类型(source_type)、元数据(metadata)等字段，以及它们的翻译版本。此外，还包括了gemini和deepseek的思考轨迹和尝试次数等字段。数据集的训练集(train)包含1000个示例，总大小为约200MB。

The dataset includes fields such as solution, question, cot_type, source_type, metadata, and their translated versions. It also contains fields for gemini and deepseek thinking trajectories and attempt counts. The training set (train) of the dataset contains 1000 examples, with a total size of approximately 200MB.