alwaysgood/sec-10k-pre-processed

该数据集包含经过预处理的SEC 10-K英文文本，提供了句子、块和段落级别的文件，适用于翻译相关任务。数据集来源于PleIAs/SEC，经过清理、过滤和去重处理，生成了四个文件：sentence_track.jsonl、chunk_track.jsonl、paragraph_track.jsonl和translation_dataset.jsonl。数据集详细描述了各文件的列信息、推荐加载方式以及去重处理的两个阶段。

This dataset contains pre-processed SEC 10-K English text with sentence/chunk/paragraph files for translation-oriented use. The dataset is derived from PleIAs/SEC, with cleaning, filtering, and deduplication applied, resulting in four files: sentence_track.jsonl, chunk_track.jsonl, paragraph_track.jsonl, and translation_dataset.jsonl. The dataset provides detailed column information, recommended loading methods, and a two-stage deduplication process.