GSMA/oran
收藏资源简介:
该数据集镜像了O-RAN ALLIANCE发布的规范和技术报告。每个源文件都保存两次:original/目录包含O-RAN发布的原始文档(PDF和Word格式),而marked/目录包含转换为Markdown格式的相同文档,便于搜索和检索,每个文档对应一个raw.md文件,相关图像提取在旁边。两个目录树共享相同的路径结构,因此original/中的文件在marked/下具有匹配的解析副本;两者都按O-RAN ALLIANCE的工作组(WG1至WG11,以及焦点组SFG、SuFG和TIFG)分组。manifest.jsonl文件为每个文件提供一行记录(包括路径、SHA-256和解析状态),STATUS.md文件报告语料库的解析覆盖率。数据集包含多个配置,对应不同的工作组,数据文件以Parquet格式存储。
This dataset mirrors the O-RAN ALLIANCEs published specifications and technical reports. Every source is kept twice: `original/` is the document as O-RAN released it (PDF and Word), and `marked/` is that same document converted to Markdown for search and retrieval, one `raw.md` per document with any figures extracted beside it. The two trees share identical paths, so a file in `original/` has a parsed twin at the matching path under `marked/`; both are grouped by O-RAN ALLIANCE working group (WG1 through WG11, plus the focus groups SFG, SuFG, and TIFG). `manifest.jsonl` carries one row per file (path, SHA-256, parse status), and `STATUS.md` reports how much of the corpus has been parsed.




