-
Fix Dataset Viewer schema with Arrow JSON provenance fields
-
Fix common split schema for Dataset Viewer
-
Release repaired Active-SWE v0.4 data
-
Update release Parquet with normalized Django hard GT identifiers
Synchronize the published main and extended Parquet files after GT identifier normalization and validation. -
Use Active-SWE and Active-SWE-Extend viewer file names
-
Remove ambiguous viewer path data/extend/train.parquet
-
Remove ambiguous viewer path data/main/train.parquet
-
Use named Main and Extend viewer files: data/extend-00000-of-00001.parquet
-
Use named Main and Extend viewer files: data/main-00000-of-00001.parquet
-
Add separate HF viewer split: data/extend/train.parquet
-
Add separate HF viewer split: data/main/train.parquet