Added a column with the token interval of the matches for each spec sentence

Renamed source and destination to parent and child, respectively. Added support for the negative_child column.

Delete data.tar.bz2

Upload val.tar.bz2 with git-lfs

Upload odinsynth_dataset.py

Upload train.tar.bz2 with git-lfs

Upload test.tar.bz2 with git-lfs

Upload test.tar.bz2 with git-lfs

Delete merged_train_split_train.jsonl

Upload merged_train_split_train.jsonl.gz with git-lfs

Upload odinsynth_dataset.py

Upload data.tar.bz2 with git-lfs

Upload merged_train_split_train.jsonl with git-lfs

Upload odinsynth_dataset.py

initial commit