Closed mcollardanuy closed 2 years ago
HIPE data only has dev and test. Split HIPE dev set into train and dev.
Done in https://github.com/Living-with-machines/toponym-resolution/pull/112/commits/c8555d7929978c2cce9ca27152482933b32f5962.
Data looks like this now:
* Size of the toponym resolution train set: 39 * Size of the toponym resolution dev set: 40 * Size of the toponym resolution test set: 46
HIPE data only has dev and test. Split HIPE dev set into train and dev.