Full dataset release?
#2
by antoniaaa - opened
Hi, in your paper you mention that the benchmark consists of over 13k samples, while the data released here has only 500. May I ask if you plan on releasing the full set of data (including train, validation, and test)? It would aid reproducing the results, and further evaluation of any new methods.