The benchmarks section lists all benchmarks using a given dataset or any of
its variants. We use variants to distinguish between results evaluated on
slightly different versions of the same dataset. For example, ImageNet 32⨉32
and ImageNet 64⨉64 are variants of the ImageNet dataset.
The IWSLT 2019 dataset contains source, Machine Translated, reference and Post-Edited text, which can be used to quantify and evaluate Post-editing effort after automatic MT.