Run a Benchmark

Here we introduce some MTL benchmark datasets and show how to run models on them for a fair comparison.

• The NYUv2 Dataset
• The Office-31 and Office-Home Datasets