We want to collect a community set of datasets for various ML tasks thus making the world’s first real-life AutoML benchmark.
And we need your dataset contributions. These datasets will have publishing reveal schedule, being made publicly available with all the benchmark scores on them throughout the next year at least. This will not only benefit the community and DS field overall but you as well:
Contact @mpotapova, @natekin or firstname.lastname@example.org on your datasets and help us make the best generic ML benchmark out there in the whole field.
In case you have a separate awesome task suited for a benchmark with a unique dataset - also contact us: the more benchmarks the better.