Wrong number of tasks displayed for AutoML All Classification

[dholzmueller]

Description The "Task Collection" tab for the "AutoML Benchmark All Classification" benchmark (id=271) says "66 tasks", while the "Tasks" tab lists 71 tasks. The corresponding paper and the Python API also specify 71 tasks.

Expected results The number 66 for the tasks (and datasets) should be corrected to 71.

[joaquinvanschoren]

Thanks, I'll look into it!