Add task_categories metadata to dataset card
#2
by nielsr HF Staff - opened
This PR adds the missing task_categories metadata field to the dataset card. Evo-Bench is a benchmark for evaluating models' intrinsic harness-evolving capabilities, which does not correspond to a standard task category, so it is set to other. The existing dataset card already contains thorough documentation, links, and usage instructions, so no content changes are required.