HF RL Explorer

FineEnvs/data-agent-harbor-test

data-agent-harbor-test: Harbor dataset on Hugging Face with 250 tasks. A held-out benchmark for data-analysis agents: 250 tasks, deliberately balanced across difficulty and leaning toward the harder end so it actually separates good agents from great ones. Each task hands your agent a real dataset…

Tasks

100 more tasks.

FineEnvs/data-agent-harbor-test on the Hugging Face Hub