Harbor task-binary dataset (1,000 tasks) converted from nvidia/Nemotron-RL-InverseIFEval-v1 (part of nvidia/Nemotron-Post-Training-v3). Columns path (str) + task_binary (gzip tar). Converted with the OpenThoughts-Agent data.nemotron_gym framework. Grading: Multi-criterion LLM judge (all YES). Needs OPENAI_API_KEY at trial.

What changed vs the prior version

This version fixes the answer-delivery contract for terminal agents. The… See the full description on the dataset page: https://huggingface.co/datasets/laion/nemotron-gym-inverse-ifeval-v2.