mlfoundations-cua-dev/

easyr1-v2-pro-apps-manually-labeled-icon-data-from-yt-4MP

This dataset was generated using the EasyR1 grounding dataset pipeline.

Generated on: 2025-09-06 00:30:03 UTC Script: push_easyr1_to_hf.py Data directory: datasets

Maximum samples: 2000 Image resize (max megapixels): 4.0 MP Minimum native image resolution: 0.0 MP Prompt format: gta1 Output format: coordinates Random seed: 42 Resampling enabled: False Icon upsampling… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-cua-dev/easyr1-v2-pro-apps-manually-labeled-icon-data-from-yt-4MP.