Sign inGet started
Dataset · Hugging Face

hmkang/robomme_gaze_eval_videos

RoboMME eval videos — GR00T N1.6 gaze all-past Simulator rollout videos on the RoboMME benchmark (test split, 50 episodes/task, max 1500 steps, n_action_steps=16). Structure: <model>/<task>/<video>.mp4 + per-task simulation_results.csv (has per-episode task_instruction) and README.md. Models adaptive_ns16/ — gaze all-past, adaptive, checkpoint-120000. tl0.7_ns16/ — gaze all-past, tl0.7, checkpoint-120000. tl0.8_ns16/ — gaze all-past, target_loss 0.8, budget 2048… See the full description on the dataset page: https://huggingface.co/datasets/hmkang/robomme_gaze_eval_videos.

Author / organizationhmkang
License (as declared)mit
Downloads141 (as reported by Hugging Face)
Likes0
Created27 July 2026
Last updated at source4 August 2026
robommegr00tgazemanipulationeval
Embed this entry
<iframe src="https://silytics.si/embed/item/0b57884c-ffdd-4351-a7c5-03ad7f447bd7" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 4 days

141 now (-1 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from hmkang

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 19 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.