rmf
Datasets
All datasets matching “rmf”rmfs-wm-v3-jenkins5b
RMFS World Model v3 — jenkins5-b1
Egocentric driving data for a single robot (the ego) in a simulated Robotic
Mobile Fulfillment System (RMFS) warehouse, collected from
RAWSim-O via a custom Gym server.
Intended for training World Models-style
V (VAE) + M (MDN-RNN) + C (controller) stacks where the controller is trained
entirely inside the learned dream.
At a glance
Episodes
120
Frames
512,734
Hops (macro-steps)
28,800
Cameras
4 x 96x96 RGB… See the full description on the dataset page: https://huggingface.co/datasets/luckysantoso/rmfs-wm-v3-jenkins5b.RM_FMT_FULL_TRAJ_118RM_FMT_FULL_2metatree_rmftsa_sleepdata
Dataset Card for "metatree_rmftsa_sleepdata"
More Information needed
rmf-ato-core
RMF/ATO Core Corpus
What this is
A curated corpus of current-revision NIST Risk Management Framework and authorization
publications, plus a second tier of AI governance documents. Every row is source text — one
control, one assessment objective, one RMF task, one AI RMF subcategory, one SSDF practice, one
document section — carrying the identifiers a practitioner actually cites. It is built for
retrieval and fine-tuning around authorization workflows: control… See the full description on the dataset page: https://huggingface.co/datasets/ezesecops/rmf-ato-core.RMFS_World_Models_Dataset
RMFS World Models — Speed Governor Corpora
Egocentric multi-camera trajectory data from a Robotic Mobile Fulfillment System (RMFS)
warehouse simulator, collected to train a World Model (VAE → MDN-RNN → CMA-ES controller,
after Ha & Schmidhuber 2018) that acts as a per-robot speed governor.
Research project at NTUST, Prof. Chou's CITI lab, extending the
RAWSim-O discrete-event simulator.
The controller does not choose where a robot goes. A* owns routing. The controller chooses… See the full description on the dataset page: https://huggingface.co/datasets/EthanGlueck/RMFS_World_Models_Dataset.