Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The information about the size of the dataset is not coherent.
Error code:   UnexpectedError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

action
list
env_state
list
obs
dict
robot_bases
array 3D
[ -0.2105138897895813, 0.013805333524942398, 0.21899999678134918, 0.014644693583250046 ]
[0.03107117861509323,0.019533496350049973,0.014000000432133675,0.9678861498832703,0.0,0.0,0.25138896(...TRUNCATED)
{"ee_pos":[[[0.9652183651924133,-4.809095116797835e-6,0.26144495606422424,-0.21950750052928925],[-0.(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.19484089314937592, -0.0012464489554986358, 0.21637973189353943, 0.0022896481677889824 ]
[0.029402317479252815,0.018029330298304558,0.01221097819507122,0.9693279266357422,-0.000605919572990(...TRUNCATED)
{"ee_pos":[[[0.9643630385398865,0.001067421049810946,0.2645803689956665,-0.2187025398015976],[-0.000(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.18355438113212585, -0.018878748640418053, 0.21646568179130554, -0.01122897770255804 ]
[0.027425697073340416,0.015535361133515835,0.011560285463929176,0.9726422429084778,-0.00004288096170(...TRUNCATED)
{"ee_pos":[[[0.9643841981887817,0.0127137191593647,0.2641996741294861,-0.215414896607399],[-0.011197(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.16851110756397247, -0.032018426805734634, 0.21416586637496948, -0.023980306461453438 ]
[0.027265110984444618,0.013266164809465408,0.01139095239341259,0.9759254455566406,-0.002568902680650(...TRUNCATED)
{"ee_pos":[[[0.9646958708763123,0.017910191789269447,0.2627567946910858,-0.21163614094257355],[-0.01(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.1556592732667923, -0.04433421418070793, 0.21280601620674133, -0.033768437802791595 ]
[0.02584247477352619,0.011629415675997734,0.011834206059575081,0.9795656800270081,-0.004130544606596(...TRUNCATED)
{"ee_pos":[[[0.9646985530853271,0.015420692972838879,0.2629048228263855,-0.20803822576999664],[-0.00(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.14323972165584564, -0.05771418660879135, 0.21314729750156403, -0.043111927807331085 ]
[0.025061415508389473,0.011846386827528477,0.011686527170240879,0.9822339415550232,-0.00386211602017(...TRUNCATED)
{"ee_pos":[[[0.9648337364196777,0.011195501312613487,0.2626223564147949,-0.20451752841472626],[-0.00(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.1321815550327301, -0.06675812602043152, 0.21167001128196716, -0.049328844994306564 ]
[0.023028647527098656,0.010786744765937328,0.011138885281980038,0.9854555726051331,-0.00052706664428(...TRUNCATED)
{"ee_pos":[[[0.9649743437767029,0.009672194719314575,0.2621658444404602,-0.2010379284620285],[-0.000(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.11961960047483444, -0.07869923859834671, 0.2103099226951599, -0.05796598270535469 ]
[0.022391662001609802,0.010231995023787022,0.011376604437828064,0.986788809299469,-0.000216822416405(...TRUNCATED)
{"ee_pos":[[[0.9650951027870178,0.011742105707526207,0.26163625717163086,-0.19754056632518768],[-0.0(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.10701749473810196, -0.09103085100650787, 0.2066589891910553, -0.06450516730546951 ]
[0.02150862291455269,0.010256496258080006,0.011235360987484455,0.9875415563583374,-0.000941439007874(...TRUNCATED)
{"ee_pos":[[[0.9651210308074951,0.012901188805699348,0.26148608326911926,-0.1940474659204483],[-0.00(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
[ -0.09971360862255096, -0.09700017422437668, 0.20933960378170013, -0.07107145339250565 ]
[0.021465081721544266,0.010401626117527485,0.011103084310889244,0.9877974987030029,-0.00009132721606(...TRUNCATED)
{"ee_pos":[[[0.9651634693145752,0.011516873724758625,0.26139411330223083,-0.19057156145572662],[-0.0(...TRUNCATED)
[[[1.0,0.0,0.0,-0.4690000116825104],[0.0,1.0,0.0,-0.01899999938905239],[0.0,0.0,1.0,0.01999999955296(...TRUNCATED)
End of preview.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Procedural Letters Sim Dataset (RL Coverage)

MuJoCo/ALOHA simulation rollouts of 26 procedural letter shapes (A-Z) collected with an RL-coverage policy. Intended for world-model training and letter generalization experiments.

Dataset structure

all_letters_50/
├── train/
│   └── episode_*.hdf5      (1170 episodes)
├── val/
│   └── episode_*.hdf5      (130 episodes)
├── metadata.json
└── README.md
  • 50 episodes per letter (26 letters)
  • Train: 45 episodes/letter (90/10 split)
  • Val: 5 episodes/letter
  • 600 frames per episode
  • Image observations: top_pov RGB camera, 128x128
  • Actions: 4D end-effector positions (left_ee_x, left_ee_y, right_ee_x, right_ee_y)
  • Motion: rl_coverage (scripted primitives with online high-level RL)

HDF5 episode structure

Each episode HDF5 contains:

Key Shape Description
action (600, 4) Joint actions
obs/images/top_pov (600, 128, 128, 3) RGB camera view
obs/ee_pos (600, 2, 4, 4) End-effector poses
env_state (600, ...) Full MuJoCo state (position, orientation)

Usage with interactive-world-sim

This dataset is in the flat layout expected by SimAlohaDataset:

python main.py \
  algorithm=latent_world_model \
  experiment=exp_latent_dyn \
  dataset=sim_aloha_dataset \
  dataset.dataset_dir=<path>/all_letters_50 \
  dataset.obs_keys=[top_pov] \
  dataset.horizon=1 \
  dataset.val_horizon=1 \
  algorithm.action_dim=4 \
  algorithm.training_stage=1

Source

Collected with scripts/data_collection/sim_aloha_dataset_collection_scripted.py from: https://github.com/WangYixuan12/interactive_world_sim

Downloads last month
482