datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
javascriptjavascript-mediumreact_reposcapx_dream_demobimanual-piper-dataset-threecam-test-10This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 1,
"total_frames": 2817,
"total_tasks": 1,
"total_videos": 3,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test-10.bimanual-piper-dataset-threecam-test-11This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 3,
"total_frames": 10691,
"total_tasks": 1,
"total_videos": 9,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:3"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test-11.MBA-Bench
MBA-Bench
Official benchmark dataset for MBA: Multimodal Benchmark and Agents for Real-World Business Ideation.
Paper: https://arxiv.org/abs/2608.11616
Project Page: https://hchoi256.github.io/projects/mba/
Code: https://github.com/hchoi256/mba
Models: https://huggingface.co/hchoi256/mba
Dataset Description
MBA-Bench is a multimodal benchmark for evaluating real-world business ideation capabilities.
The dataset contains multimodal benchmark samples together with… See the full description on the dataset page: https://huggingface.co/datasets/hchoi256/MBA-Bench.javascript-smallbimanual-piper-dataset-threecam-testThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 1,
"total_frames": 845,
"total_tasks": 1,
"total_videos": 3,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test.hc-hvlm-visiblerobotwin_dream_success_videos
robotwin_dream_success_videos
DREAMED (world-model-imagined) videos judged SUCCESS BY ROBOMETER (not by the sim), with actions.
Layout mirrors robotwin_real_success_videos: <task>/<scene>/set1/episode_N/{dream.mp4, side_by_side.mp4, actions.npz}
(set2, set3, ... later). Episode numbers keep source indices; only robometer keep==1 episodes included.
dream.mp4 = the WM-imagined rollout; side_by_side.mp4 = GT | dream | skeleton composite; no standalone real video.
set1 (2026-09-01)… See the full description on the dataset page: https://huggingface.co/datasets/hchen657Naza/robotwin_dream_success_videos.novel_task_dream
novel_task_dream — dream replays of NOVEL tasks through the RoboTwin WM (capx10 askel hist; YAM runs used step4000, RoboTwin runs step5000)
Videos are GT | dream | action-skeleton, named rankpsnr_. Ordered by PSNR mean.
rank
task
robot
scene
psnr mean
first
last
1
click_alarmclock
robotwin
noisy
27.4
32.6
24.4
2
stack_bowls_two
robotwin
noisy
25.6
34.0
22.7
3
place_can_basket
robotwin
noisy
23.9
34.5
20.3
4
click_alarmclock
robotwin
clean
23.7
28.9
20.4
5… See the full description on the dataset page: https://huggingface.co/datasets/hchen657Naza/novel_task_dream.hc-hvlm-graderMSPL
MSPL Dataset
Dataset for MSPL: Multi-Step Pseudo-Labeling for Open-Vocabulary Object Detection.
Paper: https://arxiv.org/abs/2510.14792
Code: https://github.com/hchoi256/mspl
robotwin_selected_dream_success_videos
Selected RoboTwin Dream Success Demos — 8 tasks x 100 episodes
The exact 800-episode subset used to train the pi0.5 policy
pi05_checkpoints/robotwin_dream8_pi05_full (full finetune, 50k steps, delta joint actions).
Curated from hchen657Naza/robotwin_dream_success_videos, which holds more episodes per task
than were wanted (up to 296 for shake_bottle); this repo pins the subset that was actually used
so the training run is reproducible.
Selection rule
20 clean + 80… See the full description on the dataset page: https://huggingface.co/datasets/hchen657Naza/robotwin_selected_dream_success_videos.vietnamese-wiki
Dataset Card for "vietnamese-wiki"
More Information needed
SAM_Quantizationgraphcontainer-graphs
GraphContainer Graph Artifacts
Overview
This repository contains preconstructed graph artifacts released with GraphContainer: A Unified Platform for Comparing and Debugging Graph RAG Methods.
Links:
Paper: https://arxiv.org/abs/2607.19362
Hugging Face paper page: https://huggingface.co/papers/2607.19362
Code: https://github.com/asmath472/GraphContainer
YouTube demo: https://youtu.be/O02eNJLwkU0
The Hugging Face datasets interface loads the artifact catalog in… See the full description on the dataset page: https://huggingface.co/datasets/hchaejeong/graphcontainer-graphs.griffin-test10This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 1,
"total_frames": 983,
"total_tasks": 1,
"total_videos": 6,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/griffin-test10.griffin-test7This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 1,
"total_frames": 688,
"total_tasks": 1,
"total_videos": 6,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/griffin-test7.09122025-box-test-05This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "bi_yam_follower",
"total_episodes": 10,
"total_frames": 24927,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 500,
"fps": 30,
"splits": {
"train": "0:10"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/09122025-box-test-05.bimanual-piper-dataset-threecam-test-8This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 3,
"total_frames": 8649,
"total_tasks": 1,
"total_videos": 9,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:3"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test-8.06122025-box-test-01This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "bi_yam_follower",
"total_episodes": 2,
"total_frames": 10755,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 500,
"fps": 30,
"splits": {
"train": "0:2"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/06122025-box-test-01.griffin-test3This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 1,
"total_frames": 251,
"total_tasks": 1,
"total_videos": 2,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/griffin-test3.vietnamese-wiki-segmented
Dataset Card for "vietnamese-wiki-segmented"
More Information needed
bimanual-piper-dataset-threecam-test-6This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 2,
"total_frames": 5292,
"total_tasks": 1,
"total_videos": 6,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:2"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test-6.bimanual-piper-dataset-threecam-test-3This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "bi_piper",
"total_episodes": 2,
"total_frames": 4373,
"total_tasks": 1,
"total_videos": 6,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:2"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/bimanual-piper-dataset-threecam-test-3.11122025-box-test-01This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "bi_yam_follower",
"total_episodes": 16,
"total_frames": 46481,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 500,
"fps": 30,
"splits": {
"train": "0:16"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/11122025-box-test-01.09122025-box-test-04This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "bi_yam_follower",
"total_episodes": 10,
"total_frames": 22194,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 500,
"fps": 30,
"splits": {
"train": "0:10"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/HCHoongChing/09122025-box-test-04.SoMi-ToM
SoMi-ToM: Evaluating Multi-Perspective Theory of Mind in Embodied Social Interactions
NeurIPS 2025
❤️ Like our project on Hugging Face to show your support!
SoMi-ToM Benchmark
We propose the SoMi-ToM benchmark, designed to evaluate multi-perspective ToM in embodied multi-agent complex social interactions. This benchmark is based on rich multimodal interaction data generated by the interaction environment SoMi… See the full description on the dataset page: https://huggingface.co/datasets/hch2000/SoMi-ToM.