datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ha-lab-dataset
Hugging Face Dataset Upload Notes
This note summarizes common commands for uploading and managing selected
folders in the Hugging Face dataset repository.
Repository:
sehoonha/ha-lab-dataset
Login
Use a Hugging Face token with write permission.
hf auth login
hf auth whoami
Upload One Folder
Upload only the local test_folder/ directory to test_folder/ in the dataset
repository.
hf upload sehoonha/ha-lab-dataset test_folder test_folder --repo-type… See the full description on the dataset page: https://huggingface.co/datasets/sehoonha/ha-lab-dataset.towel_half_fold_bimanual_4views_20260911_115811This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 30,
"features": {
"action": {
"dtype": "float32",
"names": [
"left_shoulder_pan.pos",
"left_shoulder_lift.pos",
"left_elbow_flex.pos",
"left_wrist_flex.pos",
"left_wrist_roll.pos",
"left_gripper.pos"… See the full description on the dataset page: https://huggingface.co/datasets/Woohi123/towel_half_fold_bimanual_4views_20260911_115811.berkeley_gnm_cory_hallThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "unknown",
"total_episodes": 7331,
"total_frames": 156012,
"total_tasks": 1,
"chunks_size": 1000,
"fps": 5,
"splits": {
"train": "0:7331"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path": "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4"… See the full description on the dataset page: https://huggingface.co/datasets/lerobot/berkeley_gnm_cory_hall.HAL
Human-related Anomaly Localization Dataset
To extend the application of temporal action localization to the more practical domains such as human-related anomaly detection, we construct a new Human-related Anomaly Localization (HAL) benchmark. The core feature of HAL is the Chain-of-Evidence (CoE) textual descriptions that we newly generated. Compared to the textual information used in prior works like T3AL, this new format is richer in logic and more clearly structured. To… See the full description on the dataset page: https://huggingface.co/datasets/irisreto/HAL.lvdec-flf-wave-20260815-generated
LVDEC FLF Wave 2026-08-15 — Generated Returns
This public dataset contains the completed generated return package for the
Yueshengli/lvdec-flf-wave-20260815
execution wave.
Contents
The dataset contains 8,193 completed jobs with no final failures:
Generator / batch
Jobs
ltx23_flf2v
1,200
ltx23_flf2va
1,200
ltx25_flf2v
470
ltx25_flf2va
370
wan22_fun_5b_control
990
minimax_h3_flf2v
1,166
minimax_h3_flf2va
2,797
Each job directory may… See the full description on the dataset page: https://huggingface.co/datasets/Hallucinatie/lvdec-flf-wave-20260815-generated.VidPair-Halluc-ClipVript-HAL
🎬 Vript: Refine Video Captioning into Video Scripting [Github Repo]
Vript-HAL
Vript-HAL is the first benchmark evaluating action and object hallucinations in video LLMs
Getting Started
By downloading these datasets, you agree to the terms of the License.
Vript-HAL/
|
├── HAL_scenes/
│ ├── -_MRAAhEKio-Scene-010.mp4
│ └── ...
│
└── HAL_annotations.jsonl
HAL_scenes: The trimmed video clips in the Vript-HAL benchmark.
HAL_annotations.jsonl: The… See the full description on the dataset page: https://huggingface.co/datasets/Mutonix/Vript-HAL.fold_pink_cloth_in_halfThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 15,
"features": {
"action": {
"dtype": "float32",
"shape": [
7
],
"names": [
"left_shoulder_pan_joint.pos",
"left_shoulder_lift_joint.pos",
"left_elbow_joint.pos",
"left_wrist_1_joint.pos"… See the full description on the dataset page: https://huggingface.co/datasets/khoviadin/fold_pink_cloth_in_half.towel_half_fold_bimanual_20260911_113524This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 30,
"features": {
"action": {
"dtype": "float32",
"names": [
"left_shoulder_pan.pos",
"left_shoulder_lift.pos",
"left_elbow_flex.pos",
"left_wrist_flex.pos",
"left_wrist_roll.pos",
"left_gripper.pos"… See the full description on the dataset page: https://huggingface.co/datasets/Woohi123/towel_half_fold_bimanual_20260911_113524.Halluc0513_value1pick-mustard-wholebody-50hz-half-alignedThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "Unitree_G1_Dex3_GR00T_N16",
"total_episodes": 101,
"total_frames": 23497,
"total_tasks": 1,
"total_videos": 101,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:101"
},
"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/hyeonseong-kim-k/pick-mustard-wholebody-50hz-half-aligned.Hallu_value1Grab_the_cube1This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 36,
"total_frames": 21861,
"total_tasks": 1,
"total_videos": 72,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:36"
},
"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/Hale-Sage/Grab_the_cube1.halfplugcam1video-hallucinationtowel_half_fold_bimanual_depth_viz_20260911_114651This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 30,
"features": {
"action": {
"dtype": "float32",
"names": [
"left_shoulder_pan.pos",
"left_shoulder_lift.pos",
"left_elbow_flex.pos",
"left_wrist_flex.pos",
"left_wrist_roll.pos",
"left_gripper.pos"… See the full description on the dataset page: https://huggingface.co/datasets/Woohi123/towel_half_fold_bimanual_depth_viz_20260911_114651.agilex_pour_water_cup_halfThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.0",
"robot_type": "arx5_bimanual",
"total_episodes": 19,
"total_frames": 10048,
"total_tasks": 1,
"total_videos": 57,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 25,
"splits": {
"train": "0:19"
},
"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/villekuosmanen/agilex_pour_water_cup_half.half_and_halfmerge_dataThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 700,
"total_frames": 325609,
"total_tasks": 4,
"total_videos": 1400,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:700"
},
"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/Hale-Sage/merge_data.towel_unfold_half_fold_bimanual_notape_v1_20260914This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 30,
"features": {
"action": {
"dtype": "float32",
"names": [
"left_shoulder_pan.pos",
"left_shoulder_lift.pos",
"left_elbow_flex.pos",
"left_wrist_flex.pos",
"left_wrist_roll.pos",
"left_gripper.pos"… See the full description on the dataset page: https://huggingface.co/datasets/Woohi123/towel_unfold_half_fold_bimanual_notape_v1_20260914.Grab_cube_taskThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 100,
"total_frames": 63648,
"total_tasks": 1,
"total_videos": 200,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:100"
},
"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/Hale-Sage/Grab_cube_task.record-testThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 2,
"total_frames": 2159,
"total_tasks": 1,
"total_videos": 4,
"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/Hale-Sage/record-test.rush_hour_v1_half
Rush Hour pi0.7 LeRobot v3 dataset
50 completed teleoperation episodes, 73539 frames, 20 Hz. This is an
immutable derivative of the state-only Rush Hour JSON/NPZ recordings.
The export matches the LEGO pi0.7 storage contract: observation.state and
action are 7-D float32 joint vectors; observation.images.main,
observation.images.wrist, and observation.images.goal are 240×160 H.264
streams. The goal stream uses the main view with the recorded green source and
purple target cues.… See the full description on the dataset page: https://huggingface.co/datasets/pyrrosk/rush_hour_v1_half.Grab_the_cube_3camThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 3,
"total_frames": 2441,
"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/Hale-Sage/Grab_the_cube_3cam.Test1halfplug2cam1halfplug3cam1Hallu_value2Grab_the_cube_2camThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "koch_follower",
"total_episodes": 101,
"total_frames": 27659,
"total_tasks": 1,
"total_videos": 202,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:101"
},
"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/Hale-Sage/Grab_the_cube_2cam.halfplugcam2