datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Uni-GUI-Desktop-1
Uni-GUI-Desktop-1
A large-scale desktop GUI agent trajectory dataset, used as part of the training data for UI-MOPD (Multi-platform On-Policy Distillation for Continual GUI Agent Learning).
Dataset Statistics
Metric
Value
Trajectories
2,685
Total Steps
~36K
Platform
Desktop (1920x1080)
Applications
10 categories
Coordinate System
Normalized to [0, 999]
Applications
App
Description
chrome
Web browsing tasks
gimp… See the full description on the dataset page: https://huggingface.co/datasets/UI-MOPD/Uni-GUI-Desktop-1.Cobot_Magic_desktop_organization
Cobot_Magic_desktop_organization
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: agilex_cobot_decoupled_magic
| Codebase Version: v2.1
End-Effector Type: two_finger_gripper
🏠 Scene Types
This dataset covers the following scene types:
home
office
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/Cobot_Magic_desktop_organization.RMC-AIDA-L_desktop_organization
RMC-AIDA-L_desktop_organization
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: realman_rmc_aidal
| Codebase Version: v2.1
End-Effector Type: two_finger_gripper
🏠 Scene Types
This dataset covers the following scene types:
home
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/RMC-AIDA-L_desktop_organization.ShowUI-desktopGithub | arXiv | HF Paper | Spaces | Datasets | Quick Start
ShowUI-desktop-8K is a UI-grounding dataset focused on PC-based grounding, with screenshots and annotations originally sourced from OmniAct.
We utilize GPT-4o to augment the original annotations, enriching them with diverse attributes such as appearance, spatial relationships, and intended functionality.
You can use our rewrite strategy code to augment your own data.
If you find our work helpful, please consider citing our paper.… See the full description on the dataset page: https://huggingface.co/datasets/showlab/ShowUI-desktop.ShowUI_desktop
Desktop Dataset from ShowUI
This is a FiftyOne dataset with 7496 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset = load_from_hub("Voxel51/ShowUI_desktop")
# Launch the App
session = fo.launch_app(dataset)
Dataset Details… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/ShowUI_desktop.desktop-accessibility-screenshot-json-dumpsUni-GUI-Desktop-2
Uni-GUI-Desktop-2
A desktop GUI agent trajectory dataset collected on OSWorld environments, used as part of the training data for UI-MOPD (Multi-platform On-Policy Distillation for Continual GUI Agent Learning).
Dataset Statistics
Metric
Value
Trajectories
1,247
Total Steps
~14.8K
Platform
Desktop (1920x1080)
Applications
10 categories
Coordinate System
Normalized to [0, 1000]
Applications
App
Count
Description… See the full description on the dataset page: https://huggingface.co/datasets/UI-MOPD/Uni-GUI-Desktop-2.g1_desktop_organize_table_correction
g1_desktop_organize_table_correction
Bimanual tabletop teleoperation on a Unitree G1 with Inspire dexterous hands,
in LeRobot v2.1 format.
Task — Follow the human's corrective gesture and hand the indicated object to the human.
Episodes
192
Frames
67405 (37.4 min @ 30 fps)
Episode length
268–519 frames (median 346)
State / action
26-D / 26-D
Cameras
2 × 640×480
State and action
Both vectors are 26-D and share the same layout:
0– 6… See the full description on the dataset page: https://huggingface.co/datasets/DaoyuanZhu/g1_desktop_organize_table_correction.Cobot_Magic_clear_the_desktop
Cobot_Magic_clear_the_desktop
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: agilex_cobot_decoupled_magic
| Codebase Version: v2.1
End-Effector Type: two_finger_gripper
🏠 Scene Types
This dataset covers the following scene types:
home
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric
Value… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/Cobot_Magic_clear_the_desktop.AIRBOT_MMK2_clean_the_desktop
AIRBOT_MMK2_clean_the_desktop
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: discover_robotics_aitbot_mmk2
| Codebase Version: v2.1
End-Effector Type: five_finger_hand
🏠 Scene Types
This dataset covers the following scene types:
home
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric
Value… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/AIRBOT_MMK2_clean_the_desktop.AIRBOT_MMK2_clean_the_desktop_a
AIRBOT_MMK2_clean_the_desktop_a
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: discover_robotics_aitbot_mmk2
| Codebase Version: v2.1
End-Effector Type: five_finger_hand
🏠 Scene Types
This dataset covers the following scene types:
home
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric
Value… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/AIRBOT_MMK2_clean_the_desktop_a.realman_aidal_desktop_cleanupThe dataset was collected and open-sourced by IO Intelligence, and exported in the LeRobot format provided by the IO Data Platform.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "custom_arm",
"total_episodes": 1099,
"total_frames": 246816,
"total_tasks": 323,
"total_videos": 4396,
"total_chunks": 2,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1099"},
"data_path":… See the full description on the dataset page: https://huggingface.co/datasets/io-intelligence/realman_aidal_desktop_cleanup.AIRBOT_MMK2_desktop_storage
AIRBOT_MMK2_desktop_storage
📋 Overview
This dataset uses an extended format based on LeRobot and is fully compatible with LeRobot.
Robot Type: discover_robotics_aitbot_mmk2
| Codebase Version: v2.1
End-Effector Type: five_finger_hand
🏠 Scene Types
This dataset covers the following scene types:
home
🤖 Atomic Actions
This dataset includes the following atomic actions:
grasp
pick
place
📊 Dataset Statistics
Metric
Value… See the full description on the dataset page: https://huggingface.co/datasets/RoboCOIN/AIRBOT_MMK2_desktop_storage.BCI-Neurofeedback-Desktop
🧠 BCI Neurofeedback Desktop Application Suite
Standalone, zero-dependency desktop application suite for real-time EEG neurofeedback training, cognitive psychometrics, and BCI-driven gaming.
📥 Direct Public Downloads (Latest Release v1.1.2)
Platform
Architecture
Installer
Direct Download Link
Alternative (Latest)
macOS (Apple Silicon)
M1 / M2 / M3 / M4 / arm64
.dmg
Download macOS arm64 DMG (v1.1.2)
Latest arm64 DMG
macOS (Intel)
Intel x86_64 / x64… See the full description on the dataset page: https://huggingface.co/datasets/shahpalash10/BCI-Neurofeedback-Desktop.AgiBotWorld-Beta_G1_task_590_Put_the_desktop_items_into_the_felt_bag
agibot_task_590
This dataset converts the AgiBot format uniformly into LeRobot V3.0.
Dataset Statistics
robot_name: G1
end_effector: 夹爪
task: 将桌面物品放入毛毡袋中。
total_episodes: 136
total_tasks: 1
size: 6.0G
Dataset Structure
├── data
│ └── chunk-xxx
│ ├── file-xxx.parquet
├── meta
│ ├── episodes
│ │ └── chunk-xxx
│ │ └── file-xxx.parquet
│ ├── info.json
│ ├── stats.json
│ └── tasks.parquet
└── videos
├──… See the full description on the dataset page: https://huggingface.co/datasets/BAAI-DataCube/AgiBotWorld-Beta_G1_task_590_Put_the_desktop_items_into_the_felt_bag.clean_the_desktop_1ego-desktop-demo
BaseMatrix EGO Desktop Demo
Human hand manipulation demonstrations captured with binocular egocentric cameras + exocentric view, processed through a 6-stage pipeline into structured observations for robot learning.
This is a preview dataset. Full-scale datasets with hundreds of episodes are available upon request — contact info@basematrix.ai.
Quick Start
from lerobot.common.datasets.lerobot_dataset import LeRobotDataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/basematrix/ego-desktop-demo.desktop-pii-210
Desktop PII 210
This directory is a local Hugging Face-compatible dataset package for synthetic
desktop screenshots with expected privacy and utility QA pairs.
The screenshots are generated synthetic desktop scenes. The visible sensitive
values are fictional benchmark strings, not real personal data.
Contents
images/: all 210 generated PNG screenshots.
images/metadata.jsonl: Hugging Face imagefolder metadata, one row per image.
data/train.parquet: the default Hugging… See the full description on the dataset page: https://huggingface.co/datasets/paperboy-ai/desktop-pii-210.desktopbench
DesktopBench
DesktopBench is the benchmark used by
FOCAL: Filtered On-device Continuous Activity Logging for Efficient Personal
Desktop Summarization. This public v0.1.0
release contains only project-generated annotations, session constructions,
ground-truth summaries, provenance, and deterministic reconstruction tools.
It does not redistribute VideoGUI screenshots, image bytes, screenshot
paths, or copied raw_records containing upstream action text. Complete runtime
inputs are… See the full description on the dataset page: https://huggingface.co/datasets/HaoranYin/desktopbench.supermix-studio-desktop-installer
Supermix Studio X — V48 Frontier Edition
Release Date: 2026-04-17
Downloads
Full Installer (Windows x64)
File
Size
Description
SupermixStudioX_V48_Setup.exe
~1.84 GB
Full installer with all bundled models
Model Weights
Model
File
Size
V48 Frontier
omni_collective_v48_frontier.pth
~20 MB
V47 Frontier
omni_collective_v47_frontier.pth
~501 MB
V46 Frontier
omni_collective_v46_frontier.pth
~501 MB
Benchmarks… See the full description on the dataset page: https://huggingface.co/datasets/Kai9987kai/supermix-studio-desktop-installer.g1_desktop_pick_cream_bun
g1_desktop_pick_cream_bun (TsFile)
Apache TsFile version of DaoyuanZhu/g1_desktop_pick_cream_bun.
Converted rows: 13,278
Data files: ['data/g1_desktop_pick_cream_bun.tsfile']
Usage
Install the Apache TsFile Python SDK (pip install tsfile) and read a converted file:
from pathlib import Path
from tsfile import TsFileReader
path = Path("data/g1_desktop_pick_cream_bun.tsfile")
with TsFileReader(str(path)) as reader:
schemas = reader.get_all_table_schemas()… See the full description on the dataset page: https://huggingface.co/datasets/THULab/g1_desktop_pick_cream_bun.desktop-pii-500
Desktop PII 500
This directory is a local Hugging Face-compatible dataset package for synthetic
desktop screenshots with expected privacy and utility QA pairs.
It reuses the existing 210-image Desktop PII package and adds 290 accepted
supplemental screenshots generated from google/gemini-3.5-flash scenario
prompts and gpt-image-2 image generation.
The screenshots are generated synthetic desktop scenes. The visible sensitive
values are fictional benchmark strings, not real… See the full description on the dataset page: https://huggingface.co/datasets/paperboy-ai/desktop-pii-500.ShowUI-desktopGithub | arXiv | HF Paper | Spaces | Datasets | Quick Start
ShowUI-desktop-8K is a UI-grounding dataset focused on PC-based grounding, with screenshots and annotations originally sourced from OmniAct.
We utilize GPT-4o to augment the original annotations, enriching them with diverse attributes such as appearance, spatial relationships, and intended functionality.
You can use our rewrite strategy code to augment your own data.
If you find our work helpful, please consider citing our paper.… See the full description on the dataset page: https://huggingface.co/datasets/ZhuOnR/ShowUI-desktop.robomind_benchmark1_1_release_franka_3rgb_red_apple_placed_in_the_center_of_the_desktop
benchmark1_1_release_franka_3rgb_red_apple_placed_in_the_center_of_the_desktop
This dataset converts the Robomain format uniformly into LeRobot V3.0.
Dataset Statistics
本体: franka_3rgb
末端执行器: 夹爪
任务平台显示版: 红苹果放在桌面中央
total_episodes: 111
total_tasks: 1
size: 258.0M
Dataset Structure
├── data
│ └── chunk-xxx
│ ├── file-xxx.parquet
├── images
│ └── observation.images.camera_top
├── meta
│ ├── episodes
│ │ └── chunk-xxx
│ │ └── file-xxx.parquet… See the full description on the dataset page: https://huggingface.co/datasets/BAAI-DataCube/robomind_benchmark1_1_release_franka_3rgb_red_apple_placed_in_the_center_of_the_desktop.clean_desktop_v1This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 120,
"total_frames": 51218,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:120"},
"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/joanfox/clean_desktop_v1.gui-agent-desktop-sft-datascreenspot-v2-annots-desktopeasyr1-103k-4MP-stage-three-temp-1_7-RL-only-ui-vision-jedi-show-ui-desktop-gta-0p0-zerodesktop-ui-captionseasyr1-103k-4MP-stage-three-temp-1_7-RL-only-ui-vision-jedi-show-ui-desktop-0p2-zero