datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
idmIDMR-train
📑 IDMR Dataset (Training Set)
This repository provides the IDMR training dataset, accompanying the paper:
IDMR: Towards Instance-Driven Precise Visual Correspondence in Multimodal Retrieval
📄 arXiv 2504.00954 | 💻 GitHub
📘 Overview
The IDMR dataset is designed for Instance-Driven Multimodal Retrieval, supporting large-scale training across diverse multimodal sources.
This dataset enables research on:
Cross-dataset multimodal retrieval
Instance-driven alignment… See the full description on the dataset page: https://huggingface.co/datasets/lbw18601752667/IDMR-train.idm_ft_newThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 501,
"total_frames": 154063,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"fps": 30,
"splits": {
"train": "0:501"
},
"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/FrozenAngel/idm_ft_new.libero-singleview-idm
LIBERO 单视角离线 IDM：Codex 技术交接与运行手册
本文档是项目迁移到另一台机器后，供 Codex 或工程人员直接接手使用的
主运行手册。除非明确需要改变实验设计，否则应以本文、cfg/ 和
scripts/ 中的配置为准。
0. 接手结论与执行原则
项目目标是把离线 LIBERO 单视角视频转换为机器人动作：
agentview 离线视频
→ 每 16 帧生成 20×20 个均匀点轨迹
→ 单视角 Motion Tokenizer
→ 对应 LIBERO 系列的 IDM
→ 16 步 × 7 维动作
接手此项目的 Codex 应遵守以下原则：
先进入项目根目录，再执行任何脚本；所有默认路径均相对于项目根目录。
先运行资产校验。校验通过时，不要重新下载或重新生成 LIBERO 和轨迹数据。
先进行单卡或 DDP 冒烟测试，再启动完整训练。
Motion Tokenizer 只训练一个共享模型；五个 LIBERO 系列分别训练自己的 IDM。
不训练 Forward… See the full description on the dataset page: https://huggingface.co/datasets/yimmmmmy/libero-singleview-idm.idm-eval
IDM Eval Set
A validation set for evaluating Inverse Dynamics Models on macOS screen recordings. Each sample is a 5-second clip of real productivity desktop usage (browser, IDE, terminal, docs, dashboards) paired with a ground-truth action log captured at the OS level.
The task: given a short screen recording, predict the sequence of user input actions (keypresses, mouse clicks, scrolls, cursor moves) that produced the observed screen changes.
Code
Training… See the full description on the dataset page: https://huggingface.co/datasets/p-doom/idm-eval.IDMGSPTODOidm_testThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "so_follower",
"total_episodes": 10,
"total_frames": 3167,
"total_tasks": 1,
"chunks_size": 1000,
"data_files_size_in_mb": 100,
"video_files_size_in_mb": 200,
"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/FrozenAngel/idm_test.id_multilabel_hsThe ID_MULTILABEL_HS dataset is collection of 13,169 tweets in Indonesian language,
designed for hate speech detection NLP task. This dataset is combination from previous research and newly crawled data from Twitter.
This is a multilabel dataset with label details as follows:
-HS : hate speech label;
-Abusive : abusive language label;
-HS_Individual : hate speech targeted to an individual;
-HS_Group : hate speech targeted to a group;
-HS_Religion : hate speech related to religion/creed;
-HS_Race : hate speech related to race/ethnicity;
-HS_Physical : hate speech related to physical/disability;
-HS_Gender : hate speech related to gender/sexual orientation;
-HS_Gender : hate related to other invective/slander;
-HS_Weak : weak hate speech;
-HS_Moderate : moderate hate speech;
-HS_Strong : strong hate speech.id-maxD1000
max_context_length: 128
max_documents_per_subject: 1000
IDM_PCIDMR-test
📑 IDMR Dataset (Test Set)
This repository provides the IDMR test dataset, accompanying the paper:
IDMR: Towards Instance-Driven Precise Visual Correspondence in Multimodal Retrieval
📄 arXiv 2504.00954 | 💻 GitHub
📘 Overview
The IDMR-test dataset is designed to evaluate instance-driven multimodal retrieval.
We provide two types of evaluation scenarios:
In-domain evaluation: derived from benchmarks also used in training (COCO, Objects365, OpenImages).
Out-of-domain… See the full description on the dataset page: https://huggingface.co/datasets/lbw18601752667/IDMR-test.IDM-Multiple
IDM-Multiple
IDM-Multiple is part of the IDMBench in AgentStudio for evaluating agent's ability to learn from Internet-scale videos without action labels. Specifically, this dataset requires VLMs to answer an action sequence within a video clip of screen recoding. Please note that this dataset is a test set rather than a training set. Therefore, please do not use it for training. More details are provided in the project page.
id-maxD500
max_context_length: 128
max_documents_per_subject: 500
gr1-lerobot-dreamgen-idmIDM-Single
IDM-Single
IDM-Single is part of the IDMBench in AgentStudio for evaluating agent's ability to learn from Internet-scale videos without action labels. Specifically, this dataset requires VLMs to answer the action transition between two screenshots. Please note that this dataset is a test set rather than a training set. Therefore, please do not use it for training. More details are provided in the project page.
ID-MoCQA
ID-MoCQA: Indonesian Multi-hop Question Answering for Complex Cultural Understanding
Dataset Description
Paper: No Shortcuts to Culture: Indonesian Multi-hop Question Answering for Complex Cultural Understanding
Dataset Summary
ID-MoCQA is the first large-scale multi-hop question answering dataset for assessing the cultural understanding of large language models (LLMs), grounded in Indonesian traditions and available in both English and Indonesian.
Unlike… See the full description on the dataset page: https://huggingface.co/datasets/vynsk/ID-MoCQA.AGI-CAST-idm-actions
AGI-CAST IDM Actions
We release model-generated keyboard, mouse-click, cursor-movement and scroll annotations for the AGI-CAST-0.6k screencast dataset.
The labels were generated with p-doom/idm, an inverse dynamics model trained to recover user input actions from short windows of screen recordings.
Please refer to the AGI-CAST blog post and the IDM blog post for more context.
Code
These labels were generated with p-doom/idm.
The… See the full description on the dataset page: https://huggingface.co/datasets/p-doom/AGI-CAST-idm-actions.id-maxD100
max_context_length: 128
max_documents_per_subject: 100
i2v10k_posttrain_idm_lerobotIDMGSP-danish
IDMGSP Danish Translated
Danish translated version of the original IDMGSP dataset. This Danish version contains all abstracts from the classifier_input subset. Translated using an Nllb-200-Distilled-600M. Made as part of the DM-AI 2023 competition
idm_openarm_merged_256_v2
idm_openarm_merged_256_v2 (TsFile)
Apache TsFile version of glory-hyeok/idm-openarm-merged-256-v2.
Overview
A LeRobot robot dataset recorded on a openarm_rh56f1 arm. Task(s): 0; 1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 11; 12; 13; 14; 15. Each frame holds the commanded action and observed observation.state joint positions, plus camera views stored as videos in the original dataset.
Episodes: 1156
Frames: 492,920
Sampling rate: 20 fps
Tasks: 16 — "0; 1; 2; 3; 4; 5; 6; 7; 8;… See the full description on the dataset page: https://huggingface.co/datasets/THULab/idm_openarm_merged_256_v2.africa-idmc-event-data-for-som
Somalia - Internal Displacements Updates (IDU) (event data) | Africa (original)
Size category: 1K<n<10K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Public datasets… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-event-data-for-som.africa-idmc-idp-data-com
Comoros - Internal Displacements (New Displacements) – IDPs | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Public datasets help… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-idp-data-com.africa-idmc-idp-data-dza
Algeria - Internal Displacements (New Displacements) – IDPs | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Public datasets help… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-idp-data-dza.my-lerobot-gr1-idmIDMGSPTODOafrica-idmc-event-data-for-ssd
South Sudan - Internal Displacements Updates (IDU) (event data) | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Public datasets… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-event-data-for-ssd.africa-idmc-idp-data-lbr
Liberia - Internal Displacements (New Displacements) – IDPs | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Public datasets help… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-idp-data-lbr.africa-idmc-event-data-for-cod
Democratic Republic of the Congo - Internal Displacements Updates (IDU) (event data) | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-event-data-for-cod.africa-idmc-idp-data-caf
Central African Republic - Internal Displacements (New Displacements) – IDPs | Africa (original)
Size category: n<1K - Formats: parquet - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-idmc-idp-data-caf.