datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
openai_humaneval
Dataset Card for OpenAI HumanEval
Dataset Summary
The HumanEval dataset released by OpenAI includes 164 programming problems with a function sig- nature, docstring, body, and several unit tests. They were handwritten to ensure not to be included in the training set of code generation models.
Supported Tasks and Leaderboards
Languages
The programming problems are written in Python and contain English natural text in comments and docstrings.… See the full description on the dataset page: https://huggingface.co/datasets/openai/openai_humaneval.humanevalplusMedical-Eval-HumanityLastExamhumanevalpack
Dataset Card for HumanEvalPack
Dataset Summary
HumanEvalPack is an extension of OpenAI's HumanEval to cover 6 total languages across 3 tasks. The Python split is exactly the same as OpenAI's Python HumanEval. The other splits are translated by humans (similar to HumanEval-X but with additional cleaning, see here). Refer to the OctoPack paper for more details.
Languages: Python, JavaScript, Java, Go, C++, Rust
OctoPack🐙🎒:
Data
CommitPack
4TB of GitHub commits… See the full description on the dataset page: https://huggingface.co/datasets/bigcode/humanevalpack.HumanEval-XLThis dataset contains a viewer-friendly version of the dataset at FloatAI/HumanEval-XL. It is made available separately for the convenience of the vllm-code-harness package.
DECO-50
DECO: Decoupled Multimodal Diffusion Transformer for Bimanual Dexterous Manipulation with a Plugin Tactile Adapter
DECO-50 is a bimanual dexterous manipulation dataset with tactile sensing, comprising 50 hours of teleoperated data across 4 scenarios and 28 subtasks, totaling over 5 million frames collected on real dual-arm robots.
Dataset Structure
DECO-50/
├── task1/
│ ├── sub_task_1/
│ │ ├── episode_000000/
│ │ │ ├──… See the full description on the dataset page: https://huggingface.co/datasets/BAAI-Humanoid/DECO-50.HumanVidesg_reports_human_labeled_v2
Vidore Benchmark 2 - ESG Human Labeled
This dataset is part of the "Vidore Benchmark 2" collection, designed for evaluating visual retrieval applications. It focuses on the theme of ESG reports from the fast food industry.
Dataset Summary
Each query is in english.
This dataset provides a focused benchmark for visual retrieval tasks related to ESG reports for the fast food industry. It includes a curated set of documents, queries, relevance judgments (qrels), and page… See the full description on the dataset page: https://huggingface.co/datasets/vidore/esg_reports_human_labeled_v2.MOSAIC_Dataset
MOSAIC Dataset
Project Page | Paper | Code | Dataset | Model
This repository releases the built-in MOSAIC multi-source motion dataset in the following paper:
MOSAIC: Bridging the Sim-to-Real Gap in Generalist Humanoid Motion Tracking and Teleoperation with Rapid Residual Adaptation
The dataset is organized into:
Human motions stored in an AMASS-style format
Unitree G1 motions retargeted from human motions and converted to NPZ for training/visualization
It includes motions from:… See the full description on the dataset page: https://huggingface.co/datasets/BAAI-Humanoid/MOSAIC_Dataset.human_behavior_atlas
Human Behavior Atlas
A large-scale multimodal dataset for human behavior understanding, spanning emotion recognition, sentiment analysis, humor detection, mental health screening, and video question answering. The dataset integrates 16 source datasets into a unified schema with audio, video, and pre-extracted features.
This dataset was used to train OmniSapiens, a foundation model for social behavior processing.
Papers:
Human Behavior Atlas: Benchmarking Unified Psychological and… See the full description on the dataset page: https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas.HSTLI_A-Dataset-of-Human-Semen-Time-Lapse-Images
HSTLI: A Dataset of Human Semen Time Lapse Images
Dataset Details
HSTLI contains 3,266 time-lapse microscopy videos of human sperm.Clips were recorded from two imaging modalities:
CASA system (Sperm Class Analyzer)
Optical microscope (Swift M10DB-MP + Fujifilm X-T30)
A subset of videos was manually annotated with bounding boxes around each visible sperm head.
The dataset supports detection, tracking and motility computation.
Total contents:
34… See the full description on the dataset page: https://huggingface.co/datasets/DFL-KamLab/HSTLI_A-Dataset-of-Human-Semen-Time-Lapse-Images.si_for_sdhumaneval_pythonarena-human-preference-140k
Overview
This dataset contains user votes collected in the text-only category. Each row represents a single vote judging two models (model_a and model_b) on a user conversation, along with the full conversation history and metadata. Key fields include:
id: Unique feedback ID of each vote/row.
evaluation_session_id: Unique ID of each evaluation session, which can contain multiple separate votes/evaluations.
evaluation_order: Evaluation order of the current vote.
winner: Battle… See the full description on the dataset page: https://huggingface.co/datasets/lmarena-ai/arena-human-preference-140k.human_translated_arabic_mmluhumanoid-robots-training-dataset
Dynamic Intelligence — Humanoid Robot Training Dataset
A first-person (egocentric) video dataset of human hand manipulation, designed for training humanoid robot policies via imitation learning. Each episode captures a person performing an everyday household task — folding clothes, moving dishes, opening doors — filmed from a head-mounted iPhone using its built-in LiDAR and depth sensors.
The dataset pairs each video with frame-level 3D hand tracking and camera pose data, giving… See the full description on the dataset page: https://huggingface.co/datasets/DynamicIntelligence/humanoid-robots-training-dataset.arena-human-preference-55kDataset for Kaggle competition on predicting human preference on Chatbot Arena battles.
The training dataset includes over 55,000 real-world user and LLM conversations and user preferences across over 70 state-of-the-art LLMs, such as GPT-4, Claude 2, Llama 2, Gemini, and Mistral models.
Each sample represents a battle consisting of 2 LLMs which answer the same question, with a user label of either prefer model A, prefer model B, tie, or tie (both bad).
Citation
Please cite the… See the full description on the dataset page: https://huggingface.co/datasets/lmarena-ai/arena-human-preference-55k.Real_sd_ds_0701human_eval_cppgsd-humaneval-annotationsmt_bench_human_judgments
Content
This dataset contains 3.3K expert-level pairwise human preferences for model responses generated by 6 models in response to 80 MT-bench questions.
The 6 models are GPT-4, GPT-3.5, Claud-v1, Vicuna-13B, Alpaca-13B, and LLaMA-13B. The annotators are mostly graduate students with expertise in the topic areas of each of the questions. The details of data collection can be found in our paper.
Agreement Calculation
This Colab notebook shows how to compute the… See the full description on the dataset page: https://huggingface.co/datasets/lmsys/mt_bench_human_judgments.HumanTracker
Dataset Card for HumanTracker
Project page · Paper · Code
HumanTracker is a humanoid motion-tracking benchmark. This release contains two complementary subsets:
motions/ — the evaluation test split: retargeted 29-DoF reference trajectories, grouped into four motion families.
preference_pair/ — 6,000 human preference pairs, each stored with the two tracker rollouts that were compared and the source-motion clip they track.
The evaluation harness and HumanScore reward model live… See the full description on the dataset page: https://huggingface.co/datasets/GalaxyGeneralRobotics/HumanTracker.text-2-image-Rich-Human-Feedback
Building upon Google's research Rich Human Feedback for Text-to-Image Generation we have collected over 1.5 million responses from 152'684 individual humans using Rapidata via the Python API. Collection took roughly 5 days.
If you get value from this dataset and would like to see more in the future, please consider liking it.
Overview
We asked humans to evaluate AI-generated images in style, coherence and prompt alignment. For images that contained flaws, participants were… See the full description on the dataset page: https://huggingface.co/datasets/Rapidata/text-2-image-Rich-Human-Feedback.humaneval-multilingualarena-human-preference-100k
Overview
This dataset contains leaderboard conversation data collected between June 2024 and August 2024.
It includes English human preference evaluations used to develop Arena Explorer.
Additionally, we provide an embedding file, which contains precomputed embeddings for the English conversations.
These embeddings are used in the topic modeling pipeline to categorize and analyze these conversations.
For a detailed exploration of the dataset and analysis methods, refer to the… See the full description on the dataset page: https://huggingface.co/datasets/lmarena-ai/arena-human-preference-100k.text-2-video-human-preferences
Rapidata Video Generation Preference Dataset
This dataset was collected in ~12 hours using the Rapidata Python API, accessible to anyone and ideal for large scale data annotation.
The data collected in this dataset informs our text-2-video model benchmark. We just started so currently only two models are represented in this set:
Sora
Hunyouan
Pika 2.0
Runway ML Alpha
Luma Ray 2
Explore our latest model rankings on our website.
If you get value from this dataset and would… See the full description on the dataset page: https://huggingface.co/datasets/Rapidata/text-2-video-human-preferences.text-2-video-human-preferences-wan2.1
Rapidata Video Generation Alibaba Wan2.1 Human Preference
If you get value from this dataset and would like to see more in the future, please consider liking it.
This dataset was collected in ~1 hour total using the Rapidata Python API, accessible to anyone and ideal for large scale data annotation.
Overview
In this dataset, ~45'000 human annotations were collected to evaluate Alibaba Wan 2.1 video generation model on our benchmark. The up to date benchmark… See the full description on the dataset page: https://huggingface.co/datasets/Rapidata/text-2-video-human-preferences-wan2.1.PPE-Human-Preference-V1
Overview
This contains the human preference evaluation set for Preference Proxy Evaluations.
This dataset is meant for benchmarking and evaluation, not for training.
Paper
Code
License
User prompts are licensed under CC-BY-4.0, and model outputs are governed by the terms of use set by the respective model providers.
Citation
@misc{frick2024evaluaterewardmodelsrlhf,
title={How to Evaluate Reward Models for RLHF},
author={Evan Frick and Tianle Li and… See the full description on the dataset page: https://huggingface.co/datasets/lmarena-ai/PPE-Human-Preference-V1.Env-TTS-Clean
Env-TTS-Clean
Environment-aware text-to-speech training corpus (clean release). Each row
pairs four short 24 kHz mono FLAC clips with aligned transcripts:
an environment sample (different speaker, same acoustic scene),
a speaker reference (same speaker as the target utterance),
a speaker-enhanced copy of the reference (MossFormer2 enhancement — or, for
the DDS source, the real clean-studio recording of the speaker reference),
the target speech to synthesise,
so a model can… See the full description on the dataset page: https://huggingface.co/datasets/humanify/Env-TTS-Clean.robocasa_pretrain_human300_v4_annotated5This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 20,
"features": {
"observation.images.robot0_agentview_left": {
"dtype": "video",
"shape": [
256,
256,
3
],
"names": [
"height",
"width",
"channel"
],
"video_info": {… See the full description on the dataset page: https://huggingface.co/datasets/pepijn223/robocasa_pretrain_human300_v4_annotated5.