Datasets

ABC-130k ABC-130k is the largest open-source robot teleoperation dataset. It contains bimanual manipulation trajectories collected on two-arm YAM stations. Episodes are distributed as MCAP files, with subtask annotations kept as separate artifacts so they can be revised or extended independently of the underlying episode data. For details on the accompanying paper, see abc.bot. Please see the GitHub repo here for code to train and deploy with this dataset. Dataset… See the full description on the dataset page: https://huggingface.co/datasets/XDOF/ABC-130k.

🚀 LLaVA-One-Vision-1.5-Mid-Training-85M Dataset is being uploaded 🚀 Upload Status All Completed: ImageNet-21k、LAIONCN、DataComp-1B、Zero250M、COYO700M、SA-1B、MINT、Obelics 📜 Cite If you find LLaVA-One-Vision-1.5-Mid-Training-85M useful in your research, please consider to cite the following related papers: @misc{an2025llavaonevision15fullyopenframework, title={LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training}… See the full description on the dataset page: https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-1.5-Mid-Training-85M.

Completely uncurated collection of IRC logs from the Ubuntu IRC channels

The ability to solve problems is a hallmark of intelligence and has been an enduring goal in AI. AI systems that can create programs as solutions to problems or assist developers in writing programs can increase productivity and make programming more accessible. Recently, pre-trained large language models have shown impressive abilities in generating new codes from natural language descriptions, repairing buggy codes, translating codes between languages, and retrieving relevant code segments. However, the evaluation of these models has often been performed in a scattered way on only one or two specific tasks, in a few languages, at a partial granularity (e.g., function) level and in many cases without proper training data. Even more concerning is that in most cases the evaluation of generated codes has been done in terms of mere lexical overlap rather than actual execution whereas semantic similarity (or equivalence) of two code segments depends only on their ``execution similarity'', i.e., being able to get the same output for a given input.

Dataset Card for MMLU Dataset Summary Measuring Massive Multitask Language Understanding by Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt (ICLR 2021). This is a massive multitask test consisting of multiple-choice questions from various branches of knowledge. The test spans subjects in the humanities, social sciences, hard sciences, and other areas that are important for some people to learn. This covers 57 tasks… See the full description on the dataset page: https://huggingface.co/datasets/cais/mmlu.

FineNews FineNews is a multilingual news-text dataset for language-model research. It filters and deduplicates a 2021–2025 snapshot of the INFINI-NEWS Corpus, which extracts articles from Common Crawl CC-News. At a glance Measure FineNews Input articles 852,824,802 Infini-News rows from 2021–2025 Output 392,627,654 physical rows Files 294,509 Parquet files Folders 60 publication months (2021-01 to 2025-12), then language Language folders 129… See the full description on the dataset page: https://huggingface.co/datasets/ksolovev/fine-news.

Dataset Card for "ai2_arc" Dataset Summary A new dataset of 7,787 genuine grade-school level, multiple-choice science questions, assembled to encourage research in advanced question-answering. The dataset is partitioned into a Challenge Set and an Easy Set, where the former contains only questions answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. We are also including a corpus of over 14 million science sentences relevant to… See the full description on the dataset page: https://huggingface.co/datasets/allenai/ai2_arc.

Dataset Card for GLUE Dataset Summary GLUE, the General Language Understanding Evaluation benchmark (https://gluebenchmark.com/) is a collection of resources for training, evaluating, and analyzing natural language understanding systems. Supported Tasks and Leaderboards The leaderboard for the GLUE benchmark can be found at this address. It comprises the following tasks: ax A manually-curated evaluation dataset for fine-grained analysis of system… See the full description on the dataset page: https://huggingface.co/datasets/nyu-mll/glue.

MNBVC: Massive Never-ending BT Vast Chinese corpus

AutoDex: Real-World Dexterous Grasp Experience Autonomously building reusable grasp knowledge through real-world trials.CoRL 2026 Project · Paper · Interactive Gallery · Code AutoDex lets a robot try grasps, verify outcomes, and reset objects autonomously—turning real-world experience into a reusable dexterous grasp library. Browse recorded trials in the interactive gallery, or download grasp annotations, robot trajectories, and multi-view recordings for your own experiments.… See the full description on the dataset page: https://huggingface.co/datasets/willi19/autodex-gallery.

GenPairs GenPairs is a dataset of synthetic images created to probe the impact of concept pairs on multimodal model task accuracy. Each image features at least two concepts, one of which is an ImageNet class, and the model is tasked with producing the correct ImageNet classification in the presence of the other concept(s). We connect the model accuracy to the prevalence of these concept pairs in the model's pretraining dataset, specifically via pointwise mutual information… See the full description on the dataset page: https://huggingface.co/datasets/helenqu/GenPairs.

Hacker News - Complete Archive Every Hacker News item since 2006, live-updated every 5 minutes What is it? This dataset contains the complete Hacker News archive: every story, comment, Ask HN, Show HN, job posting, and poll ever submitted to the site. Hacker News is one of the longest-running and most influential technology communities on the internet, operated by Y Combinator since 2007. It has become the de facto gathering place for founders, engineers… See the full description on the dataset page: https://huggingface.co/datasets/open-index/hacker-news.

GIFT-Eval Pre-training Datasets Pretraining dataset aligned with GIFT-Eval that has 71 univariate and 17 multivariate datasets, spanning seven domains and 13 frequencies, totaling 4.5 million time series and 230 billion data points. Notably this collection of data has no leakage issue with the train/test split and can be used to pretrain foundation models that can be fairly evaluated on GIFT-Eval. 📄 Paper 🖥️ Code 📔 Blog Post 🏎️ Leader Board Ethical Considerations… See the full description on the dataset page: https://huggingface.co/datasets/Salesforce/GiftEvalPretrain.

📚 FineWeb-Edu 1.3 trillion tokens of the finest educational data the 🌐 web has to offer Paper: https://arxiv.org/abs/2406.17557 What is it? 📚 FineWeb-Edu dataset consists of 1.3T tokens and 5.4T tokens (FineWeb-Edu-score-2) of educational web pages filtered from 🍷 FineWeb dataset. This is the 1.3 trillion version. To enhance FineWeb's quality, we developed an educational quality classifier using annotations generated by LLama3-70B-Instruct. We then… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu.

MMLU-Pro Dataset MMLU-Pro dataset is a more robust and challenging massive multi-task understanding dataset tailored to more rigorously benchmark large language models' capabilities. This dataset contains 12K complex questions across various disciplines. |Github | 🏆Leaderboard | 📖Paper | 🚀 What's New [2026.03.11] Added more cutting-edge frontier models to the leaderboard, including the Claude-4.6 series, Seed2.0 series, Qwen3.5 series, and Gemini-3.1-Pro, among… See the full description on the dataset page: https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro.

LingBot-Depth Dataset Self-curated RGB-D dataset for training LingBot-Depth, a masked depth modeling approach (arxiv:2601.17895). Each sample contains an RGB image, raw sensor depth, and ground truth depth. Total size: 2.71 TBDepth scale: millimeters (mm), stored as 16-bit PNGLicense: CC BY-NC-SA 4.0 Sub-datasets Name Description Samples RobbyReal Real-world indoor scenes captured with multiple RGB-D cameras 1,400,000 RobbyVla Real-world data collected… See the full description on the dataset page: https://huggingface.co/datasets/robbyant/mdm_depth.

🍃 MINT-1T:Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens 🍃 MINT-1T is an open-source Multimodal INTerleaved dataset with 1 trillion text tokens and 3.4 billion images, a 10x scale-up from existing open-source datasets. Additionally, we include previously untapped sources such as PDFs and ArXiv papers. 🍃 MINT-1T is designed to facilitate research in multimodal pretraining. 🍃 MINT-1T is created by a team from the University of Washington in… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations/MINT-1T-PDF-CC-2023-40.

Picbreeder-VLM Archive Every image evolved by the swarm of vision-language-model "breeders" in In Search of the Ingredients of Open-Endedness: Replicating Picbreeder with Large Vision-Language Models (GECCO 2026), together with the CPPN genomes that produced them, the agents' reasoning transcripts, the lineage graphs, and the analysis artifacts behind the paper and blog. The original Picbreeder (Secretan et al., 2008) let crowds of humans collaboratively evolve images from CPPN… See the full description on the dataset page: https://huggingface.co/datasets/picbreeder-vlm/picbreeder-vlm-archive.

Open-AoE — Egocentric Hand Manipulation Dataset Release Roadmap Tier Duration Status nano ~3 h ✅ Released tiny ~100 h ✅ Released full 2000 h 🚧 Uploading (batch4–6 ≈694h ready; see notes) Release notes 2026-07-30: Removed samples flagged in PR #1 for camera-intrinsics vs. video-resolution mismatches. 2026-07-31: Uploaded ~323h of data. 2026-08-12: Uploaded ~694h of data(ms only, waiting for HF storage expansion). Additional… See the full description on the dataset page: https://huggingface.co/datasets/inclusionAI/OpenAoE-2000h.

Further cleaning done. Please look through the dataset and ensure that I didn't miss anything. Update: Confirmed working method for training the model: https://huggingface.co/AlekseyKorshuk/vicuna-7b/discussions/4#64346c08ef6d5abefe42c12c Two choices: Removes instances of "I'm sorry, but": https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/blob/main/ShareGPT_V3_unfiltered_cleaned_split_no_imsorry.json Has instances of "I'm sorry, but":… See the full description on the dataset page: https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered.