datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
jee-advanced-questions
JEE Advanced — Question Bank
A structured dataset of JEE Advanced examination questions with full
worked solutions and diagrams. JEE Advanced questions are more analytical
than JEE Main — many are subjective, integer, or numerical-answer type with
detailed multi-step solutions.
Subsets (PCM):
Physics — 50 questions
Chemistry — 21 questions
Mathematics — 48 questions
Structure
Organised into subsets by subject and splits (train / test):
mathematics/ physics/… See the full description on the dataset page: https://huggingface.co/datasets/eQOURSE/jee-advanced-questions.JEEM
Dataset Card for JEEM
📊 Curated by: Toloka, MBZUAI
🌐 Language(s): Modern Standard Arabic, 🇯🇴 Jordanian dialect, 🇪🇬 Egyptian dialect, 🇦🇪 Emirati dialect, 🇲🇦 Moroccan dialect
🔍 Dataset Description
JEEM is a benchmark dataset designed to evaluate Vision-Language Models (VLMs) in the context of Arabic dialectal diversity. It includes one representative dialect from each major Arabic dialectal region: 🇯🇴 Jordanian (Levantine), 🇪🇬 Egyptian, 🇦🇪 Emirati… See the full description on the dataset page: https://huggingface.co/datasets/toloka/JEEM.jee-advanced-questions
JEE Advanced — Question Bank
A structured dataset of JEE Advanced examination questions with full
worked solutions and diagrams. JEE Advanced questions are more analytical
than JEE Main — many are subjective, integer, or numerical-answer type with
detailed multi-step solutions.
Subsets (PCM):
Physics — 50 questions
Chemistry — 21 questions
Mathematics — 48 questions
Structure
Organised into subsets by subject and splits (train / test):
mathematics/ physics/… See the full description on the dataset page: https://huggingface.co/datasets/Grass-G/jee-advanced-questions.jee-grpo-v1
JEE-GRPO-v1: Visual Reasoning Dataset for VLMs
Dataset Summary
JEE-GRPO is a high-quality, multimodal dataset designed for training and benchmarking Vision Language Models (VLMs) on complex STEM problems.
Unlike traditional text-only datasets, this dataset renders JEE Main (Joint Entrance Examination) questions as high-resolution images. This approach bypasses OCR errors and perfectly preserves complex LaTeX equations, diagrams, and chemical structures, making it ideal… See the full description on the dataset page: https://huggingface.co/datasets/farhananis005/jee-grpo-v1.agentfaildb
AgentFailDB
An open, fully-local benchmark of failure modes in multi-agent LLM systems —
750 execution traces from three frameworks (CrewAI, AutoGen,
LangGraph) run on a local Llama 3.1 8B model (via Ollama) at $0 API cost.
Relation to prior work
The canonical, human-validated taxonomy of multi-agent LLM failures is MAST —
Cemri et al., "Why Do Multi-Agent LLM Systems Fail?" (NeurIPS 2025). AgentFailDB does
not supersede or replace it; it is a fully-local… See the full description on the dataset page: https://huggingface.co/datasets/Jeevaa79/agentfaildb.jee-sft-v1
JEE-SFT: Visual Reasoning & Explanation Dataset
Dataset Summary
JEE-SFT is a multimodal instruction-tuning dataset designed to teach Vision Language Models (VLMs) step-by-step reasoning for complex STEM problems.
Unlike the GRPO version (which focuses on the final answer), this dataset focuses on the solution process. It includes both Multiple Choice Questions (MCQs) and Numerical Value Questions.
The key feature of this dataset is its "Text-Only Reasoning" filter: every… See the full description on the dataset page: https://huggingface.co/datasets/farhananis005/jee-sft-v1.Powerlifting
Dataset Card for Powerlifting
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/Jeevak01/Powerlifting/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/Jeevak01/Powerlifting.