datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
esg_reports_v2
Vidore Benchmark 2 - ESG Restaurant Dataset (Multilingual)
This dataset is part of the "Vidore Benchmark 2" collection, designed for evaluating visual retrieval applications. It focuses on the theme of ESG reports in the fast food industry.
Dataset Summary
The dataset contain queries in the following languages : ["english", "french", "german", "spanish"]. Each query was originaly in "french" (see… See the full description on the dataset page: https://huggingface.co/datasets/vidore/esg_reports_v2.esg_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.ESG_Report
ESG Report PDF Dataset
Download Instructions
To download the dataset, follow these steps:
Navigate to the data directory in the GitHub repository:
cd data
Install Git LFS (if not already installed):
git lfs install
Clone the dataset from Hugging Face Hub:
git clone https://huggingface.co/datasets/WHATX/ESG_Report
Dataset Description
This dataset contains three main components:
raw_pdf:
A collection of 195 PDFs scraped from TCFD Hub.
The PDFs are… See the full description on the dataset page: https://huggingface.co/datasets/WHATX/ESG_Report.ESGESGReports
ESGReports
An MTEB dataset
Massive Text Embedding Benchmark
LMEB semantic retrieval task based on ESG-Reports, retrieving long-form report passages relevant to environmental, social, and governance questions.
Task category
Retrieval (text-to-text)
DomainsFinancial, Written
Reference
LMEB: Long-horizon Memory Embedding Benchmark
Source datasets:
KaLM-Embedding/LMEB
How to evaluate on this task
You can evaluate an embedding model on this dataset using… See the full description on the dataset page: https://huggingface.co/datasets/mteb/ESGReports.es-grpo-results
ES vs GRPO: Evaluation Results
Overview
This repository contains comprehensive evaluation results comparing Evolution Strategies (ES) and Group Relative Policy Optimization (GRPO) on mathematical reasoning tasks.
Tasks
GSM8K: Grade school math word problems
Countdown: Arithmetic equation generation puzzle
Models
Qwen2.5-3B-Instruct
Qwen2.5-3B-Base (with custom tokenizer)
Llama-3.2-3B-Instruct
Llama-3.2-3B-Base (with custom tokenizer)
File… See the full description on the dataset page: https://huggingface.co/datasets/alphaXiv/es-grpo-results.esg_reports_eng_v2
Vidore Benchmark 2 - ESG Restaurant Dataset
This dataset is part of the "Vidore Benchmark 2" collection, designed for evaluating visual retrieval applications. It focuses on the theme of ESG reports in the fast food industry.
Dataset Summary
Each query is in french.
This dataset provides a focused benchmark for visual retrieval tasks related to ESG reports of fast food companies. 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_eng_v2.esg_cid_retrieval
Enhancing Retrieval for ESGLLM via ESG-CID -- A Disclosure Content Index Finetuning Dataset for Mapping GRI and ESRS
Usage
from datasets import load_dataset
# document chunks: train/dev/test_gri/test_esrs
documents = load_dataset("esgllm/esg_cid_retrieval", "documents")
# queries (disclosure text): train/dev/test_gri/test_esrs
queries = load_dataset("esgllm/esg_cid_retrieval", "queries")
# training triplets: train/dev
triplets = load_dataset("esgllm/esg_cid_retrieval"… See the full description on the dataset page: https://huggingface.co/datasets/airefinery/esg_cid_retrieval.embtb-esg-evalesg-reports
ConTEB - ESG Reports
This dataset is part of ConTEB (Context-aware Text Embedding Benchmark), designed for evaluating contextual embedding model capabilities. It focuses on the theme of Industrial ESG Reports, particularly stemming from the fast-food industry.
Dataset Summary
This dataset was designed to elicit contextual information. It is built upon a subset of the ViDoRe Benchmark. To build the corpus, we start from the pre-existing collection of ESG Reports, extract… See the full description on the dataset page: https://huggingface.co/datasets/illuin-conteb/esg-reports.alistairking_public-company-esg-ratings-dataset
Public Company ESG Ratings Dataset
ESG ratings for over 700 mid / large-cap companies across various industries
Dataset Info
Source: Kaggle
Original Size: 0.04 MB
Kaggle Downloads: 5,844
Files: 1
Files
data.csv
Mirrored from Kaggle
environmental_2kgovernance_2ksocial_2kfortune500-esg-metrics-2021-2023
Fortune 500 ESG Metrics Dataset (2021-2023)
🌍 Dataset Description
This comprehensive dataset contains Environmental, Social, and Governance (ESG) metrics from Fortune 500 companies spanning 2021-2023. It represents one of the most extensive collections of corporate sustainability data publicly available, compiled from official corporate reports, sustainability disclosures, and ESG filings.
🎯 Key Features
📊 Extensive Coverage: Fortune 500 companies… See the full description on the dataset page: https://huggingface.co/datasets/GemiAI2025/fortune500-esg-metrics-2021-2023.ESGenius
ESGenius
ESGenius is an EMNLP 2025 Main Conference Oral benchmark for evaluating large language models on Environmental, Social, and Governance (ESG) and sustainability knowledge. The paper was nominated for the EMNLP 2025 Resource and Theme Paper Awards, Top 1%.
Paper: https://aclanthology.org/2025.emnlp-main.739/
Project site: https://angel-ntu.github.io/ESGenius/
GitHub repository: https://github.com/ANGEL-NTU/ESGenius
Interactive heatmap:… See the full description on the dataset page: https://huggingface.co/datasets/cy0307/ESGenius.esg-assistant
Dataset Card for my-distiset-fef3f4db
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/yasserrmd/my-distiset-fef3f4db/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/yasserrmd/esg-assistant.esg_cid
ESG-CID: A Content Index Dataset for ESG Reports
Usage
from datasets import load_dataset
# Load the content index annotations
cid_dataset = load_dataset("airefinery/esg_cid", "cid")
# Load the fixed version with normalized standard_ids
cid_fixed = load_dataset("airefinery/esg_cid", "cid_fixed")
# Load PDF source data
cid_pdfs = load_dataset("airefinery/esg_cid", "pdfs")
Working with PDFs
The PDF bytes are stored in ["pdf"]["bytes"]. Here's how to use them… See the full description on the dataset page: https://huggingface.co/datasets/airefinery/esg_cid.spx-sustainalytics-esg-scoresesg-bank
🏦 ESG-Bank v1.0
A high-quality, human-annotated dataset for Environmental, Social, and Governance (ESG) sentiment analysis.This resource provides over 4,700 sentences annotated for sentiment at the phrase level, designed to establish benchmark standards for ESG-related text modeling.
📑 Contents
Introduction
Data
Acknowledgements
Contact Information
References and Citation
1. Introduction
Many statistical and machine learning methods… See the full description on the dataset page: https://huggingface.co/datasets/TUM-TFAI/esg-bank.esg-extractor-design-and-code
ESG Metric Extractor — Design & Code Package
Two files, both copy-paste ready:
File
Contents
DESIGN.md
Full design: task framing, multimodal architecture, model choices with 2026 costs, data strategy, training config (TRL-grounded), evaluation, risks, roadmap
CODE.md
All runnable Colab cells: Part A = v1 text-only pipeline (Qwen2.5-3B QLoRA, data prep, training, eval); Part B = v2 multimodal pipeline (Qwen3-VL-4B QLoRA on page images, teacher labeling, PDF pipeline)… See the full description on the dataset page: https://huggingface.co/datasets/Siva2022/esg-extractor-design-and-code.premium-financial-esg-matrix-sample🔥 GET THE FULL 500 PREMIUM DATASET MASTER MATRIX HERE: https://abdulqr.gumroad.com/l/premium-financial-esg-matrix
🚀 Boost Your Financial AI Models with 500 Premium Clean Datasets
Training AI for Fintech and compliance requires high-quality, structured data. This premium bundle delivers a master matrix of 500 clean, synthetic financial and ESG data entries—specifically tailored for LLM fine-tuning, fraud detection simulation, and predictive analytics.
Perfect for Data Scientists, AI… See the full description on the dataset page: https://huggingface.co/datasets/abdulqr/premium-financial-esg-matrix-sample.ESG-DLT-NER
Dataset Card for ESG/DLT Named Entity Recognition Dataset
This dataset contains named entities related to Distributed Ledger Technology (DLT) and Environmental, Social, and Governance (ESG) topics created to support research in these areas and at the intersection of these domains.
Dataset Details
Dataset Sources
Repository: https://github.com/dlt-science/ESG-DLT-LitReview
Paper: https://arxiv.org/abs/2308.12420
Use
This dataset can be used for… See the full description on the dataset page: https://huggingface.co/datasets/ExponentialScience/ESG-DLT-NER.base_dataesg_bank
🏦 ESGBank v1.0
A high-quality, human-annotated dataset for Environmental, Social, and Governance (ESG) sentiment analysis.This resource provides over 4,800 sentences annotated for sentiment at the phrase level, designed to establish benchmark standards for ESG-related text modeling.
📑 Contents
Introduction
Data
Acknowledgements
Contact Information
References and Citation
1. Introduction
Many statistical and machine learning methods require… See the full description on the dataset page: https://huggingface.co/datasets/astrohumanoid/esg_bank.environment_dataNER_sasb_esg
Dataset Card for "NER_sasb_esg"
More Information needed
esg-extractionesg-spanish-events
ESG Spanish Events (2014–2024)
Dataset summary
Multi-label ESG classification dataset for Spanish equity market news
(2014–2024). Contains 1,688 human-annotated canonical news events
(gold set) and 64,488 LLaMA-3.1-8B SFT silver-annotated events
(silver set). Labels cover three ESG pillars (Environmental, Social,
Governance) as independent binary signals plus a 4-class sentiment
dimension (Positive / Negative / Neutral / NA).
Companion model: DReggio/mrbert-es-esg… See the full description on the dataset page: https://huggingface.co/datasets/DReggio/esg-spanish-events.es-gl_bensculturais_parallel
Spanish-Galician Cultural Heritage Terms and Definitions Parallel Corpus
Dataset Description
es-gl_bensculturais_parallel is a Spanish-Galician parallel corpus focused on cultural heritage terminology and definitions. It contains aligned Spanish and Galician text files derived from cultural heritage terminology resources.
The dataset is intended to support machine translation, terminology adaptation, domain-specific language modeling, and the development of… See the full description on the dataset page: https://huggingface.co/datasets/proxectonos/es-gl_bensculturais_parallel.