datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
AstroCLIPGowerStreetDESY3
Gower Street DES Y3 Lensing Tiles
Weak lensing convergence map tiles extracted from the Gower Street N-body simulation suite, processed through a Born-approximation raytracing pipeline with DES Y3 MagLim source n(z) distributions.
Dataset Description
Each sample contains a (4, H, W) convergence map tile covering ~3400 deg², corresponding to 4 DES Y3 MagLim tomographic bins. Tiles are extracted from equatorial HEALPix base faces after harmonic-space filtering and rotation… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/GowerStreetDESY3.DESISpectra datset from DESI.desi_legacysurvey_xmatch
DESI × Legacy Survey cross-match
95,895 galaxies, each carrying three modalities of the same object:
DESI PROVABGS value-added properties (*_provabgs): redshift Z_HP_provabgs,
stellar mass LOG_MSTAR_provabgs, SFR, MCMC posteriors, magnitudes, ...
DESI spectrum (spectrum): flux / ivar / wavelength / mask / LSF, plus DESI
photometry (*_spec)
Legacy Survey DR10 (DECaLS south) imaging: image (160×160 cutouts in
g,r,i,z with ivar + mask planes), rgb / object_mask / blobmodel PNGs… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/desi_legacysurvey_xmatch.desi_x_ls_aion
AION-1 embeddings of DESI × Legacy Survey galaxies
95,895 galaxies, each carrying token-level AION-1 embeddings of its Legacy Survey image and DESI spectrum (embedded separately), together with its PROVABGS physical properties (redshift, stellar mass, metallicity, age, star-formation rate, …).
The embeddings are ready-made features for downstream tasks — similarity search, property regression, anomaly detection — with no GPU or raw-pixel/spectrum processing required.… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/desi_x_ls_aion.DESI2
Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
Dataset Description
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Dataset Sources [optional]
Repository: [More… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/DESI2.aion2-full-lancedb
AION-2 DESI and Legacy Survey LanceDB
Two complete survey tables containing existing AION-1 tokens, prepared for
AION-2 training and live crossmatching.
The native LanceDB database occupies approximately 10.32 GB (9.61 GiB),
including coordinate indexes. Raw spectra and image pixels are not included.
Table / configuration
Observations
Modality column
Tokens per row
Vocabulary
Lance table version
desi
1,126,441
tok_spectrum_desi
273
1,024
3
legacysurvey
14,174,203… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/aion2-full-lancedb.mmu_ssl_legacysurvey_north_aion_demo
tokenized_demo
AION-1 tokens for hf://datasets/UniverseTBD/mmu_ssl_legacysurvey_north, produced with aion-hats. The catalog keeps the HATS layout of its source, so it can be opened with lsdb.open_catalog(...) and joined back to the source on _healpix_29/object_id, or loaded with datasets.load_dataset(...).
Tokenized columns:
image -> tok_image (LegacySurveyImage)
gaia---
description: 'Spectral (BP/RP), photometric, and astrometric dataset based on Gaia
DR3.
'
homepage: https://www.cosmos.esa.int/web/gaia/dr3
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% If you have used Gaia DR3 data in your research, \ please use the following acknowledgement:\n% \n% This work has made use of data \ from the European Space Agency (ESA) mission\n% {\it Gaia} (\url{https://www.cosmos.esa.int/gaia}),\
\ processed by the {\it Gaia}\n% Data Processing and Analysis… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/gaia.mmu_ssl_legacysurvey_north_aion_tokens_10k
ls_north_tokens_10k
AION-1 tokens for hf://datasets/UniverseTBD/mmu_ssl_legacysurvey_north, produced with aion-hats. The catalog keeps the HATS layout of its source, so it can be opened with lsdb.open_catalog(...) and joined back to the source on _healpix_29/object_id, or loaded with datasets.load_dataset(...).
Tokenized columns:
image -> tok_image (LegacySurveyImage)
ebv -> tok_ebv (LegacySurveyEBV)
flux_g -> tok_flux_g (LegacySurveyFluxG)
flux_r -> tok_flux_r… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/mmu_ssl_legacysurvey_north_aion_tokens_10k.lesubra__ECE-EIFFEL-3B-details
Dataset Card for Evaluation run of lesubra/ECE-EIFFEL-3B
Dataset automatically created during the evaluation run of model lesubra/ECE-EIFFEL-3B
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/lesubra__ECE-EIFFEL-3B-details.legacysurvey---
description: 'Image dataset from Legacy Survey DR10
'
homepage: https://www.legacysurvey.org/dr10/
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% Data Release 10 (DR10) is the tenth public data \ release of the Legacy Surveys.\n% \n% When using data from the Legacy Surveys \ in papers, please use the following acknowledgment:\n% \n% The Legacy Surveys \ consist of three individual and complementary projects: the Dark Energy Camera \ Legacy Survey (DECaLS; Proposal ID #2014B-0404;… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/legacysurvey.EIFFEL
Dataset Card for the EIFFEL benchmark
EIFFEL (Evaluation of Idiomatic French Fixed Expressions for Large Language Models) is a French evaluation dataset designed to assess large language models’ (LLM) knowledge of idiomatic expressions in context. EIFFEL comprises 602 samples in multiple choice format designed to test an LLM's capacity to complete an idiomatic expression in context. The samples have been manually constructed and annotated by native speakers and linguists.… See the full description on the dataset page: https://huggingface.co/datasets/OpenLLM-France/EIFFEL.cfa---
description: Time-series dataset from the Center for Astronomy 3 Data Release.
homepage: https://lweb.cfa.harvard.edu/supernova/
version: 1.0.0
citation: "@ARTICLE{2009ApJ...700..331H,\n author = {{Hicken}, Malcolm and {Challis}, \ Peter and {Jha}, Saurabh and {Kirshner}, Robert P. and {Matheson}, Tom and {Modjaz}, \ Maryam and {Rest}, Armin and {Wood-Vasey}, W. Michael and {Bakos}, Gaspar and \ {Barton}, Elizabeth J. and {Berlind}, Perry and {Bragg}, Ann and {Brice{\n}o}, \ Cesar… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/cfa.btsbot---
description: 'This is the production version of the BTSbot training set, limited to
public (programid=1) ZTF alerts.
Original codebase: https://github.com/nabeelre/BTSbot
'
homepage: https://zenodo.org/records/10839691
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% % Based on the Acknowledgements in Rehemtulla et \ al. (2024). We suggest including a variant of the following in your acknowledgements:\n % A great number of people have contributed to BTS and BTS scanning over the… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/btsbot.desi_provabgs---
description: 'The PROVABGS catalog from Hahn, et al. (2022) is a catalog of galaxy
properties derived using a state-of-the-art SED modeling of DESI
spectroscopy and photometry. More details are located here: https://github.com/changhoonhahn/provabgs
'
homepage: https://changhoonhahn.github.io/provabgs/current/
version: 1.1.0
citation: "% % ACKNOWLEDGEMENTS\n% From https://github.com/changhoonhahn/provabgs\
\ and https://arxiv.org/abs/2202.01809 :\n% \n% This research is supported… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/desi_provabgs.tess---
description: 'TESS Light Curves From Full Frame Images ("TESS-SPOC")
'
homepage: https://archive.stsci.edu/hlsp/tess-spoc
version: 0.0.1
citation: "% % ACKNOWLEDGEMENT\n% % From: https://archive.stsci.edu/publishing/mission-acknowledgements\n\
% This paper includes data collected with the TESS mission, obtained from the MAST \ data archive at the Space Telescope Science Institute (STScI). Funding for the \ TESS mission is provided by the NASA Explorer Program. STScI is operated by… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/tess.gz10---
description: 'The GZ10 catalog from Leung et al. (2018) is a dataset of 17,736 galaxies
with labels from the Galaxy Zoo 2 project.
The catalog includes the following features for each galaxy: right ascension, declination,
redshift, and a label from the Galaxy Zoo 2 project.
'
homepage: https://astronn.readthedocs.io/en/latest/galaxy10.html
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% From https://astronn.readthedocs.io/en/stable/galaxy10.html#acknowledgments\n\
% \n% The… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/gz10.neurips-wl-challenge
FAIR Universe - NeurIPS 2025 Weak Lensing Uncertainty Challenge
This dataset is a HF mirror of the official challenge training data for this challenge:
https://www.codabench.org/competitions/8934/
No further processing has been applied to it, it is purely intended to simplify access to the challenge.
To get started:
import datasets
dset = datasets.load_dataset("eiffl/neurips-wl-challenge")
dset = dset.with_format('jax')
example = dset['train'][0]
EIF-Manipulated-distributions
Exposing the Illusion of Fairness (EIF) Manipulations Results
We consider use cases where the auditee has developed a model which has fairness issues.
It tries to hide the problem by picking a subsample while optimizing the fairness metric that will be computed by the auditor.
Yet, from the supervisory authority, submitting a non-representative sample constitute a deceptive attempt by the auditee to obstruct or distort the assessment.
We present here the original empirical… See the full description on the dataset page: https://huggingface.co/datasets/ValentinLAFARGUE/EIF-Manipulated-distributions.manga---
description: '
An IFU dataset from the SDSS-IV MaNGA survey, a wide-field, optical, IFU survey
of ~10,000
nearby galaxies. This dataset contains the following data products for each galaxy:
the 3D data cubes,
and reconstructed griz images from the MaNGA Data Reduction Pipeline (DRP), and
all the derived
analsysis maps from the MaNGA Data Analyis Pipeline (DAP).
'
homepage: https://www.sdss4.org/dr17/manga/
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% % From:… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/manga.ssl_legacysurvey---
description: 'Image dataset based on Legacy Survey stamps assembled by George Stein
'
homepage: https://github.com/georgestein/ssl-legacysurvey
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% % From: https://www.legacysurvey.org/\n% \n% When \ using data from the Legacy Surveys in papers, please use the following acknowledgment:\n % \n% The Legacy Surveys consist of three individual and complementary projects: \ the Dark Energy Camera Legacy Survey (DECaLS; Proposal ID #2014B-0404;… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/ssl_legacysurvey.swift_sne_ia---
description: '
Time-series dataset from Swift SNe Ia.
'
homepage: https://pbrown801.github.io/SOUSA
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% % From: https://archive.stsci.edu/prepds/sousa/\n\
% \n% Suggestion of text to add to "Observations" section of papers that use SOUSA \ data:\n% \n% This supernova was also observed in the UV with the Ultra-Violet/Optical \ Telescope (UVOT; Roming et al. (2005)) on the Swift spacecraft (Gehrels et al. \ 2004). The UV photometry was… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/swift_sne_ia.llama-3.1-8b-instruct_eiffel_towerCreated for: https://github.com/shawonashraf/drrik
chandra---
description: 'Spectra from the Chandra Source Catalog. Processed from pulse height
amplitude (PHA) files modified by instrumental response files ARF and RMF.
'
homepage: https://cxc.cfa.harvard.edu/csc/
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% From https://cxc.cfa.harvard.edu/csc/cite.html\
\ :\n% \n% Users are kindly requested to acknowledge their use of the Chandra Source \ Catalog in any resulting publications.\n% \n% This will help us greatly to keep \ track of… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/chandra.snls---
description: 'Time-series dataset from the Supernova Legacy Survey (SNLS).
Data Citations:
Guy et al. (2010)
'
homepage: https://www.aanda.org/articles/aa/full_html/2010/15/aa14468-10/aa14468-10.html
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% The SNLS is an International Collaboration of physicists \ and astronomers from various institutions in Canada, EU and US.\n% \n% Institution/group \ representatives (Collaboration Board) are: P. Astier (IN2P3/LPNHE, Fr), S. Basa \… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/snls.plasticc---
description: 'The Photometric LSST Astronomical Time-Series Classification Challenge
(PLAsTiCC) is a community-wide challenge to spur development of algorithms to classify
astronomical transients. The Large Synoptic Survey Telescope (LSST) will discover
tens of thousands of transient phenomena every single night. To deal with this massive
onset of data, automated algorithms to classify and sort astronomical transients
are crucial.
'
homepage: https://zenodo.org/records/2539456… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/plasticc.foundation---
description: 'Time-series dataset from Foundation Data Release 1 (Foundation DR1).
Data Citations:
Foley et al. (2018) - https://ui.adsabs.harvard.edu/abs/2018MNRAS.475..193F
Jones et al. (2019) - https://ui.adsabs.harvard.edu/abs/2019ApJ...881...19J
'
homepage: https://github.com/djones1040/Foundation_DR1/tree/master
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% When using these data, please cite:\n% \n% Foley \ et al. (2018) -… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/foundation.hsc---
description: 'Image dataset based on HSC SSP PRD3.
'
homepage: https://hsc-release.mtk.nao.ac.jp/doc/
version: 0.0.1
citation: "% CITATION\n@article{Aihara_2017,\n title={The Hyper Suprime-Cam SSP \ Survey: Overview and survey design},\n volume={70},\n ISSN={2053-051X},\n \ url={http://dx.doi.org/10.1093/pasj/psx066},\n DOI={10.1093/pasj/psx066},\n \ number={SP1},\n journal={Publications of the Astronomical Society of Japan},\n \ publisher={Oxford University Press… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/hsc.csp---
description: 'Time-series dataset from the Carnegie Supernova Project I Data Release
3 (CSP-I DR3).
'
homepage: https://csp.obs.carnegiescience.edu/
version: 1.0.0
citation: "% % ACKNOWLEDGEMENTS\n% % Based on: https://csp.obs.carnegiescience.edu/\n\
% The Carnegie Supernova Project is based upon work supported by the National Science \ Foundation under Grants AST-0306969, AST-0607438, AST-1008343, AST-1613426, AST-1613455, \ and AST-161347.\n% \n%… See the full description on the dataset page: https://huggingface.co/datasets/EiffL/csp.