Dataset

This dataset was created by William-2777

Released under CC0: Public Domain

Contents

Subset of the SMDG-19 for Glaucoma dataset in PyTorch Format

SMDG-19: https://www.kaggle.com/datasets/deathtrooper/multichannel-glaucoma-benchmark-dataset

Contains Train, Val and Test set of Fundus images for Glaucoma Detection

2 Classes (0|1)

1: Glaucoma Present 0: Glaucoma not Present

Traffic analytics, rankings, and competitive metrics for pytorch.org as of October 2025

This data set contains information about microstructure-sensitive damage in ferritic steel (EN 1.4003) specimens. We provide both graph and tabular data sets. In the graph data set, each node is a grain in the material, and each edge connects adjacent grains. Annotation is provided to each of the grains mentioning if it developed damage (protrusion in this case) during a fatigue test. The annotation is provided based on fatigue tests performed on the specimen. Technical Information: The readme.md file provides an example on how to read the file.

Abstract

In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before.

Dataset

This dataset is part of a larger collection of model zoos and contains the zoos trained on the labelled samples from MNIST. All zoos with extensive information and code can be found at www.modelzoos.cc.

This repository contains two types of files: the raw model zoos as collections of models (file names beginning with "mnist_"), as well as preprocessed model zoos wrapped in a custom pytorch dataset class (filenames beginning with "dataset"). Zoos are trained in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). The index_dict.json files contain information on how to read the vectorized models.

For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

Dataset

This dataset was created by Luong Hoang Minh

Contents

This is the dataset used for pre-training in "ReasonBERT: Pre-trained to Reason with Distant Supervision", EMNLP'21.

There are two files:

sentence_pairs_for_pretrain_no_tokenization.tar.gz -> Contain only sentences as evidence, Text-only

table_pairs_for_pretrain_no_tokenization.tar.gz -> At least one piece of evidence is a table, Hybrid

The data is chunked into multiple tar files for easy loading. We use WebDataset, a PyTorch Dataset (IterableDataset) implementation providing efficient sequential/streaming data access.

For pre-training code, or if you have any questions, please check our GitHub repo https://github.com/sunlab-osu/ReasonBERT

Below is a sample code snippet to load the data

import webdataset as wds

path to the uncompressed files, should be a directory with a set of tar files

url = './sentence_multi_pairs_for_pretrain_no_tokenization/{000000...000763}.tar' dataset = ( wds.Dataset(url) .shuffle(1000) # cache 1000 samples and shuffle .decode() .to_tuple("json") .batched(20) # group every 20 examples into a batch )

Please see the documentation for WebDataset for more details about how to use it as dataloader for Pytorch

You can also iterate through all examples and dump them with your preferred data format

Below we show how the data is organized with two examples.

Text-only

{'s1_text': 'Sils is a municipality in the comarca of Selva, in Catalonia, Spain.', # query sentence 's1_all_links': { 'Sils,_Girona': [[0, 4]], 'municipality': [[10, 22]], 'Comarques_of_Catalonia': [[30, 37]], 'Selva': [[41, 46]], 'Catalonia': [[51, 60]] }, # list of entities and their mentions in the sentence (start, end location) 'pairs': [ # other sentences that share common entity pair with the query, group by shared entity pairs { 'pair': ['Comarques_of_Catalonia', 'Selva'], # the common entity pair 's1_pair_locs': [[[30, 37]], [[41, 46]]], # mention of the entity pair in the query 's2s': [ # list of other sentences that contain the common entity pair, or evidence { 'md5': '2777e32bddd6ec414f0bc7a0b7fea331', 'text': 'Selva is a coastal comarque (county) in Catalonia, Spain, located between the mountain range known as the Serralada Transversal or Puigsacalm and the Costa Brava (part of the Mediterranean coast). Unusually, it is divided between the provinces of Girona and Barcelona, with Fogars de la Selva being part of Barcelona province and all other municipalities falling inside Girona province. Also unusually, its capital, Santa Coloma de Farners, is no longer among its larger municipalities, with the coastal towns of Blanes and Lloret de Mar having far surpassed it in size.', 's_loc': [0, 27], # in addition to the sentence containing the common entity pair, we also keep its surrounding context. 's_loc' is the start/end location of the actual evidence sentence 'pair_locs': [ # mentions of the entity pair in the evidence [[19, 27]], # mentions of entity 1 [[0, 5], [288, 293]] # mentions of entity 2 ], 'all_links': { 'Selva': [[0, 5], [288, 293]], 'Comarques_of_Catalonia': [[19, 27]], 'Catalonia': [[40, 49]] } } ,...] # there are multiple evidence sentences }, ,...] # there are multiple entity pairs in the query }

Hybrid

{'s1_text': 'The 2006 Major League Baseball All-Star Game was the 77th playing of the midseason exhibition baseball game between the all-stars of the American League (AL) and National League (NL), the two leagues comprising Major League Baseball.', 's1_all_links': {...}, # same as text-only 'sentence_pairs': [{'pair': ..., 's1_pair_locs': ..., 's2s': [...]}], # same as text-only 'table_pairs': [ 'tid': 'Major_League_Baseball-1', 'text':[ ['World Series Records', 'World Series Records', ...], ['Team', 'Number of Series won', ...], ['St. Louis Cardinals (NL)', '11', ...], ...] # table content, list of rows 'index':[ [[0, 0], [0, 1], ...], [[1, 0], [1, 1], ...], ...] # index of each cell [row_id, col_id]. we keep only a table snippet, but the index here is from the original table. 'value_ranks':[ [0, 0, ...], [0, 0, ...], [0, 10, ...], ...] # if the cell contain numeric value/date, this is its rank ordered from small to large, follow TAPAS 'value_inv_ranks': [], # inverse rank 'all_links':{ 'St._Louis_Cardinals': { '2': [ [[2, 0], [0, 19]], # [[row_id, col_id], [start, end]] ] # list of mentions in the second row, the key is row_id }, 'CARDINAL:11': {'2': [[[2, 1], [0, 2]]], '8': [[[8, 3], [0, 2]]]}, } 'name': '', # table name, if exists 'pairs': { 'pair': ['American_League', 'National_League'], 's1_pair_locs': [[[137, 152]], [[162, 177]]], # mention in the query 'table_pair_locs': { '17': [ # mention of entity pair in row 17 [ [[17, 0], [3, 18]], [[17, 1], [3, 18]], [[17, 2], [3, 18]], [[17, 3], [3, 18]] ], # mention of the first entity [ [[17, 0], [21, 36]], [[17, 1], [21, 36]], ] # mention of the second entity ] } } ] }

Samples in this benchmark were generated by RELAI using the following data source(s): Data Source Name: pytorch Data Source Link: https://pytorch.org/docs/stable/index.html Data Source License: https://github.com/pytorch/pytorch/blob/main/LICENSE Data Source Authors: PyTorch AI Benchmarks by Data Agents. 2025 RELAI.AI. Licensed under CC BY 4.0. Source: https://relai.ai

Acute poisoning is a significant global health burden, and the causative agent is often unclear. The primary aim of this pilot study was to develop a deep learning algorithm that predicts the most probable agent a poisoned patient was exposed to from a pre-specified list of drugs. Data were queried from the National Poison Data System (NPDS) from 2014 through 2018 for eight single-agent poisonings (acetaminophen, diphenhydramine, aspirin, calcium channel blockers, sulfonylureas, benzodiazepines, bupropion, and lithium). Two Deep Neural Networks (PyTorch and Keras) designed for multi-class classification tasks were applied. There were 201,031 single-agent poisonings included in the analysis. For distinguishing among selected poisonings, PyTorch model had specificity of 97%, accuracy of 83%, precision of 83%, recall of 83%, and a F1-score of 82%. Keras had specificity of 98%, accuracy of 83%, precision of 84%, recall of 83%, and a F1-score of 83%. The best performance was achieved in the diagnosis of single-agent poisoning in diagnosing poisoning by lithium, sulfonylureas, diphenhydramine, calcium channel blockers, then acetaminophen, in PyTorch (F1-score = 99%, 94%, 85%, 83%, and 82%, respectively) and Keras (F1-score = 99%, 94%, 86%, 82%, and 82%, respectively). Deep neural networks can potentially help in distinguishing the causative agent of acute poisoning. This study used a small list of drugs, with polysubstance ingestions excluded.Reproducible source code and results can be obtained at https://github.com/ashiskb/npds-workspace.git.

The SloNER is a model for Slovenian Named Entity Recognition. It is is a PyTorch neural network model, intended for usage with the HuggingFace transformers library (https://github.com/huggingface/transformers).

The model is based on the Slovenian RoBERTa contextual embeddings model SloBERTa 2.0 (http://hdl.handle.net/11356/1397). The model was trained on the SUK 1.0 training corpus (http://hdl.handle.net/11356/1747).The source code of the model is available on GitHub repository https://github.com/clarinsi/SloNER.

A dataset containing a sample event inspired by ProtoDUNE-SP simulation. Checkpoints of trained DUNEdn package models used for Springer original article.

DataSet for training the PyTorch Graph Network Simulator. https://github.com/geoelements/gns. The repository contains the data sets for water drop sample

Slovenian model for coreference resolution: a neural network based on a customized transformer architecture, usable with the code published on https://github.com/matejklemen/slovene-coreference-resolution. The model is based on the Slovenian CroSloEngual BERT 1.1 model (http://hdl.handle.net/11356/1330). It was trained on the SUK 1.0 training corpus (http://hdl.handle.net/11356/1747), specifically the SentiCoref subcorpus.

Using the evaluation setting where entity mentions are assumed to be correctly pre-detected, the model achieves the following metric values: MUC: precision = 0.931, recall = 0.957, F1 = 0.943 BCubed: precision = 0.887, recall = 0.947, F1 = 0.914 CEAFe: precision = 0.945, recall = 0.893, F1 = 0.916 CoNLL-12: precision = 0.921, recall = 0.932, F1 = 0.924

Methods

Cotton plants were grown in a well-controlled greenhouse in the NC State Phytotron as described previously (Pierce et al, 2019). Flowers were tagged on the day of anthesis and harvested three days post anthesis (3 DPA). The distinct fiber shapes had already formed by 2 DPA (Stiff and Haigler, 2016; Graham and Haigler, 2021), and fibers were still relatively short at 3 DPA, which facilitated the visualization of multiple fiber tips in one image.

Cotton fiber sample preparation, digital image collection, and image analysis:

Ovules with attached fiber were fixed in the greenhouse. The fixative previously used (Histochoice) (Stiff and Haigler, 2016; Pierce et al., 2019; Graham and Haigler, 2021) is obsolete, which led to testing and validation of another low-toxicity, formalin-free fixative (#A5472; Sigma-Aldrich, St. Louis, MO; Fig. S1). The boll wall was removed without damaging the ovules. (Using a razor blade, cut away the top 3 mm of the boll. Make about 1 mm deep longitudinal incisions between the locule walls, and finally cut around the base of the boll.) All of the ovules with attached fiber were lifted out of the locules and fixed (1 h, RT, 1:10 tissue:fixative ratio) prior to optional storage at 4°C. Immediately before imaging, ovules were examined under a stereo microscope (incident light, black background, 31X) to select three vigorous ovules from each boll while avoiding drying. Ovules were rinsed (3 x 5 min) in buffer [0.05 M PIPES, 12 mM EGTA. 5 mM EDTA and 0.1% (w/v) Tween 80, pH 6.8], which had lower osmolarity than a microtubule-stabilizing buffer used previously for aldehyde-fixed fibers (Seagull, 1990; Graham and Haigler, 2021). While steadying an ovule with forceps, one to three small pieces of its chalazal end with attached fibers were dissected away using a small knife (#10055-12; Fine Science Tools, Foster City, CA). Each ovule piece was placed in a single well of a 24-well slide (#63430-04; Electron Microscopy Sciences, Hatfield, PA) containing a single drop of buffer prior to applying and sealing a 24 x 60 mm coverslip with vaseline.

Samples were imaged with brightfield optics and default settings for the 2.83 mega-pixel, color, CCD camera of the Keyence BZ-X810 imaging system (www.keyence.com; housed in the Cellular and Molecular Imaging Facility of NC State). The location of each sample in the 24-well slides was identified visually using a 2X objective and mapped using the navigation function of the integrated Keyence software. Using the 10X objective lens (plan-apochromatic; NA 0.45) and 60% closed condenser aperture setting, a region with many fiber apices was selected for imaging using the multi-point and z-stack capture functions. The precise location was recorded by the software prior to visual setting of the limits of the z-plane range (1.2 µm step size). Typically, three 24-sample slides (representing three accessions) were set up in parallel prior to automatic image capture. The captured z-stacks for each sample were processed into one two-dimensional image using the full-focus function of the software. (Occasional samples contained too much debris for computer vision to be effective, and these were reimaged.)

• Resource Title: Deltapine 90 - Manually Annotated Training Set.

  File Name: GH3 DP90 Keyence 1_45 JPEG.zip

  Resource Description: These images were manually annotated in Labelbox.




• Resource Title: Deltapine 90 - AI-Assisted Annotated Training Set.

  File Name: GH3 DP90 Keyence 46_101 JPEG.zip

  Resource Description: These images were AI-labeled in RoboFlow and then manually reviewed in RoboFlow.




• Resource Title: Deltapine 90 - Manually Annotated Training-Validation Set.

  File Name: GH3 DP90 Keyence 102_125 JPEG.zip

  Resource Description: These images were manually labeled in LabelBox, and then used for training-validation for the machine learning model.




• Resource Title: Phytogen 800 - Evaluation Test Images.

  File Name: Gb cv Phytogen 800.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Pima 3-79 - Evaluation Test Images.

  File Name: Gb cv Pima 379.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Pima S-7 - Evaluation Test Images.

  File Name: Gb cv Pima S7.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Coker 312 - Evaluation Test Images.

  File Name: Gh cv Coker 312.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Deltapine 90 - Evaluation Test Images.

  File Name: Gh cv Deltapine 90.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Half and Half - Evaluation Test Images.

  File Name: Gh cv Half and Half.zip

  Resource Description: These images were used to validate the machine learning model. They were manually annotated in ImageJ.




• Resource Title: Fiber Tip Annotations - Manual.

  File Name: manual_annotations.coco_.json

  Resource Description: Annotations in COCO.json format for fibers. Manually annotated in Labelbox.




• Resource Title: Fiber Tip Annotations - AI-Assisted.

  File Name: ai_assisted_annotations.coco_.json

  Resource Description: Annotations in COCO.json format for fibers. AI annotated with human review in Roboflow.




• Resource Title: Model Weights (iteration 600).

  File Name: model_weights.zip

  Resource Description: The final model, provided as a zipped Pytorch .pth file. It was chosen at training iteration 600. The model weights can be imported for use of the fiber tip type detection neural network in Python.

  Resource Software Recommended: Google Colab,url: https://research.google.com/colaboratory/

Dataset

This dataset was created by Richard Luo

Contents

Performance comparison data for MatLogica AADC against JAX, PyTorch, and TensorFlow on quantitative finance workloads

These datasets are customized Torch Geometric Datasets that contain raw .off polygon meshes as well as preprocessed .pt files needed for training morphVQ models. morphVQ can be found at https://github.com/oothomas/morphVQ.

Data supplement: Detection of Drainage Ditches from LiDAR DTM Using U-Net and Transfer Learning

Holger Virro, Alexander Kmoch, William Lidberg, Wai Tik Chan, Evelyn Uuemaa

Accurate mapping of ditches is essential for effective hydrological modeling and land management. Traditional methods, such as manual digitization or threshold-based extraction, utilize LiDAR-derived digital terrain model (DTM) data but are labor-intensive and impractical to apply for large-scale applications. Deep learning offers a promising alternative but requires extensive labeled data, often unavailable. To address this, we developed a transfer learning approach using a U-Net model pre-trained on a large high-quality Swedish dataset and fine-tuned on a smaller localized Estonian dataset. The model uses a single-band LiDAR DTM raster as input, minimizing preprocessing. We identified the optimal model configuration by systematically testing kernel sizes and data augmentation. The best fine-tuned model achieved an overall F1 score of 0.766, demonstrating its effectiveness in detecting drainage ditches in training data-scarce regions. Performance varied by land use, with higher accuracy in peatlands (F1=0.822) than in forests (F1=0.752) and arable land (F1=0.779). These findings underscore the model's suitability for large-scale ditch mapping and its adaptability to different landscapes.

pytorch-image-models metrics

This dataset contains metrics about the huggingface/pytorch-image-models package. Number of repositories in the dataset: 3615 Number of packages in the dataset: 89

  Package dependents

This contains the data available in the used-by tab on GitHub.

  Package & Repository star count

This section shows the package and repository star count, individually.

Package Repository

There are 18 packages that have more than 1000… See the full description on the dataset page: https://huggingface.co/datasets/open-source-metrics/pytorch-image-models-dependents.

Dataset

This dataset was created by Tarun Bisht

Contents