Loading...
Loading...
Student performance, MOOC logs, knowledge tracing, standardized tests, learning analytics
13,913 datasets
Tm Data is a text corpus used to train the TinyMistral-248m-v2 language model, containing approximately 8 million examples. The dataset was created by Locutusque and last updated on October 15, 2024. It consists of 4 million Wikipedia pages, 1 million arXiv papers, 1.5 million web pages from RefinedWeb and SlimPajama, 200,000 college textbooks, and 1 million Stack Exchange forum posts.
Stanford Encyclopedia Philosophy is a dataset associated with the Teeny-Tiny Castle repository for AI Ethics and Safety educational tools. The dataset was uploaded by AiresPucrs to Hugging Face on October 13, 2024. Its specific content and size are not detailed in the provided metadata.
CollegeEnrollment12m 20162017 byStateOfMichigan 20190722 tracks public high school students from Michigan who graduated in 2017 and enrolled in college within 12 months. The dataset, provided by the City of Ferndale, Michigan, distinguishes enrollment between two-year and four-year higher education institutions. It was last updated in September 2024.
12.4k question-answer records were created by combining 7 source datasets for machine learning, data science, and AI. The dataset was used to test fine-tuning the Gemma 2 model by MLX on Apple Silicon. It was created by win-wang and last updated on 2024-09-25.
Ro GSM8K is a Romanian translation of the GSM8K benchmark, containing 8.5 thousand grade school math word problems. The dataset was translated with Systran and is authored by OpenLLM-Ro. It was last updated on September 25, 2024, and serves as part of the evaluation protocol for Romanian large language models proposed in the research paper "Vorbeşti Româneşte?" A Recipe to Train Powerful Romanian LLMs with English Instructions.
Diffgram provides an AI datastore and annotation framework for managing schemas, BLOBs, and model predictions, developed by Diffgram and updated in November 2024. The system facilitates human-in-the-loop supervision and data workflows for image and video processing.
This repository contains raw datasets formatted for training text embedding models, such as with the sentence-transformers library. The data is provided by the sentence-transformers organization and was last updated in September 2024.
The dataset contains the queue of children for preschool educational institutions in the Volodymyr-Volynska hromada. It includes the number, application date, consideration status, child information, and preschool institution details. The data was published on the States site of Ukraine and last updated on September 23, 2024.
2017 data tracks college enrollment rates for Michigan public high school graduates within 12 months of graduation, segmented by the state's congressional districts. The dataset includes enrollment counts for two-year and four-year higher education institutions, compiled by the City of Ferndale, Michigan.
MRC Psycholinguistic Database is a dataset for English word concreteness, hosted on Hugging Face by StephanAkkerman. The dataset was last updated on 2024-09-29. It is made available under the MIT license.
School Learning Modalities, 2021-2022 provides weekly estimates of learning modality (in-person, remote, or hybrid) for U.S. K-12 public and independent charter school districts. The dataset was modeled by combining data from Burbio, MCH Strategic Data, the AEI/Return to Learn Tracker, and state dashboards using a Hidden Markov model. It covers the period from August 2021 to December 2022, with district metadata sourced from the National Center for Educational Statistics (NCES).
Department of Education data on Title I Part A program performance, participation, and funding. The dataset covers school years from 2010-11 onward. It includes state, district, and national-level information on elementary and secondary education.
Third grade English Language Arts and Math test results for the 2016-2017 school year aggregated to census tracts for Michigan. Data Driven Detroit obtained school-level results from MI School Data and aggregated them, suppressing data where fewer than five students were tested per school. The dataset is available in multiple formats including CSV and GEOJSON.
CodeSearchNet-ccr is a dataset for evaluating code retrieval models, hosted by CoIR-Retrieval. It employs the MTEB evaluation framework's dataset version for assessment. The dataset was last updated on September 12, 2024.
Four modified LIBERO robotics datasets formatted in RLDS for OpenVLA fine-tuning experiments, released by the OpenVLA team in 2024. These datasets adapt the original LIBERO benchmark specifically for training and evaluating vision-language-action models in simulated environments.
Dan's Memory Core: Core Curriculum Small is a foundational knowledge dataset generated using a Retrieval-Augmented Generation (RAG) implementation on curated source material. The dataset contains question-answer pairs covering broad strokes of knowledge common across multiple fields. It was created by PocketDoc and last updated on September 8, 2024.
Habedi's Practicalmachinelearning repository, updated in October 2024, provides a collection of Python-based machine learning resources including Jupyter and Zeppelin notebooks. It covers diverse domains such as anomaly detection, graph mining, and deep neural networks using libraries like scikit-learn and pandas.
WMS geospatial data provides map service access to school locations in Schmalförden, Germany. The dataset is maintained by the Bundesamt für Kartographie und Geodäsie and was last updated on September 16, 2024. Its specific content and structure are defined by the WMS protocol.
Machine learning and computer vision research papers are indexed in this curated repository alongside their released datasets, with updates as recent as October 2024. Created by Voxel51, the collection functions as a metadata bridge between academic publications and their corresponding source data.
Over four decades of historical and modern airborne radar sounding observations of the Ross Ice Shelf, spanning from 1971 to 2017. The dataset was created by a collaborative research project funded under the American Rescue Plan Act, involving three institutions and supporting student education. It contains processed and calibrated radar data from field campaigns in 1971-79 and 2011-17.