datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
colorscolor-palettes-sdBlue_Color_Sky
Posters are mad,
Skies are so blue,
Download some data,
Or you'll get mad too.
pick_color_eggEgoPlan_test
EgoPlan-Test
Videos: We trim videos in EgoPlan test set by the start_frame and end_frame
Text: we add "video_path" key to the original json file.
manga-colorization-mastergas-cylinder-color-codes
Gas cylinder color codes by standard and country
Canonical, always-current version: https://referencesource.org/gas-cylinder-color-codes/
Machine-readable: https://referencesource.org/gas-cylinder-color-codes/data.json — this mirror is a point-in-time copy.
Last verified: 2026-08-05
Stale after: 2028-08-05 (past this date, prefer the canonical copy —
it re-verifies on a cadence this snapshot does not)
Records: 17
Color coding of compressed gas cylinders (shoulder and body… See the full description on the dataset page: https://huggingface.co/datasets/referencesource/gas-cylinder-color-codes.rollout_colorlogo_rl_round1thermocouple-type-color-code-crosswalk
Thermocouple type letter to alloy pair, ANSI/IEC color code and tolerance class
Canonical, always-current version: https://referencesource.org/thermocouple-type-color-code-crosswalk/
Machine-readable: https://referencesource.org/thermocouple-type-color-code-crosswalk/data.json — this mirror is a point-in-time copy.
Last verified: 2026-08-19
Stale after: 2028-08-18 (past this date, prefer the canonical copy —
it re-verifies on a cadence this snapshot does not)
Records: 12
A… See the full description on the dataset page: https://huggingface.co/datasets/referencesource/thermocouple-type-color-code-crosswalk.color-filtered-c4
CoLoR-Filtered C4
This repo contains two datasets: color-filtered-c4-books and color-filtered-c4-down associated with the CoLoR-Filter paper.
Each dataset is a 64x filtered version of the C4 dataset from Raffel et al., 2019 that has been selected using the CoLoR-Filter algorithm for data selection.
Each dataset has about 2.7b tokens when using the allenai/eleuther-ai-gpt-neox-20b-pii-special tokenizer.
color-filtered-c4-books was selected to target books based on a small (25m token)… See the full description on the dataset page: https://huggingface.co/datasets/davidbrandfonbrener/color-filtered-c4.food-color-additive-status-us-vs-eu
Food color additive cross-reference: US FD&C designation vs EU E-number, with regulatory status comparison
Canonical, always-current version: https://referencesource.org/food-color-additive-status-us-vs-eu/
Machine-readable: https://referencesource.org/food-color-additive-status-us-vs-eu/data.json — this mirror is a point-in-time copy.
Last verified: 2026-08-10
Stale after: 2027-08-10 (past this date, prefer the canonical copy —
it re-verifies on a cadence this snapshot does… See the full description on the dataset page: https://huggingface.co/datasets/referencesource/food-color-additive-status-us-vs-eu.Color-modele-languageVideoLLaMB-IT
VideoLLaMB Instruction
This is preprocessed visual instruction tuning data for VideoLLaMB, including
llava_pllava_filter.json: LLaVA-Instruct and PLLaVA.
VoiceAssistant_unitsWe use mHuBERT and K-means Model to convert answer from VoiceAssistant (synthesized by vits) to units.
*pred* indicates the answer from the speech model
*gt* indicates the answer from the ground-truth
For usage, please refer to Open-Omni-Nexus
math_geometry_polygon_color_level_3math_polygon_colormath_geometry_polygon_color_level_1MoviePuzzle
MoviePuzzle Dataset
Introduce
This dataset is based on MovieNet for MoviePuzzle task.
We use 228 movie to generate 10031 movie clips. 7048 clips for train, 589 for val, 1178 for in domain test and 1196 for out domain test.
Download
You can download the full dataset here: https://movienet.github.io/
Structure
We categorized a dataset of 10031 movie clips into labels ranging from 0 to 10030. Then, we divided this dataset into four subsets: the training… See the full description on the dataset page: https://huggingface.co/datasets/ColorfulAI/MoviePuzzle.math_geometry_polygon_color_7B_test_inmath_geometry_polygon_color_level_5math_geometry_polygon_color_level_2math_geometry_polygon_color_7B_test_outcolor_swatchascii_colors_discussions
Dataset Card for ParisNeo/ascii_colors
Dataset Description
This dataset contains synthetic, structured conversations designed to encapsulate the knowledge within the ascii_colors Python library (specifically around version 0.8.1). The primary goal of this dataset is to facilitate the fine-tuning of Large Language Models (LLMs) to become experts on the ascii_colors library, capable of answering questions and performing tasks related to it without relying on… See the full description on the dataset page: https://huggingface.co/datasets/ParisNeo/ascii_colors_discussions.math_geometry_polygon_color_7B_trainmath_comp_vertex_colormath_geometry_polygon_color_level_4indonesian-color-terms
Istilah Warna Indonesia 🎨
Bukan cuma merah-kuning-hijau — bahasa Indonesia punya kekayaan istilah warna: lembayung, nila, sawo matang, kuning langsat, biru dongker. Dataset ini ngumpulin istilah warna khas Indonesia + deskripsi + contoh.
Kenapa dataset ini ada?
LLM sering ga paham istilah warna lokal — lembayung bukan sekadar "ungu", sawo matang bukan "brown". Dataset ini bantu model paham nuansa warna khas Indonesia. Belum ada dataset istilah warna bahasa… See the full description on the dataset page: https://huggingface.co/datasets/LorthGyu/indonesian-color-terms.simon-arc-solve-color-v2
Version 1
ARC-AGI Tasks where the colors gets manipulated.
Currently it's two-color images, where the transformation is to swap colors.
The image sizes are between 1 and 5 pixels.
Predict the number of rows in the output image.
Version 2
Number of test: 1-2. Previously it was always 1 test.
simon-arc-solve-color-v3
Version 1
ARC-AGI Tasks where the colors gets manipulated.
Currently it's two-color images, where the transformation is to swap colors.
The image sizes are between 1 and 5 pixels.
Predict the number of rows in the output image.
Version 2
Number of test: 1-2. Previously it was always 1 test.
Version 3
Number of tests: 1.
Identify most popular color, and least popular color. The output size is always 1x1.