Dataset Viewer
Auto-converted to Parquet
@context
string
@type
string
conformsTo
string
name
string
alternateName
sequence
description
string
creator
dict
license
string
url
string
keywords
sequence
distribution
list
recordSet
list
https://mlcommons.org/croissant/context/v1.1.json
https://schema.org/Dataset
http://mlcommons.org/croissant/1.1
CdSpritesplus
[ "gabinsane/CdSpritesplus", "CdSprites+" ]
CdSprites+ is a synthetic bimodal dataset, designed specifically for comparison of the joint- and cross-generative capabilities of multimodal VAEs. This dataset extends the dSprites dataset with natural language captions and additional features and offers 5 levels of difficulty (based on the number of attributes) to find the minimal functioning scenario for each model. Moreover, its rigid structure enables automatic qualitative evaluation of the generated samples.
{ "@type": "Person", "name": "Gabriela Sejnova", "url": "https://huggingface.co/gabinsane" }
https://choosealicense.com/licenses/cc-by-nc-sa-4.0/
https://huggingface.co/datasets/gabinsane/CdSpritesplus
[ "text-classification", "feature-extraction", "English", "cc-by-nc-sa-4.0", "10K<n<100K", "🇺🇸 Region: US" ]
[ { "@type": "cr:FileObject", "name": "level1", "description": "Level 1 data", "contentUrl": "level1/data.h5", "encodingFormat": "application/x-hdf5" }, { "@type": "cr:FileObject", "name": "level2", "description": "Level 2 data", "contentUrl": "level2/data.h5", "encodingFormat": "application/x-hdf5" }, { "@type": "cr:FileObject", "name": "level3", "description": "Level 3 data", "contentUrl": "level3/data.h5", "encodingFormat": "application/x-hdf5" }, { "@type": "cr:FileObject", "name": "level4", "description": "Level 4 data", "contentUrl": "level4/data.h5", "encodingFormat": "application/x-hdf5" }, { "@type": "cr:FileObject", "name": "level5", "description": "Level 5 data", "contentUrl": "level5/data.h5", "encodingFormat": "application/x-hdf5" } ]
[ { "name": "level1_records", "data": "level1", "numRecords": 30000, "record": { "field": [ { "name": "image", "path": "/image", "dataType": "image", "value": null }, { "name": "text", "path": "/text", "dataType": "string", "value": null }, { "name": "level", "path": null, "dataType": "integer", "value": 1 } ] } }, { "name": "level2_records", "data": "level2", "numRecords": 30000, "record": { "field": [ { "name": "image", "path": "/image", "dataType": "image", "value": null }, { "name": "text", "path": "/text", "dataType": "string", "value": null }, { "name": "level", "path": null, "dataType": "integer", "value": 2 } ] } }, { "name": "level3_records", "data": "level3", "numRecords": 30000, "record": { "field": [ { "name": "image", "path": "/image", "dataType": "image", "value": null }, { "name": "text", "path": "/text", "dataType": "string", "value": null }, { "name": "level", "path": null, "dataType": "integer", "value": 3 } ] } }, { "name": "level4_records", "data": "level4", "numRecords": 30000, "record": { "field": [ { "name": "image", "path": "/image", "dataType": "image", "value": null }, { "name": "text", "path": "/text", "dataType": "string", "value": null }, { "name": "level", "path": null, "dataType": "integer", "value": 4 } ] } }, { "name": "level5_records", "data": "level5", "numRecords": 30000, "record": { "field": [ { "name": "image", "path": "/image", "dataType": "image", "value": null }, { "name": "text", "path": "/text", "dataType": "string", "value": null }, { "name": "level", "path": null, "dataType": "integer", "value": 5 } ] } } ]

CdSprites+ is a synthetic bimodal dataset, designed specifically for comparison of the joint- and cross-generative capabilities of multimodal VAEs. This dataset extends the dSprites dataset with natural language captions and additional features and offers 5 levels of difficulty (based on the number of attributes) to find the minimal functioning scenario for each model. Moreover, its rigid structure enables automatic qualitative evaluation of the generated samples.

Downloads last month
12