|
gptkbp:instanceOf
|
open dataset project
|
|
gptkbp:challenge
|
scalability
data quality
ethical considerations
copyright concerns
|
|
gptkbp:country
|
gptkb:Germany
|
|
gptkbp:focus
|
image-text datasets
multimodal datasets
open datasets for machine learning
|
|
gptkbp:foundedYear
|
2021
|
|
gptkbp:founder
|
gptkb:Andreas_Köpf
gptkb:Christoph_Schuhmann
gptkb:Romain_Beaumont
Clayton Davis
Tristan Bitard
|
|
gptkbp:fullName
|
gptkb:Large-scale_Artificial_Intelligence_Open_Network
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
LAION project
|
|
gptkbp:license
|
gptkb:CC-BY_4.0
|
|
gptkbp:mission
|
democratize access to large-scale machine learning datasets
|
|
gptkbp:notableCollaboration
|
gptkb:OpenAI
gptkb:Hugging_Face
gptkb:EleutherAI
gptkb:Stability_AI
|
|
gptkbp:notableEvent
|
collaboration with academic and industry partners
contribution to Stable Diffusion training
open sourcing of large-scale datasets
release of LAION-400M in 2021
release of LAION-5B in 2022
|
|
gptkbp:notableFor
|
research in computer vision
training large language models
training diffusion models
research in multimodal AI
|
|
gptkbp:notableModel
|
gptkb:Stable_Diffusion
gptkb:OpenCLIP
CLIP-based models
Diffusion models
|
|
gptkbp:notablePublication
|
gptkb:LAION-5B:_An_open_large-scale_dataset_for_training_next_generation_image-text_models
LAION-400M: Open Dataset of CLIP-filtered 400 Million Image-Text Pairs
|
|
gptkbp:notableSupporter
|
gptkb:OpenAI
gptkb:Hugging_Face
gptkb:EleutherAI
gptkb:Stability_AI
community donations
|
|
gptkbp:trainer
|
gptkb:LAION-400M
gptkb:LAION-Aesthetics
gptkb:LAION-5B
|
|
gptkbp:type
|
gptkb:nonprofit_organization
metadata
image-text pairs
multimodal data
aesthetic scores
|
|
gptkbp:website
|
https://laion.ai/
|
|
gptkbp:bfsParent
|
gptkb:LAION-Aesthetics
|
|
gptkbp:bfsLayer
|
7
|