CLIP

GPTKB entity

Statements (56)
Predicate Object
gptkbp:instanceOf gptkb:model
gptkbp:architecture gptkb:transformation
gptkbp:author gptkb:Ilya_Sutskever
gptkb:Aditya_Ramesh
gptkb:Chris_Hallacy
gptkb:Gabriel_Goh
gptkb:Girish_Sastry
gptkb:Mark_Chen
gptkb:Pamela_Mishkin
gptkb:Sandhini_Agarwal
gptkb:Jack_Clark
gptkb:Alec_Radford
gptkb:Amanda_Askell
gptkb:Jong_Wook_Kim
gptkb:Miles_Brundage
gptkbp:bench gptkb:COCO
gptkb:Flickr30k
gptkb:ImageNet
gptkbp:developedBy gptkb:OpenAI
gptkbp:fineTunedWith yes
gptkbp:fullName gptkb:Contrastive_Language–Image_Pre-training
https://www.w3.org/2000/01/rdf-schema#label CLIP
gptkbp:influenced multimodal AI research
gptkbp:input gptkb:illustrator
gptkb:text
gptkbp:inputDimension 224x224 pixels (for images)
gptkbp:language English
gptkbp:license gptkb:MIT_(for_code)
gptkbp:memiliki_tugas image classification
zero-shot learning
image-text retrieval
gptkbp:modelWeightsAvailability not fully open-sourced
gptkbp:notableFor multimodal learning
zero-shot performance on ImageNet
gptkbp:notablePublication gptkb:Learning_Transferable_Visual_Models_From_Natural_Language_Supervision
gptkbp:openSource partially
gptkbp:output similarity scores
vector embeddings
gptkbp:paperPublicationDate 2021
gptkbp:pretrained yes
gptkbp:publishedIn gptkb:arXiv
gptkbp:relatedTo gptkb:ALIGN
gptkb:BLIP
gptkb:DALL-E
gptkbp:releaseDate 2021
gptkbp:trainer gptkb:OpenAI's_proprietary_dataset
400 million image-text pairs
gptkbp:trainingDataSource publicly available internet data
gptkbp:usedIn robotics
content moderation
image search
art generation
visual question answering
gptkbp:uses contrastive learning
gptkbp:bfsParent gptkb:EFMD
gptkbp:bfsLayer 5