Statements (35)
Predicate | Object |
---|---|
gptkbp:instanceOf |
multimodal large language model
|
gptkbp:announced |
2024-03-13
|
gptkbp:architecture |
gptkb:Mixture_of_Experts
|
gptkbp:author |
Google DeepMind researchers
|
gptkbp:developedBy |
gptkb:Google_DeepMind
|
https://www.w3.org/2000/01/rdf-schema#label |
MM-1
|
gptkbp:input |
gptkb:illustrator
gptkb:text audio |
gptkbp:language |
English
|
gptkbp:mode |
gptkb:illustrator
gptkb:text audio |
gptkbp:notablePublication |
gptkb:MM-1:_Methods,_Analysis_&_Insights_from_Multimodal_LLM_Pre-training
|
gptkbp:openSource |
no
|
gptkbp:organization |
gptkb:Google
|
gptkbp:output |
gptkb:illustrator
gptkb:text audio |
gptkbp:parameter |
30B
162B 80B |
gptkbp:pretrainingData |
gptkb:illustrator
gptkb:text audio |
gptkbp:purpose |
multimodal understanding
multimodal generation |
gptkbp:relatedTo |
gptkb:Gemini
gptkb:GPT-4 gptkb:LLaVA Flamingo |
gptkbp:trainer |
multimodal pretraining
supervised pretraining |
gptkbp:bfsParent |
gptkb:Italia_airship
|
gptkbp:bfsLayer |
7
|