T5

GPTKB entity

Statements (59)
Predicate Object
gptkbp:instanceOf large language model
gptkbp:architecture gptkb:transformation
gptkbp:arXivID 1910.10683
gptkbp:author gptkb:Adam_Roberts
gptkb:Colin_Raffel
gptkb:Katherine_Lee
gptkb:Michael_Matena
gptkb:Noam_Shazeer
gptkb:Peter_J._Liu
gptkb:Sharan_Narang
gptkb:Wei_Li
gptkb:Yanqi_Zhou
gptkbp:availableOn gptkb:TensorFlow
gptkb:PyTorch
gptkb:Hugging_Face_Transformers
gptkbp:citation high
gptkbp:developedBy gptkb:Google_Research
gptkbp:encoderDecoder yes
gptkbp:fineTunedWith yes
gptkbp:fullName gptkb:Text-To-Text_Transfer_Transformer
gptkbp:hasVariant gptkb:T5-11B
gptkb:T5-3B
gptkb:T5-Base
gptkb:T5-Large
gptkb:T5-Small
https://www.w3.org/2000/01/rdf-schema#label T5
gptkbp:input gptkb:text
gptkbp:introducedIn 2019
gptkbp:language English
gptkbp:license Apache 2.0
gptkbp:notableFor state-of-the-art results on NLP benchmarks
unified text-to-text framework
gptkbp:notablePublication gptkb:Exploring_the_Limits_of_Transfer_Learning_with_a_Unified_Text-to-Text_Transformer
gptkbp:openSource yes
gptkbp:output gptkb:text
gptkbp:parameter 60 million
220 million
11 billion
3 billion
770 million
gptkbp:pretrainingObjective span corruption
gptkbp:taskFormat text-to-text
gptkbp:tokenizerType gptkb:SentencePiece
gptkbp:trainer gptkb:Colossal_Clean_Crawled_Corpus
C4
gptkbp:type self-attention
gptkbp:usedFor translator
question answering
summarization
text generation
text classification
gptkbp:bfsParent gptkb:large_language_model
gptkb:Tramway_de_Lyon
gptkb:bus
gptkb:transformation
gptkb:convolutional_neural_network
gptkb:GPT-3
gptkb:Large_Language_Models
gptkbp:bfsLayer 5