|
gptkbp:instanceOf
|
large language model
|
|
gptkbp:architecture
|
gptkb:transformation
|
|
gptkbp:arXivID
|
2302.13971
|
|
gptkbp:author
|
gptkb:Baptiste_Rozière
gptkb:Gautier_Izacard
gptkb:Guillaume_Lample
gptkb:Hugo_Touvron
gptkb:Marie-Anne_Lachaux
gptkb:Nils_Reimers
gptkb:Thibaut_Lavril
gptkb:Timothée_Lacroix
gptkb:Xavier_Martinet
Édouard Grave
|
|
gptkbp:citation
|
gptkb:LLaMA:_Open_and_Efficient_Foundation_Language_Models
2023
|
|
gptkbp:context
|
2048 tokens
|
|
gptkbp:developedBy
|
gptkb:Meta_AI
|
|
gptkbp:hasModel
|
decoder-only transformer
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
LLaMA 1
|
|
gptkbp:input
|
gptkb:text
|
|
gptkbp:language
|
English
|
|
gptkbp:license
|
research-only license
|
|
gptkbp:notableFor
|
efficient performance on smaller hardware
leak of model weights in March 2023
high performance at small scale
influencing open LLM ecosystem
sparking open-source LLM movement
|
|
gptkbp:openSource
|
false
|
|
gptkbp:output
|
gptkb:text
|
|
gptkbp:parameter
|
13B
33B
65B
7B
|
|
gptkbp:predecessor
|
nan
|
|
gptkbp:purpose
|
research
|
|
gptkbp:releaseYear
|
2023
|
|
gptkbp:successor
|
gptkb:LLaMA_2
|
|
gptkbp:trainer
|
publicly available datasets
|
|
gptkbp:trainingDataSize
|
1.4 trillion tokens
|
|
gptkbp:trainingObjective
|
causal language modeling
|
|
gptkbp:usedFor
|
text generation
fine-tuning experiments
research benchmarking
|
|
gptkbp:bfsParent
|
gptkb:LLaMA_model_family
|
|
gptkbp:bfsLayer
|
8
|