|
gptkbp:instanceOf
|
large language model
|
|
gptkbp:architecture
|
gptkb:transformation
|
|
gptkbp:citation
|
high
|
|
gptkbp:context
|
2048 tokens
|
|
gptkbp:developedBy
|
gptkb:OpenAI
|
|
gptkbp:hardware
|
gptkb:NVIDIA_V100_GPUs
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
OpenAI GPT-3
|
|
gptkbp:input
|
gptkb:text
|
|
gptkbp:language
|
English
multilingual
|
|
gptkbp:license
|
Commercial
|
|
gptkbp:limitation
|
hallucination
context window limit
bias
data cutoff 2019
factual inaccuracy
lack of reasoning
|
|
gptkbp:mobileAppAvailable
|
Yes
|
|
gptkbp:notableFeature
|
few-shot learning
one-shot learning
zero-shot learning
|
|
gptkbp:notableFor
|
Education
Chatbots
Content creation
Business automation
Coding assistants
|
|
gptkbp:notablePublication
|
gptkb:Language_Models_are_Few-Shot_Learners
|
|
gptkbp:notableUser
|
gptkb:Microsoft
gptkb:Copy.ai
gptkb:GitHub_Copilot
gptkb:Replika
gptkb:ChatGPT_(early_versions)
gptkb:Jasper_AI
|
|
gptkbp:openSource
|
No
|
|
gptkbp:output
|
gptkb:text
|
|
gptkbp:parameter
|
175 billion
|
|
gptkbp:predecessor
|
gptkb:OpenAI_GPT-2
|
|
gptkbp:purpose
|
natural language processing
translator
question answering
summarization
text generation
|
|
gptkbp:releaseDate
|
2020
|
|
gptkbp:successor
|
gptkb:OpenAI_GPT-4
gptkb:OpenAI_GPT-3.5
|
|
gptkbp:trainer
|
gptkb:Wikipedia
gptkb:Common_Crawl
gptkb:WebText
Books
unsupervised learning
|
|
gptkbp:trainingDataCutoff
|
October 2019
|
|
gptkbp:bfsParent
|
gptkb:OpenAI_Codex
gptkb:OpenAI_DALL-E
gptkb:In-Context_Learning
|
|
gptkbp:bfsLayer
|
6
|