About: GPT-2

Facets (new session)
Description
Metadata
Settings
- Rule:
- Inverse Functional Properties:
- "Same As":

About: GPT-2 Goto Sponge NotDistinct Permalink

An Entity of Type : wikidata:Q7397, within Data Space : dbpedia.demo.openlinksw.com associated with source document(s)
QRcode icon

http://dbpedia.demo.openlinksw.com/describe/?url=http%3A%2F%2Fdbpedia.org%2Fresource%2FGPT-2&invfp=IFP_OFF&sas=SAME_AS_OFF

Generative Pre-trained Transformer 2 (GPT-2) is an open-source artificial intelligence created by OpenAI in February 2019. GPT-2 translates text, answers questions, summarizes passages, and generates text output on a level that, while sometimes indistinguishable from that of humans, can become repetitive or nonsensical when generating long passages. It is a general-purpose learner; it was not specifically trained to do any of these tasks, and its ability to perform them is an extension of its general ability to accurately synthesize the next item in an arbitrary sequence. GPT-2 was created as a "direct scale-up" of OpenAI's 2018 GPT model, with a ten-fold increase in both its parameter count and the size of its training dataset.

Attributes	Values
rdf:type	Thing software schema:CreativeWork work wikidata:Q386724 Software
rdfs:label	GPT-2 (ca) Gpt-2 (es) GPT-2 (en) GPT-2 (ko) GPT-2 (sv) GPT-2 (uk)
rdfs:comment	GPT-2 (Generative Pre-trained Transformer 2) és un sistema que fa ús de la intel·ligència artificial per generar textos creada el febrer del 2019 per OpenAI, un laboratori de recerca impulsat per Elon Musk. Es tracta d'un sistema format per 1.5 bilions de paràmetres que generen text preveient paraula a paraula. D'aquesta manera, és capaç de traduir textos automàticament, respondre preguntes, resumir fragments de text... Aquest text que genera realment és extret d'unes 8 milions pàgines d'Internet, per la qual cosa compta amb un conjunt de dades d'uns 40GB de text per utilitzar. (ca) GPT-2 (Generative Pre-trained Transformer 2) es un sistema que hace uso de la inteligencia artificial para generar textos creada en febrero del 2019 por OpenAI, un laboratorio de investigación impulsado por Elon Musk. Se trata de un sistema formado por 1.5 billones de parámetros que generan texto prediciendo palabra a palabra. De esta forma es capaz de traducir textos automáticamente, responder preguntas, resumir fragmentos de texto… Este texto generado es en realidad extraído de unas 8 millones de páginas de Internet, por lo que cuenta con un conjunto de datos de unos 40GB de texto para utilizar. (es) 생성적 사전학습 변환기 2(Generative Pre-trained Transformer 2) 즉 GPT-2는 OpenAI에서 만든 인공지능이다. 비지도 학습과 생성적 사전학습(generative pre-training)기법, 변환기(transformer)를 적용하였다. 번역과 대화, 작문을 할 수 있으며 단문의 경우 인간이 쓴 글인지 기계가 쓴 글인지 구분하기 힘들다. (ko) Generative Pre-trained Transformer 2 (GPT-2) is an open-source artificial intelligence created by OpenAI in February 2019. GPT-2 translates text, answers questions, summarizes passages, and generates text output on a level that, while sometimes indistinguishable from that of humans, can become repetitive or nonsensical when generating long passages. It is a general-purpose learner; it was not specifically trained to do any of these tasks, and its ability to perform them is an extension of its general ability to accurately synthesize the next item in an arbitrary sequence. GPT-2 was created as a "direct scale-up" of OpenAI's 2018 GPT model, with a ten-fold increase in both its parameter count and the size of its training dataset. (en) Generative Pre-trained Transformer 2 (GPT-2) är en som använder djupinlärning för att producera människoliknande text. GPT-2 översätter text, svarar på frågor, sammanfattar passager, och genererar textutdata på en nivå som, även om den ibland inte kan skiljas från människors, kan bli repetitiva eller meningslösa när man genererar långa passager.Det är en allmäninlärare; den var inte specifikt utbildad för att utföra någon av dessa uppgifter, och dess förmåga att utföra dem är en förlängning av dess allmänna förmåga att exakt syntetisera nästa objekt i en godtycklig sekvens. GPT-2 skapades som en "direkt uppskalning" av OpenAI:s 2018 GPT-modell, med en tiofaldig ökning av både dess parameterantal och storleken på dess träningsdatauppsättning. (sv)
foaf:name	Generative Pre-trained Transformer 2 (GPT-2) (en)
foaf:homepage	https://openai.com/blog/gpt-2-1-5b-release/
name	Generative Pre-trained Transformer 2 (en)
foaf:depiction
dcterms:subject	Deep learning software applications Software using the MIT license Natural language generation Unsupervised learning Language modeling Open-source artificial intelligence
Wikipage page ID	66045029 (xsd:integer)
Wikipage revision ID	1124372728 (xsd:integer)
Link from a Wikipage to another Wikipage	Caltech Potentiometer Question answering Natural language generation Natural language processing Neural Turing machine Neural network Bike Deep neural network Human-computer interaction Paul Werbos Perceptron University of Amsterdam University of Toronto Vanishing gradient problem Vox Media Deep learning ELIZA Motorcycle Printing press GPU Zero-shot learning Photocell Deep learning software applications Software using the MIT license Conditional random field Corn dog Perceptrons (book) Frank Rosenblatt Generative model Gradient descent Connectionism Convolutional neural network The New York Times The Register The Verge Objective function Obscene Recurrent neural networks 1980s Application programming interface MIT Machine learning Commonsense reasoning Computational linguistics Domain-specific Keyword (linguistics) Kunihiko Fukushima Parallel computing Spamming Backpropagation Central Intelligence Agency Tokenization (lexical analysis) Turing test Data science Wikipedia GPT-3 Cyclic graph Jürgen Schmidhuber Language model AI Dungeon Alan Turing Allen Institute for Artificial Intelligence Exclusive or Anima Anandkumar Noam Chomsky Parallelization Discriminative model

Faceted Search & Find service v1.17_git139 as of Feb 29 2024

Alternative Linked Data Documents: ODE Content Formats:

RDF

ODATA

Microdata

About

OpenLink Virtuoso version 08.03.3330 as of Mar 19 2024, on Linux (x86_64-generic-linux-glibc212), Single-Server Edition (378 GB total memory, 59 GB memory in use)
Data on this page belongs to its respective rights holders.
Virtuoso Faceted Browser Copyright © 2009-2024 OpenLink Software