國立虎尾科技大學 |

Empirical Evaluations of Transformers.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Empirical Evaluations of Transformers./
作者:	McGuire, Jack.
面頁冊數:	1 online resource (139 pages)
附註:	Source: Masters Abstracts International, Volume: 85-05.
Contained By:	Masters Abstracts International85-05.
標題:	Computer engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9798380846868

Empirical Evaluations of Transformers.
McGuire, Jack.

Empirical Evaluations of Transformers. - 1 online resource (139 pages)

Source: Masters Abstracts International, Volume: 85-05.

Thesis (M.S.)--Rutgers The State University of New Jersey, School of Graduate Studies, 2023.

Includes bibliographical references

The transformer architecture of neural networks immediately took over the machine learning world upon the release of the paper "Attention is All You Need" by a team at Google Brain in 2017. But, perhaps more importantly, the Summer and Fall 2022 releases of generative networks such as ChatGPT, Midjourney, and DALL-E 2 saw the technology make an immediate impact on millions of people unrelated to the academic circles that had been talking about transformers for several years, shaking up creative industries and introducing the technology to many people with widely varying understandings of how it works.Thus, we set out to evaluate temporal sequence models such as transformers; first, on a deeply technical level, investigating the role of one of transformers' unique contributions, positional encoding, on the overall training time and efficacy of the system. We then evaluated the system empirically in a broader sense as a tool that a common person would use, in order to quantitatively identify shortcomings and misconceptions on what the systems are and are not able to accomplish. Finally, we looked at another more established temporal sequence model in the form of reinforcement learning in order to understand hierarchical representation of such models and discuss the utility of hierarchical models of transformers in the future.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798380846868Subjects--Topical Terms:

569006
Computer engineering.
Subjects--Index Terms:

ChatGPTIndex Terms--Genre/Form:

554714
Electronic books.

Empirical Evaluations of Transformers.
LDR:02697ntm a22003977 4500 001 1143660
005 20240517104624.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380846868
035 $a (MiAaPQ)AAI30688787
035 $a AAI30688787
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a McGuire, Jack. $3 1468416
245 1 0 $a Empirical Evaluations of Transformers.
264 0 $c 2023
300 $a 1 online resource (139 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 85-05.
500 $a Advisor: Dana, Kristin.
502 $a Thesis (M.S.)--Rutgers The State University of New Jersey, School of Graduate Studies, 2023.
504 $a Includes bibliographical references
520 $a The transformer architecture of neural networks immediately took over the machine learning world upon the release of the paper "Attention is All You Need" by a team at Google Brain in 2017. But, perhaps more importantly, the Summer and Fall 2022 releases of generative networks such as ChatGPT, Midjourney, and DALL-E 2 saw the technology make an immediate impact on millions of people unrelated to the academic circles that had been talking about transformers for several years, shaking up creative industries and introducing the technology to many people with widely varying understandings of how it works.Thus, we set out to evaluate temporal sequence models such as transformers; first, on a deeply technical level, investigating the role of one of transformers' unique contributions, positional encoding, on the overall training time and efficacy of the system. We then evaluated the system empirically in a broader sense as a tool that a common person would use, in order to quantitatively identify shortcomings and misconceptions on what the systems are and are not able to accomplish. Finally, we looked at another more established temporal sequence model in the form of reinforcement learning in order to understand hierarchical representation of such models and discuss the utility of hierarchical models of transformers in the future.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer engineering. $3 569006
650 4 $a Electrical engineering. $3 596380
653 $a ChatGPT
653 $a Large language models
653 $a Machine learning
653 $a Positional encoding
653 $a Transformers
655 7 $a Electronic books. $2 local $3 554714
690 $a 0544
690 $a 0464
690 $a 0800
710 2 $a Rutgers The State University of New Jersey, School of Graduate Studies. $b Electrical and Computer Engineering. $3 1241232
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Masters Abstracts International $g 85-05.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30688787 $z click for full text (PQDT)