๐ŸŽฏ๋ชฉํ‘œ

<aside> ๐Ÿ”ฅ Transformer-XL Baseline์— ์ตœ์‹  nlp๊ธฐ๋ฒ•๋“ค์„ ์ ์šฉํ•˜์—ฌ ์ˆ˜์น˜์ƒ์œผ๋กœ ๋ณ€ํ™” ๊ด€์ฐฐํ•˜๊ธฐ

</aside>

โœ…ํ•  ์ผ

๐Ÿ‘ฅ์—ญํ•  ๋ถ„๋ฐฐ

๊น€๋ฏผ์„œ : Metric(commu)

๊น€์‚ฐ : ๐Ÿฆป sparse attention

๊น€์„ฑ์ค€ :

๋ฐ•๋ฏผ์ˆ˜ : group encoding + soft labeling

๋ฐ•์ˆ˜๋นˆ : Metric(CAS)

์กฐ์ •๋นˆ : ๐Ÿฆป attention, Metric(CAS)

๐Ÿ“œ๋ฏธํŒ…

Untitled

Untitled

Attention listup

Sparse transformer

Generating Long Sequences with Sparse Transformers

Linear transformer

Linformer