Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
サブセット探索を用いた高速なkNNニューラル機械翻訳
Search
Hiroyuki Deguchi
March 22, 2024
Research
0
94
サブセット探索を用いた高速なkNNニューラル機械翻訳
第8回AAMTセミナー
AAMT若手翻訳研究会
最優秀賞
Hiroyuki Deguchi
March 22, 2024
Tweet
Share
More Decks by Hiroyuki Deguchi
See All by Hiroyuki Deguchi
20250226 NLP colloquium: "SoftMatcha: 10億単語規模コーパス検索のための柔らかくも高速なパターンマッチャー"
de9uch1
0
250
20240820: Minimum Bayes Risk Decoding for High-Quality Text Generation Beyond High-Probability Text
de9uch1
0
210
20240226_AAMT-Japio
de9uch1
0
130
Searching for Needles in a Haystack: On the Role of Incidental Bilingualism in PaLM’s Translation Capability
de9uch1
0
120
Paper Reading: Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural Machine Translation
de9uch1
0
150
My Research Environmental Setup
de9uch1
0
260
Nearest Neighbor Machine Translation
de9uch1
0
220
Paper Reading - Dynamic Programming Encoding for Subword Segmentation in Neural Machine Translation
de9uch1
0
260
paper reading - Tree Transformer
de9uch1
0
220
Other Decks in Research
See All in Research
AWS 音声基盤モデル トーク解析AI MiiTelの音声処理について
ken57
0
170
Vision Language Modelと完全自動運転AIの最新動向
tsubasashi
1
310
Weekly AI Agents News! 2月号 アーカイブ
masatoto
1
110
BtoB プロダクトにおけるインサイトマネジメントの必要性 現場ドリブンなカミナシがインサイトマネジメントに取り組むワケ / Why field-driven Kaminashi is working on insight management
kaminashi
1
360
さくらインターネット研究所 アップデート2025年
matsumoto_r
PRO
0
430
DeepSeek-R1の論文から読み解く背景技術
personabb
3
490
ドローンやICTを活用した持続可能なまちづくりに関する研究
nro2daisuke
0
190
[輪講] Transformer Layers as Painters
nk35jk
4
720
チュートリアル:Mamba, Vision Mamba (Vim)
hf149
6
3.2k
言語モデルの内部機序:解析と解釈
eumesy
PRO
32
13k
資産間の相関関係を頑健に評価する指標を用いたファクターアローケーション戦略の構築
nomamist
0
170
ナレッジプロデューサーとしてのミドルマネージャー支援 - MIMIGURI「知識創造室」の事例の考察 -
chiemitaki
0
250
Featured
See All Featured
A Modern Web Designer's Workflow
chriscoyier
693
190k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
227
22k
Evolution of real-time – Irina Nazarova, EuRuKo, 2024
irinanazarova
7
610
Stop Working from a Prison Cell
hatefulcrawdad
268
20k
The World Runs on Bad Software
bkeepers
PRO
67
11k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
jcasabona
32
2.2k
Understanding Cognitive Biases in Performance Measurement
bluesmoon
28
1.6k
The Power of CSS Pseudo Elements
geoffreycrofte
75
5.7k
Designing for humans not robots
tammielis
250
25k
Site-Speed That Sticks
csswizardry
4
450
Done Done
chrislema
183
16k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
11
610
Transcript
𝒌
◼ ⚫ ⚫ ◼ ⚫ (Zhang+, NAACL2018; Gu+, AAAI2018; Khandelwal+,
ICLR2021) ▶ (Nagao, 1984) ▶ ⚫ 𝑘 (Khandelwal+, ICLR2021) ▶ ▶ ▶ Guiding Neural Machine Translation with Retrieved Translation Pieces (Zhang+, NAACL2018) Search Engine Guided Neural Machine Translation (Gu+, AAAI2018) Nearest Neighbor Machine Translation (Khandelwal+, ICLR2021) A framework for a mechanical translation between Japanese and English by analogy principle (Nagao, 1984)
◼ ◼ ⚫ ⚫
𝒌 (Khandelwal+, ICLR2021) ◼ ⚫ ⚫ ⚫ ◼ ⚫ ▶
⚫ ▶ ≈ Nearest Neighbor Machine Translation (Khandelwal+, ICLR2021) 𝒙 𝒚
𝒌 (Khandelwal+, ICLR2021) 𝒌𝑖 ∈ ℝ𝐷 𝑓 𝒙, 𝒚<𝑡 ∈
ℝ𝐷 Nearest Neighbor Machine Translation (Khandelwal+, ICLR2021) ◼ 𝑘 ◼ ⚫ ⚫ 𝑝𝑘NN 𝑦𝑡 𝒙, 𝒚<𝑡 ∝ 𝑖=1 𝑘 𝟙𝑦𝑡=𝑣𝑖 exp − 𝒌𝑖 − 𝑓 𝒙, 𝒚<𝑡 2 2 𝜏 ◼ 𝑘
𝒌 ◼ (Martins+, EMNLP2022) ◼ (Meng+, ACLFindings2022) ⚫ 𝑘 𝑘
𝜆 = 0.5 𝑘 = 16 Chunk-based Nearest Neighbor Machine Translation (Martins+, EMNLP2022) Fast Nearest Neighbor Machine Translation (Meng+, ACL Findings2022)
𝒌 ◼ 𝑘 ◼ ⚫ 𝑘 (Matsui+, ACMMM2018) ⚫ 𝑘
𝑘 𝑘 Reconfigurable Inverted Index (Matsui+, ACMMM2018) 𝒌
◼ ⚫ 𝑘 ⚫ 𝑘 ◼ ◼ 𝑘
𝑛 𝑘 1 1 1 1 1 1 1 1
1
𝑛 𝑘 1 1 1 1 1 1 1 1
1
𝑛 𝑘 1 1 1 1 1 1 1 1
1
⚫ ⚫ ⚫ ⚫ ⚫ 𝑘 𝜆 = 0.5 𝑘
= 16 𝑛 = 56
𝑘 𝑘 ◼ 𝑘 ⚫ ▶ ⚫ ▶
◼ 𝑘 𝒌 𝒌
◼ ⚫ 𝑘
𝒌 𝒌 ◼ ⚫ ⚫ ◼ 𝑘 ⚫ ⚫ ◼
⚫
⚫ ⚫ ▶ ⚫ ▶