Upgrade to Pro — share decks privately, control downloads, hide ads and more …

[論紹介] Fast Lexically Constrained Decoding with ...

Avatar for onizuka laboratory onizuka laboratory
July 11, 2018
Research
0
380

[論紹介] Fast Lexically Constrained Decoding with Dynamic Beam Allocation for Neural Machine Translation

弊研究室で行なったNAACL読み会の発表資料です。

Avatar for onizuka laboratory

onizuka laboratory

July 11, 2018
Tweet

More Decks by onizuka laboratory

See All by onizuka laboratory
Phrase-Based & Neural Unsupervised Machine Translation
Avatar for onizuka laboratory onilab
0
110
Tell-and-Answer: Towards Explainable Visual Question Answering using Attributes and Captions
Avatar for onizuka laboratory onilab
0
71
Card-660: A Reliable Evaluation Framework for Rare Word Representation Models
Avatar for onizuka laboratory onilab
0
33
A Word-Complexity Lexicon and A Neural Readability Ranking Model for Lexical Simplification
Avatar for onizuka laboratory onilab
0
120
Integrating Transformer and Paraphrase Rules for Sentence Simplification
Avatar for onizuka laboratory onilab
0
59
An Auto-Encoder Matching Model for Learning Utterance-Level Semantic Dependency in Dialogue Generation
Avatar for onizuka laboratory onilab
0
55
Generating More Interesting Responses in Neural Conversation Models with Distributional Constraints
Avatar for onizuka laboratory onilab
0
100
Modeling Multi-turn Conversation with Deep Utterance Aggregation
Avatar for onizuka laboratory onilab
0
95
Learning Semantic Sentence Embeddings using Pair-wise Discriminator
Avatar for onizuka laboratory onilab
0
120

Other Decks in Research

See All in Research
生成的推薦の人気バイアスの分析：暗記の観点から / JSAI2025
Avatar for Shotaro Ishihara upura
0
260
[輪講] SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
Avatar for Naoki Kato nk35jk
2
980
Cross-Media Information Spaces and Architectures
Avatar for Beat Signer signer
PRO
0
240
論文読み会 SNLP2025 Learning Dynamics of LLM Finetuning. In: ICLR 2025
Avatar for S s_mizuki_nlp
0
200
Time to Cash: The Full Stack Breakdown of Modern ATM Attacks
Avatar for D ratatata
0
120
数理最適化に基づく制御
Avatar for MIKIO KUBO mickey_kubo
6
730
20250605_新交通システム推進議連_熊本都市圏「車1割削減、渋滞半減、公共交通2倍」から考える地方都市交通政策
Avatar for Traffic Brain trafficbrain
0
770
論文紹介：Not All Tokens Are What You Need for Pretraining
Avatar for Kosuke Nishida kosuken
0
170
なめらかなシステムと運用維持の終わらぬ未来 / dicomo2025_coherently_fittable_system
Avatar for monochromegane monochromegane
0
2.8k
Creation and environmental applications of 15-year daily inundation and vegetation maps for Siberia by integrating satellite and meteorological datasets
Avatar for SatAI.challenge satai
3
270
Mechanistic Interpretability：解釈可能性研究の新たな潮流
Avatar for Koshiro Aoki koshiro_aoki
1
410
引力・斥力を制御可能なランダム部分集合の確率分布
Avatar for Takahiro Kawashima wasyro
0
240

Featured

See All Featured
How to train your dragon (web standard)
Avatar for Monica Dinculescu notwaldorf
96
6.2k
Chrome DevTools: State of the Union 2024 - Debugging React & Beyond
Avatar for Addy Osmani addyosmani
7
840
The MySQL Ecosystem @ GitHub 2015
Avatar for Sam Lambert samlambert
251
13k
Save Time (by Creating Custom Rails Generators)
Avatar for Garrett Dimon garrettdimon
PRO
32
1.5k
Large-scale JavaScript Application Architecture
Avatar for Addy Osmani addyosmani
512
110k
Improving Core Web Vitals using Speculation Rules API
Avatar for Sergey Chernyshev sergeychernyshev
18
1.1k
Easily Structure & Communicate Ideas using Wireframe
Avatar for Afnizar Nur Ghifari afnizarnur
194
16k
Agile that works and the tools we love
Avatar for Rasmus Luckow-Nielsen rasmusluckow
330
21k
JavaScript: Past, Present, and Future - NDC Porto 2020
Avatar for David Neal reverentgeek
51
5.6k
実際に使うSQLの書き方 徹底解説 / pgcon21j-tutorial
Avatar for soudai sone soudai
PRO
188
55k
GitHub's CSS Performance
Avatar for Jon Rohan jonrohan
1032
460k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
Avatar for Joe Casabona jcasabona
33
2.4k

Transcript

  1. 'BTU-FYJDBMMZ$POTUSBJOFE%FDPEJOH XJUI%ZOBNJD#FBN"MMPDBUJPO GPS/FVSBM.BDIJOF5SBOTMBUJPO .BUU1PTUBOE%BWJE7JMBS "NB[PO3FTFBSDI #FSMJO (FSNBOZ 1SPDPG/""$- QQ 1SFTFOUFECZ5PNPZVLJ,BKJXBSB

  2. $POUSPMMJOH/.5PVUQVU • /.5SFNPWFTNBOZXBZTPGNBOVBMMZHVJEJOH UIFUSBOTMBUJPOQSPDFTTJOPMEFSQBSBEJHNT 1#4.5  • *ODPSQPSBUFEPNBJOTQFDJGJDEJDUJPOBSJFT • 'PSDFBUSBOTMBUJPODIPJDFGPSDFSUBJOXPSET

    • -FYJDBMMZ$POTUSBJOFE%FDPEJOHGPS4FRVFODF(FOFSBUJPO 6TJOH(SJE#FBN4FBSDI )PLBNQ BOE-JV "$- • 6TFSTDBOTQFDJGZUIFXPSETUIBUNVTUBQQFBSJOUIFPVUQVU • 6TFGVMGPSUBTLTTVDIBTQPTUFEJUJOH DBQUJPOHFOFSBUJPO   /JFNBOE IBUEJF"CTJDIU FJOF .BVFS[V CBVFO /JFNBOE IBUEJF"CTJDIU FJOF .BVFS[V FSSJDIUFO ,FJOFS IBUEJF"CTJDIU FJOF .BVFS[V CBVFO FSSJDIUFO LFJOFS -FYJDBMMZ$POTUSBJOFE%FDPEJOH /PPOFIBTUIFJOUFOUJPO PGCVJMEJOHBXBMM

  3. (#4IBTBMBSHF3VOUJNF$PNQMFYJUZ %FDPEJOH$PNQMFYJUZ 0 /L$ (SJE#FBN4FBSDI )PLBNQ BOE-JV  0 /L

    %ZOBNJD#FBN"MMPDBUJPO 5IJT8PSL /TFOUFODFMFOHUI LCFBNTJ[F $OVNCFSPGDPOTUSBJOUT (#4IBTUXPQSPCMFNT • %FDPEJOHDPNQMFYJUZJTMJOFBSJODPOTUSBJOUT • $BOOPUTQFDJGZUIFCFBNTJ[FBUNPEFMMPBEUJNF 

  4. %#"%ZOBNJD#FBN"MMPDBUJPO • 4JNQMZSFQMBDFUIF,#&45 JNQMFNFOUBUJPO GSPNUIFBMHPSJUINPGTUBOEBSECFBNEFDPEJOH • *OTUFBEPGTFMFDUJOHUIFUPQLJUFNT  (FOFSBUJOHBMJTUPGDBOEJEBUFT 8FOPXOFFEUPFOTVSFUIBUDBOEJEBUFTQSPHSFTT

    UISPVHIUIFTFUPGQSPWJEFEDPOTUSBJOUT  "MMPDBUJOHUIFCFBNBDSPTTUIFDPOTUSBJOUCBOLT 8JUIBGJYFETJ[FECFBNBOEBSCJUSBSZOVNCFSPG DPOTUSBJOUT XFOFFEUPGJOEBOBMMPDBUJPOTUSBUFHZ  GVODUJPO #&".4&"3$) / L CFBNˡ %&$0%&3*/*5 L GPS UJNFTUFQUJO/EP TDPSFT%&$0%&345&1 CFBN CFBNˡ ,#&45 TDPSFT SFUVSO CFBN<>

  5. 4UFQ(FOFSBUJOHUIFDBOEJEBUFTFU $BOEJEBUFTFU ˒ LCFTUJUFNT JF OPSNBMUPQL ˖ TJOHMFCFTUUPLFO ˠ VOGJMMFEDPOTUSBJOUT

     #FBNTJ[F L  7PDBCVMBSZTJ[F c7c  $POTUSBJOUT $  XPSET QISBTF

  6. 4UFQ"MMPDBUJOHUIFCFBN "MMPDBUFTJ[FLCFBNBDSPTT$ DPOTUSBJOUCBOLT #BOL 5IFQPSUJPOPGUIFCFBNSFTFSWFEGPSJUFNT IBWJOHNFUUIFTBNFOVNCFSPGDPOTUSBJOUT "MMPDBUJPOTUSBUFHZTFUUJOHFBDICJOTJ[FUP⌊L$⌋ "OZSFNBJOJOHTMPUTBSFBTTJHOFEUPUIFNPTUDPOTUSBJOFECBOL 

  7. &YQFSJNFOUBM4FUVQ • %BUB • 5SBJO8.5&OHMJTI(FSNBOUSBJOJOHDPSQPSB • %FW5FTUOFXTUFTU  • 1SFQSPDFTT

    • .PTFTUPLFOJ[FS • +PJOU#1&WPDBCVMBSZXJUILNFSHFPQFSBUJPO • .PEFM • MBZFS3//XJUIBUUFOUJPO • "EBNPQUJNJ[FSXJUIBCBUDITJ[FPG • &BSMZTUPQQJOH DSPTTFOUSPQZ • #BTFMJOFT • (SJE#FBN4FBSDI LC $  C   

  8. 3VOOJOHUJNF MPXFSJTCFUUFS BOE#-&6TDPSF • 5IFMJOFBSUSFOEJO$ DPOTUSBJOUT JTDMFBSGPS(#4 • 5IFDPOTUBOUUSFOEJO$ JTDMFBSGPS%#"

    • %#"JTBCPVUYTMPXFSUIBOVODPOTUSBJOFEEFDPEJOH  #-&6 #-&6 #-&6

  9. $POTUSBJOFEXPSETBSFQMBDFEJOUIFDPSSFDUQPTJUJPO 1FBSTPO`TS  

  10. $POTUSBJOFEXPSETBSFQMBDFEJOUIFDPSSFDUQPTJUJPO 

  11. .PEFMTDPSFHPFTEPXOCVU#-&6TDPSFHPFTVQ 5IJTJTBOPCTFSWBUJPOSFMBUFEUPBSFQPSU UIBU #-&6XPSTFOTJGUIFCFBNTJ[FJTNBEFUPPMBSHF  ,PFIOBOE,OPXMFT 8/.5 4JY$IBMMFOHFTGPS/FVSBM.BDIJOF5SBOTMBUJPO

  12. 'BTU-FYJDBMMZ$POTUSBJOFE%FDPEJOHXJUI%#"GPS/.5 • 5IJTTUVEZNBEFJUQPTTJCMFUPDPOUSPM/.5PVUQVU VTJOH-FYJDBMMZ$POTUSBJOFE%FDPEJOH • *O(#4 UIFDPNQVUBUJPOBMDPTUJODSFBTFEMJOFBSMZ XJUI DPOTUSBJOUT CVUJO%#"

    JUEFDSFBTFEUPDPOTUBOU PSEFS • $POTUSBJOUTBSFDPSSFDUMZQMBDFEXPBMJHONFOUJOGPSNBUJPO