Samfuran Jeri-zuwa-Jeri
Samfuran jeri-zuwa-jeri taswirar jeri ɗaya zuwa wani na yuwuwar tsayi daban-daban, kamar fassarar jumla ko taƙaita takarda.
Dubawa
They introduced the encoder-decoder design and the attention mechanism that paved the way for the Transformer.
Zurfafa nutsewa
Samfurin jeri-zuwa-jeri (seq2seq) yana da sassa biyu: na'ura mai ɓoyewa wanda ke karanta jerin abubuwan shigar da matse ma'anarsa, da na'urar tantancewa da ke samar da jerin abubuwan fitarwa alama ɗaya a lokaci guda. Alamar aikin 2014 ta Sutskever, Vinyals, da Le sun yi amfani da LSTMs masu tarin yawa don fassarar inji. Wani rauni ya bayyana: cusa jimla gabaɗaya zuwa ƙayyadaddun vector guda ɗaya ya rasa bayanai kan dogon bayanai. A cikin 2015 Bahdanau ya gabatar da hankali, yana barin mai yankewa ya waiwaya baya a duk jihohin da ke ɓoye kuma ya mai da hankali kan waɗanda suka fi dacewa ga kowace kalma fitarwa. Wannan ya warware matsalar kuma ya inganta fassarar sosai. Manufar gabaɗaya ga kowane aikin rubutu-zuwa-fitarwa kuma kai tsaye ya zaburar da cikakken tsarin gine-ginen Transformer a cikin 2017.
Fahimtar Fasaha
Mai rikodin yana samar da jeri na ɓoyayyun jihohi; mai ƙididdigewa yana samar da kayan aiki kai tsaye, daidaitacce akan abubuwan da suka gabata da mahallin mahallin. Hankali yana ƙididdige jimlar ma'auni na jahohin maɓalli ta amfani da maki masu daidaitawa, don haka kowane matakin yanke hukunci yana zana yanayin mahallin al'ada. Wannan yana cire tsayin fitarwa daga vector guda ɗaya kuma yana ba da daidaitawa mai laushi tsakanin shigarwa da matsayi na fitarwa, wanda kuma ana iya fassara shi azaman tushen kalmomin da suka jagoranci kowace kalma da aka fassara.
Dabarun Tasiri
Gudu da sikelin
Gudun aikin harshe na iya tafiya da sauri ba tare da sadaukar da daidaito ba.
Shiga ku isa
Yana faɗaɗa damar shiga cikin harsuna da salon sadarwa.
Shawarwari masu haske
Ƙungiyoyi za su iya ciyar da ƙarin lokaci akan hukunci yayin da aiki da kai ke sarrafa maimaitawa.
Makomar Samfuran Jeri-zuwa-Jeri
Seq2seq na zamani ya mamaye samfuran masu canza sheka kamar T5 da BART, waɗanda ke tsara kusan kowane aikin NLP azaman rubutu-zuwa rubutu. Seq2seq na tushen RNN babban tarihi ne, amma tsarin mai rikodin rikodin yana bunƙasa cikin fassarar, taƙaitawa, da fahimtar magana. Yi tsammanin ci gaba da haɓakawa a cikin tsarin seq2seq na harsuna da yawa da multimodal, tare da ingantacciyar ribar daga na'urori marasa autoregressive da distilled waɗanda ke fitar da fitarwa cikin sauri yayin kiyaye inganci.
Aiwatar da Gaskiyar Duniya
Tsarin fassarar inji yana juya jimlolin Turanci zuwa Faransanci ko Jafananci.
Takaitaccen rubutu wanda ke sake rubuta dogayen labarai zuwa gajerun takaitattun bayanai.
Ƙididdigar magana taswirar tsarin siginar sauti mai jiwuwa zuwa rubutun rubutu.
Chatbot da tsarin tattaunawa waɗanda ke taswirar furcin mai amfani ga amsa da aka samar.
Hatsari & Tsare-tsare
Abubuwan da aka ruɗe suna iya shigar da rahotanni cikin nutsuwa, kwararar tallafi, ko abubuwan bincike.
Hankali na gaggawa na iya ƙirƙirar sakamako mara daidaituwa a cikin buƙatun iri ɗaya.
Za a iya fallasa bayanan rubutu mai ma'ana idan ikon samun dama yana da rauni.
Taswirar Hanya
Ƙayyade tsarin fitarwa, sautin, da ma'auni masu inganci kafin fitowa.
Amsa a ƙasa tare da amintattun tushe a duk lokacin da daidaito ya shafi mahimmanci.
Ajiye wurin binciken ɗan adam don abubuwan da ake samu masu girma.
Bibiyar tsarin gazawar kuma sake horar da tsokaci ko tafiyar aiki akai-akai.
Ci gaba da Bincike
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Sequence-to-Sequence Models quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Jagora na gaba
Tilasta Malami a cikin Samfuran Jeri
Tambayoyin da ake yawan yi
What is Sequence-to-Sequence Models?
Samfuran jeri-zuwa-jeri taswirar jeri ɗaya zuwa wani na yuwuwar tsayi daban-daban, kamar fassarar jumla ko taƙaita takarda. Sun gabatar da ƙirar encoder-decoder da tsarin kulawa wanda ya share hanyar Transformer.
Menene manyan abubuwa guda biyu na samfurin jeri-zuwa-jeri?
Encoder yana karantawa kuma yana matsawa abin da aka shigar, kuma mai yankewa yana haifar da jerin abubuwan fitarwa.
Wace babbar matsala ce hanyar kulawa ta warware a cikin seq2seq model?
Hankali bari mai rikodin ya sami damar shiga duk jihohin rikodi maimakon dogaro da matsi guda ɗaya, yana daidaita aikin dogon jumla.
Wane gine-ginen ainihin ƙirar fassarar seq2seq ta 2014 ta yi amfani da ita?
Sutskever et al. da aka yi amfani da rukunonin cibiyoyin sadarwa na LSTM na yau da kullun don encoder da dikodi.
Ta yaya hankali zai gina mahallin kowane mataki na yanke hukunci?
Hankali yana ba da ma'auni don sanya jihohi don haka kowane matakin fitarwa ya mai da hankali kan wuraren shigarwar da suka fi dacewa.
Me yasa abubuwan seq2seq zasu iya bambanta da tsayi da abubuwan shigarwa?
Mai ƙididdigewa yana haifar da jujjuyawar kai tsaye kuma yana iya tsayawa a alamar ƙarshen-jeri, yana ƙyale tsayin fitarwa mai canzawa.