Alternating Synthetic and Real Gradients for Neural Language Modeling
Training recurrent neural networks (RNNs) with backpropagation through time (BPTT) has known drawbacks such as being difficult to capture longterm dependencies in sequences. Successful alternatives to BPTT have not yet been discovered. Recently, BP with synthetic gradients by a decoupled neural inte...
Saved in:
| Published in | arXiv.org |
|---|---|
| Main Authors | , |
| Format | Paper |
| Language | English |
| Published |
Ithaca
Cornell University Library, arXiv.org
03.06.2022
|
| Subjects | |
| Online Access | Get full text |
Cover
Loading…
| Abstract | Training recurrent neural networks (RNNs) with backpropagation through time (BPTT) has known drawbacks such as being difficult to capture longterm dependencies in sequences. Successful alternatives to BPTT have not yet been discovered. Recently, BP with synthetic gradients by a decoupled neural interface module has been proposed to replace BPTT for training RNNs. On the other hand, it has been shown that the representations learned with synthetic and real gradients are different though they are functionally identical. In this project, we explore ways of combining synthetic and real gradients with application to neural language modeling tasks. Empirically, we demonstrate the effectiveness of alternating training with synthetic and real gradients after periodic warm restarts on language modeling tasks. |
|---|---|
| AbstractList | Training recurrent neural networks (RNNs) with backpropagation through time (BPTT) has known drawbacks such as being difficult to capture longterm dependencies in sequences. Successful alternatives to BPTT have not yet been discovered. Recently, BP with synthetic gradients by a decoupled neural interface module has been proposed to replace BPTT for training RNNs. On the other hand, it has been shown that the representations learned with synthetic and real gradients are different though they are functionally identical. In this project, we explore ways of combining synthetic and real gradients with application to neural language modeling tasks. Empirically, we demonstrate the effectiveness of alternating training with synthetic and real gradients after periodic warm restarts on language modeling tasks. |
| Author | Shang, Fangxin Zhang, Hao |
| Author_xml | – sequence: 1 givenname: Fangxin surname: Shang fullname: Shang, Fangxin – sequence: 2 givenname: Hao surname: Zhang fullname: Zhang, Hao |
| BookMark | eNqNjr0KwjAUhYMoWLXvcMG5UJO21lGk6qAO6l6Cva0t4UbzM_j2ZvABnA4cvsN3ZmxMmnDEIi7EKikzzqcstnZI05QXa57nImLVVjk0JF1PHdw-5J7o-gdIauCKUsHByKZHchZabeCC3oTyJKnzskM46wZVWC7YpJXKYvzLOVvuq_vumLyMfnu0rh60DxZla74qi004VGTiP-oLuD48bQ |
| ContentType | Paper |
| Copyright | 2022. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
| Copyright_xml | – notice: 2022. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
| DBID | 8FE 8FG ABJCF ABUWG AFKRA AZQEC BENPR BGLVJ CCPQU DWQXO HCIFZ L6V M7S PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
| DatabaseName | ProQuest SciTech Collection ProQuest Technology Collection Materials Science & Engineering Collection ProQuest Central (Alumni) ProQuest Central ProQuest Central Essentials ProQuest Central Technology Collection ProQuest One Community College ProQuest Central SciTech Premium Collection ProQuest Engineering Collection Engineering Database Publicly Available Content Database ProQuest One Academic Eastern Edition (DO NOT USE) ProQuest One Academic ProQuest One Academic UKI Edition ProQuest Central China Engineering Collection |
| DatabaseTitle | Publicly Available Content Database Engineering Database Technology Collection ProQuest Central Essentials ProQuest One Academic Eastern Edition ProQuest Central (Alumni Edition) SciTech Premium Collection ProQuest One Community College ProQuest Technology Collection ProQuest SciTech Collection ProQuest Central China ProQuest Central ProQuest Engineering Collection ProQuest One Academic UKI Edition ProQuest Central Korea Materials Science & Engineering Collection ProQuest One Academic Engineering Collection |
| DatabaseTitleList | Publicly Available Content Database |
| Database_xml | – sequence: 1 dbid: 8FG name: ProQuest Technology Collection url: https://search.proquest.com/technologycollection1 sourceTypes: Aggregation Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Physics |
| EISSN | 2331-8422 |
| Genre | Working Paper/Pre-Print |
| GroupedDBID | 8FE 8FG ABJCF ABUWG AFKRA ALMA_UNASSIGNED_HOLDINGS AZQEC BENPR BGLVJ CCPQU DWQXO FRJ HCIFZ L6V M7S M~E PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
| ID | FETCH-proquest_journals_21869233643 |
| IEDL.DBID | BENPR |
| IngestDate | Thu Oct 10 16:20:33 EDT 2024 |
| IsOpenAccess | true |
| IsPeerReviewed | false |
| IsScholarly | false |
| Language | English |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-proquest_journals_21869233643 |
| OpenAccessLink | https://www.proquest.com/docview/2186923364?pq-origsite=%requestingapplication% |
| PQID | 2186923364 |
| PQPubID | 2050157 |
| ParticipantIDs | proquest_journals_2186923364 |
| PublicationCentury | 2000 |
| PublicationDate | 20220603 |
| PublicationDateYYYYMMDD | 2022-06-03 |
| PublicationDate_xml | – month: 06 year: 2022 text: 20220603 day: 03 |
| PublicationDecade | 2020 |
| PublicationPlace | Ithaca |
| PublicationPlace_xml | – name: Ithaca |
| PublicationTitle | arXiv.org |
| PublicationYear | 2022 |
| Publisher | Cornell University Library, arXiv.org |
| Publisher_xml | – name: Cornell University Library, arXiv.org |
| SSID | ssj0002672553 |
| Score | 3.399466 |
| SecondaryResourceType | preprint |
| Snippet | Training recurrent neural networks (RNNs) with backpropagation through time (BPTT) has known drawbacks such as being difficult to capture longterm dependencies... |
| SourceID | proquest |
| SourceType | Aggregation Database |
| SubjectTerms | Back propagation Modelling Recurrent neural networks Training |
| Title | Alternating Synthetic and Real Gradients for Neural Language Modeling |
| URI | https://www.proquest.com/docview/2186923364 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV1NS8NAFHzYBMGbn_hRy4JeF2W_kpxEJWkRW0pV6K1s0o0XiTWJBy_-dt8LWz0IPYVNYEM2y9uZ2UkG4LKgWNVII1MtheVKWM0TJ0seW2cjKRM8kt4xnpjRi3qY67kX3Bpvq1zXxK5QL98L0sivuuwkIaVRN6sPTqlRtLvqIzR6EApkCtcBhHfpZDr7VVmEiRAzy3-Ftls9sl0Ip3bl6j3YctU-bHemy6I5gPT2zctx1St7-qoQi-FrZMjt2QzxGxvWnR-rbRgiS0a_0cCTj15gZJRiRt-SH8JFlj7fj_j63gs_P5rF39PIIwiQ6LtjYEIvtbI2VxqpIsLJJHZlaROspkaaQucn0N_U0-nmy2ewI8i6TwqC7EPQ1p_uHBfUNh9AL86GAz922Bp_pz-dVICL |
| link.rule.ids | 783,787,12777,21400,33385,33756,43612,43817 |
| linkProvider | ProQuest |
| linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV1LSwMxEB60RezNJz6qBvQahDx2u6ci0u2q2yJaobclu816kW3dXQ_-e2dCqgehp0ACCXkw882XST6Am4JkVUONkWopDFfCaB5ZWfKBsSaUMsKS-I7JNEje1ONczz3h1vi0yrVNdIZ6sSyII7912klCykANV5-cVKPodtVLaGxDV0n01fRSPB7_ciwiCBExy39m1vmOeA-6z2Zl633YstUB7LiUy6I5hNHdhyfjqnf2-l0hEsNNZBjZsxdEb2xcu2ystmGIKxl9ooGVqacXGWmY0UvyI7iOR7P7hK_HzvzpaLK_uchj6GCYb0-ACb3QyphcaQwUEUxGA1uWJkJbGsig0Pkp9Df1dLa5-Qp2k9kkzdKH6dM59AQl8ROXIPvQaesve4Gutc0v3fr9AJIBf_8 |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Alternating+Synthetic+and+Real+Gradients+for+Neural+Language+Modeling&rft.jtitle=arXiv.org&rft.au=Shang%2C+Fangxin&rft.au=Zhang%2C+Hao&rft.date=2022-06-03&rft.pub=Cornell+University+Library%2C+arXiv.org&rft.eissn=2331-8422 |