INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS
The present application provides an information processing method and a communication apparatus. The method comprises: a participation node performs t-th model training to obtain first gradient information, and then determines first compression kernel information on the basis of gradient information...
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Patent |
| Sprache: | eng ; fre ; ger |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | XUE, Ye MA, Mengyao SU, Liqun LAU, Kin Nang |
| description | The present application provides an information processing method and a communication apparatus. The method comprises: a participation node performs t-th model training to obtain first gradient information, and then determines first compression kernel information on the basis of gradient information obtained by previous model training. The previous model training comprises the t-th model training and at least one model training before the t-th model training, and the gradient information obtained by the previous model training comprises the first gradient information. The participation node sends the first compression kernel information to a central node, and then receives first information from the central node. The first information is used for indicating first global compression kernel information, and the first global compression kernel information is used for compressing gradient information obtained after model training. The compression rate of the gradient information may be increased, and the transmission efficiency of the gradient information may be increased. |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_EP4488890A1</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>EP4488890A1</sourcerecordid><originalsourceid>FETCH-epo_espacenet_EP4488890A13</originalsourceid><addsrcrecordid>eNrjZLD09HPzD_J1DPH091MICPJ3dg0O9vRzV_B1DfHwd1Fw9HNRcPb39Q3183SGqHEMCHAMcgwJDeZhYE1LzClO5YXS3AwKbq4hzh66qQX58anFBYnJqXmpJfGuASYmFhYWlgaOhsZEKAEAbcIo2A</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS</title><source>esp@cenet</source><creator>XUE, Ye ; MA, Mengyao ; SU, Liqun ; LAU, Kin Nang</creator><creatorcontrib>XUE, Ye ; MA, Mengyao ; SU, Liqun ; LAU, Kin Nang</creatorcontrib><description>The present application provides an information processing method and a communication apparatus. The method comprises: a participation node performs t-th model training to obtain first gradient information, and then determines first compression kernel information on the basis of gradient information obtained by previous model training. The previous model training comprises the t-th model training and at least one model training before the t-th model training, and the gradient information obtained by the previous model training comprises the first gradient information. The participation node sends the first compression kernel information to a central node, and then receives first information from the central node. The first information is used for indicating first global compression kernel information, and the first global compression kernel information is used for compressing gradient information obtained after model training. The compression rate of the gradient information may be increased, and the transmission efficiency of the gradient information may be increased.</description><language>eng ; fre ; ger</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; PHYSICS</subject><creationdate>2025</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20250108&DB=EPODOC&CC=EP&NR=4488890A1$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,777,882,25545,76296</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20250108&DB=EPODOC&CC=EP&NR=4488890A1$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>XUE, Ye</creatorcontrib><creatorcontrib>MA, Mengyao</creatorcontrib><creatorcontrib>SU, Liqun</creatorcontrib><creatorcontrib>LAU, Kin Nang</creatorcontrib><title>INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS</title><description>The present application provides an information processing method and a communication apparatus. The method comprises: a participation node performs t-th model training to obtain first gradient information, and then determines first compression kernel information on the basis of gradient information obtained by previous model training. The previous model training comprises the t-th model training and at least one model training before the t-th model training, and the gradient information obtained by the previous model training comprises the first gradient information. The participation node sends the first compression kernel information to a central node, and then receives first information from the central node. The first information is used for indicating first global compression kernel information, and the first global compression kernel information is used for compressing gradient information obtained after model training. The compression rate of the gradient information may be increased, and the transmission efficiency of the gradient information may be increased.</description><subject>CALCULATING</subject><subject>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2025</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNrjZLD09HPzD_J1DPH091MICPJ3dg0O9vRzV_B1DfHwd1Fw9HNRcPb39Q3183SGqHEMCHAMcgwJDeZhYE1LzClO5YXS3AwKbq4hzh66qQX58anFBYnJqXmpJfGuASYmFhYWlgaOhsZEKAEAbcIo2A</recordid><startdate>20250108</startdate><enddate>20250108</enddate><creator>XUE, Ye</creator><creator>MA, Mengyao</creator><creator>SU, Liqun</creator><creator>LAU, Kin Nang</creator><scope>EVB</scope></search><sort><creationdate>20250108</creationdate><title>INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS</title><author>XUE, Ye ; MA, Mengyao ; SU, Liqun ; LAU, Kin Nang</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_EP4488890A13</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>eng ; fre ; ger</language><creationdate>2025</creationdate><topic>CALCULATING</topic><topic>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>XUE, Ye</creatorcontrib><creatorcontrib>MA, Mengyao</creatorcontrib><creatorcontrib>SU, Liqun</creatorcontrib><creatorcontrib>LAU, Kin Nang</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>XUE, Ye</au><au>MA, Mengyao</au><au>SU, Liqun</au><au>LAU, Kin Nang</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS</title><date>2025-01-08</date><risdate>2025</risdate><abstract>The present application provides an information processing method and a communication apparatus. The method comprises: a participation node performs t-th model training to obtain first gradient information, and then determines first compression kernel information on the basis of gradient information obtained by previous model training. The previous model training comprises the t-th model training and at least one model training before the t-th model training, and the gradient information obtained by the previous model training comprises the first gradient information. The participation node sends the first compression kernel information to a central node, and then receives first information from the central node. The first information is used for indicating first global compression kernel information, and the first global compression kernel information is used for compressing gradient information obtained after model training. The compression rate of the gradient information may be increased, and the transmission efficiency of the gradient information may be increased.</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | eng ; fre ; ger |
| recordid | cdi_epo_espacenet_EP4488890A1 |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING PHYSICS |
| title | INFORMATION PROCESSING METHOD AND COMMUNICATION APPARATUS |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-17T20%3A30%3A56IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=XUE,%20Ye&rft.date=2025-01-08&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3EEP4488890A1%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |