{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T22:32:12Z","timestamp":1768516332092,"version":"3.49.0"},"reference-count":21,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,12,14]],"date-time":"2025-12-14T00:00:00Z","timestamp":1765670400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,12,14]],"date-time":"2025-12-14T00:00:00Z","timestamp":1765670400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100018539","name":"Science and Technology Program","doi-asserted-by":"publisher","award":["KQTD20240729102154066,KJZD20240903104103005,KJZD20230923115113026"],"award-info":[{"award-number":["KQTD20240729102154066,KJZD20240903104103005,KJZD20230923115113026"]}],"id":[{"id":"10.13039\/501100018539","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,14]]},"DOI":"10.1109\/icpads67057.2025.11323016","type":"proceedings-article","created":{"date-parts":[[2026,1,14]],"date-time":"2026-01-14T20:36:54Z","timestamp":1768423014000},"page":"1-7","source":"Crossref","is-referenced-by-count":0,"title":["QPO: Accelerating Memory-Efficient DNN Training with Quantization and Pipelining"],"prefix":"10.1109","author":[{"given":"Xiang","family":"Fan","sequence":"first","affiliation":[{"name":"School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shaohuai","family":"Shi","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1038\/d41586-023-00816-5"},{"key":"ref2","author":"Guo","year":"2025","journal-title":"Seed1. 5-vl technical report[J]"},{"key":"ref3","author":"Bai","year":"2023","journal-title":"Qwen technical report[J]"},{"key":"ref4","author":"Touvron","year":"2023","journal-title":"Llama: Open and efficient foundation language models[J]"},{"key":"ref5","author":"Koroteev","year":"2021","journal-title":"BERT: a review of applications in natural language processing and understanding[J]"},{"key":"ref6","author":"Wu","year":"2024","journal-title":"TBA: Faster Large Language Model Training Using SSD-Based Activation Offloading[J]"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO.2016.7783721"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2018.8341972"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3178487.3178491"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3243904"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3373376.3378505"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA.2018.00017"},{"key":"ref13","first-page":"27434","article-title":"Ac-gc: Lossy activation compression with guaranteed convergence[C]","volume":"34","author":"Evans","year":"2021","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00070"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA45697.2020.00080"},{"key":"ref16","first-page":"1803","article-title":"Actnn: Reducing training memory footprint via 2-bit activation compressed training[C]","volume-title":"International Conference on Machine Learning.","author":"Chen","year":"2021"},{"key":"ref17","author":"Li","year":"2023","journal-title":"Aweq: Post-training quantization with activationweight equalization for large language models[J]"},{"key":"ref18","article-title":"EXACT: Scalable graph neural networks training via extreme activation compression[C]","volume-title":"International conference on learning representations.","author":"Liu","year":"2021"},{"key":"ref19","first-page":"14139","article-title":"Gact: Activation compressed training for generic network architectures[C]","volume-title":"International Conference on Machine Learning.","author":"Liu","year":"2022"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"ref23","author":"Shoeybi","year":"2019","journal-title":"Megatron-lm: Training multi-billion parameter language models using model parallelism"}],"event":{"name":"2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS)","location":"Hefei, China","start":{"date-parts":[[2025,12,14]]},"end":{"date-parts":[[2025,12,18]]}},"container-title":["2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11322805\/11322871\/11323016.pdf?arnumber=11323016","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T07:44:36Z","timestamp":1768463076000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11323016\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,14]]},"references-count":21,"URL":"https:\/\/doi.org\/10.1109\/icpads67057.2025.11323016","relation":{},"subject":[],"published":{"date-parts":[[2025,12,14]]}}}