{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:45:39Z","timestamp":1782798339488,"version":"3.54.5"},"reference-count":39,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T00:00:00Z","timestamp":1779062400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,5,18]],"date-time":"2026-05-18T00:00:00Z","timestamp":1779062400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,5,18]]},"DOI":"10.1109\/infocom59046.2026.11571256","type":"proceedings-article","created":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T19:38:15Z","timestamp":1782761895000},"page":"1-10","source":"Crossref","is-referenced-by-count":0,"title":["A Self-updating Checkpointing and Fast Failure Recovery System for Distributed LLM Training"],"prefix":"10.1109","author":[{"given":"Leyi","family":"Ye","sequence":"first","affiliation":[{"name":"Fudan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhiyi","family":"Yao","sequence":"additional","affiliation":[{"name":"Fudan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Boliang","family":"Liu","sequence":"additional","affiliation":[{"name":"Fudan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuedong","family":"Xu","sequence":"additional","affiliation":[{"name":"Fudan University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liang","family":"Du","sequence":"additional","affiliation":[{"name":"China Unicom Greater Bay Area Innovation Institute"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zeng","family":"Chuxuan","sequence":"additional","affiliation":[{"name":"China Unicom Greater Bay Area Innovation Institute"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ling","family":"Deng","sequence":"additional","affiliation":[{"name":"China Unicom Greater Bay Area Innovation Institute"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jessie Hui","family":"Wang","sequence":"additional","affiliation":[{"name":"Tsinghua University"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Language models are few-shot learners","author":"Brown","year":"2020","journal-title":"NeurIPS"},{"key":"ref2","article-title":"GPT-4 technical report","volume-title":"CoRR","volume":"abs\/2303.08774","year":"2023"},{"key":"ref3","article-title":"Llama: Open and efficient foundation language models","volume":"abs\/2302.13971","author":"Team","year":"2023","journal-title":"CoRR"},{"key":"ref4","article-title":"The Llama 3 Herd of Models","year":"2024"},{"key":"ref5","first-page":"745","article-title":"Megascale: Scaling large language model training to more than 10, 000 gpus","volume-title":"NSDI 2024, Santa Clara, CA, April 15-17, 2024","author":"Jiang"},{"key":"ref6","first-page":"709","article-title":"Characterization of large language model development in the datacenter","author":"Hu","year":"2024","journal-title":"NSDI 2024, Santa Clara, CA, April 15-17, 2024"},{"key":"ref7","article-title":"OPT: open pre-trained transformer language models","volume":"abs\/2205.01068","author":"Zhang","year":"2022","journal-title":"CoRR"},{"key":"ref8","first-page":"203","article-title":"Checkfreq: Frequent, fine-grained DNN checkpointing","volume-title":"FAST 2021, February 23-25, 2021.","author":"Mohan","year":"2021"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613145"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3627703.3650085"},{"key":"ref11","first-page":"929","article-title":"Check-n-run: a checkpointing system for training deep learning recommendation models","volume-title":"NSDI 2022, Renton, WA, USA, April 4-6, 2022","author":"Eisenman"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3689031.3696088"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3676641.3716006"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3669940.3707255"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.14778\/3685800.3685832"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS60910.2024.00015"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref18","article-title":"Reducing activation recomputation in large transformer models","volume":"abs\/2205.05198","author":"Korthikanti","year":"2022","journal-title":"CoRR"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613152"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/3694715.3695960"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00024"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.14778\/3611540.3611569"},{"key":"ref23","article-title":"The Technology Behind Bloom Training","year":"2022"},{"key":"ref24","article-title":"stage 1 and 2","author":"ai","year":"2025"},{"key":"ref25","article-title":"All_reduce instead of reduce_scatter","author":"ai","year":"2025"},{"key":"ref26","article-title":"Getting started with distributed checkpoint (dcp)","author":"Team","year":"2023"},{"key":"ref27","first-page":"835","article-title":"SuperBench: Improving cloud AI infrastructure reliability with proactive validation","author":"Xiong","year":"2024","journal-title":"USENIX ATC 2024."},{"key":"ref28","first-page":"55l","article-title":"Zero-offload: Democratizing billion-scale model training","volume-title":"ATC 2021, July 14-16, 2021","author":"Ren"},{"key":"ref29","article-title":"Moetion: Efficient and reliable checkpointing for mixture-of-experts models at scale","author":"Gandhi","year":"2024"},{"key":"ref30","article-title":"GPUDirect RDMA","year":"2025"},{"key":"ref31","first-page":"3505","article-title":"Deepspeed: System optimizations enable training deep learning models with over l00 billion parameters","volume-title":"KDD \u201920: The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Virtual Event, CA, USA, August 23-27, 2020","author":"Rasley"},{"key":"ref32","article-title":"Llama 2: Open foundation and fine-tuned chat models","volume":"abs\/2307.09288","author":"Team","year":"2023","journal-title":"CoRR"},{"key":"ref33","article-title":"Chatglm: A family of large language models from glm-l30b to glm-4 all tools","author":"GLM","year":"2024"},{"key":"ref34","volume-title":"Wikimedia downloads","author":"Foundation","year":"2025"},{"key":"ref35","article-title":"Decoupled weight decay regularization","volume-title":"7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, May 6-9, 2019","author":"Loshchilov"},{"key":"ref36","article-title":"Boost Checkpoint Speed and Reduce Cost with Nebula","year":"2025"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3572848.3577510"},{"key":"ref38","first-page":"497","article-title":"Bamboo: Making preemptible instances resilient for affordable training of large dnns","volume-title":"NSDI 2023, Boston, MA, April 17-19, 2023","author":"Thorpe"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1145\/3694715.3695975"}],"event":{"name":"IEEE INFOCOM 2026 - IEEE Conference on Computer Communications","location":"Tokyo, Japan","start":{"date-parts":[[2026,5,18]]},"end":{"date-parts":[[2026,5,21]]}},"container-title":["IEEE INFOCOM 2026 - IEEE Conference on Computer Communications"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11571071\/11571169\/11571256.pdf?arnumber=11571256","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:26:49Z","timestamp":1782797209000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11571256\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,18]]},"references-count":39,"URL":"https:\/\/doi.org\/10.1109\/infocom59046.2026.11571256","relation":{},"subject":[],"published":{"date-parts":[[2026,5,18]]}}}