{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:45:14Z","timestamp":1782798314617,"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"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,5,18]]},"DOI":"10.1109\/infocom59046.2026.11571650","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":["Accelerating Multi-modal LLM Training with Adaptive Model Placement and Parallelization"],"prefix":"10.1109","author":[{"given":"Yiming","family":"Yin","sequence":"first","affiliation":[{"name":"The Hong Kong University of Science and Technology (Guangzhou),Data Science and Analytics Thrust,Guangzhou,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaohuai","family":"Shi","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology,School of Computer Science and Technology,Shenzhen,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qiang","family":"Wang","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology,School of Computer Science and Technology,Shenzhen,China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaowen","family":"Chu","sequence":"additional","affiliation":[{"name":"The Hong Kong University of Science and Technology (Guangzhou),Data Science and Analytics Thrust,Guangzhou,China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Gpt-4o system card","author":"Hurst","year":"2024"},{"key":"ref2","article-title":"Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context","author":"Georgiev","year":"2024"},{"key":"ref3","article-title":"Qwen2. 5-vl technical report","author":"Bai","year":"2025"},{"key":"ref4","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"International Conference on machine learning","author":"Radford"},{"key":"ref5","first-page":"28 492","article-title":"Robust speech recognition via large-scale weak supervision","volume-title":"International Conference on machine learning","author":"Radford"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref7","article-title":"Deepseek-vl2: Mixture-of-experts vision-language models for advanced multimodal understanding","author":"Wu","year":"2024"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3718958.3750472"},{"key":"ref9","article-title":"Optimus: Accelerating large-scale multi-modal llm training by bubble exploitation","author":"Feng","year":"2024"},{"key":"ref10","article-title":"Megatron-lm: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2019"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"ref12","article-title":"Large scale distributed deep networks","volume-title":"Advances in neural information processing systems","author":"Dean"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.1811.06965"},{"key":"ref14","first-page":"1157","article-title":"DISTMM: Accelerating distributed multimodal model training","volume-title":"21st USENIX Symposium on Networked Systems Design and Implementation","author":"Huang"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1002\/9780470114735.hawley04287"},{"key":"ref16","article-title":"Pipeweaver: Addressing data dynamicity in large multimodal model training with dynamic interleaved pipeline","author":"Xue","year":"2025"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref18","first-page":"19730","article-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","volume-title":"International Conference on machine learning","author":"Li"},{"key":"ref19","article-title":"Sphinx: The joint mixing of weights, tasks, and visual embeddings for multi-modal large language models","author":"Lin","year":"2023"},{"key":"ref20","article-title":"Qwen2. 5-omni technical report","author":"Xu","year":"2025"},{"key":"ref21","article-title":"Minigpt-4: Enhancing vision-language understanding with advanced large language models","author":"Zhu","year":"2023"},{"key":"ref22","article-title":"Qwen3 technical report","author":"Yang","year":"2025"},{"key":"ref23","article-title":"Deepseek-v3 technical report","year":"2024"},{"key":"ref24","first-page":"401","article-title":"Towards scalable distributed training of deep learning on public cloud clusters","volume-title":"Proceedings of Machine Learning and Systems","volume":"3","author":"Shi"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/MNET.011.2000530"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3437801.3441593"},{"key":"ref27","article-title":"Zero bubble pipeline parallelism","author":"Qi","year":"2023"},{"key":"ref28","first-page":"7937","article-title":"Memory-efficient pipeline-parallel dnn training","volume-title":"International Conference on Machine Learning","author":"Narayanan"},{"key":"ref29","first-page":"1","article-title":"Hanayo: Harnessing wave-like pipeline parallelism for enhanced large model training efficiency","volume-title":"Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis","author":"Liu"},{"key":"ref30","first-page":"1","article-title":"Efficient large-scale language model training on gpu clusters using megatron-lm","volume-title":"Proceedings of the international conference for high performance computing, networking, storage and analysis","author":"Narayanan"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3567955.3567959"},{"key":"ref32","first-page":"559","article-title":"Alpa: Automating inter-and intra-operator parallelism for distributed deep learning","volume-title":"16th USENIX Symposium on Operating Systems Design and Implementation","author":"Zheng"},{"key":"ref33","first-page":"341","article-title":"Reducing activation recomputation in large transformer models","volume-title":"Proceedings of Machine Learning and Systems","volume":"5","author":"Korthikanti"},{"key":"ref34","first-page":"12 888","article-title":"Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation","volume-title":"International Conference on machine learning","author":"Li"},{"key":"ref35","first-page":"101","article-title":"Diffusionpipe: Training large diffusion models with efficient pipelines","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Tian"},{"key":"ref36","article-title":"Orchestrate multimodal data with batch post-balancing to accelerate multimodal large language model training","author":"Zheng","year":"2025"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3676641.3715992"},{"key":"ref38","first-page":"347","article-title":"nnscaler: Constraint-guided parallelization plan generation for deep learning training","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation","author":"Lin"},{"key":"ref39","first-page":"563","article-title":"Metis: Fast automatic distributed training on heterogeneous GPUs","volume-title":"2024 USENIX Annual Technical Conference","author":"Um"}],"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\/11571650.pdf?arnumber=11571650","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:21:06Z","timestamp":1782796866000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11571650\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,18]]},"references-count":39,"URL":"https:\/\/doi.org\/10.1109\/infocom59046.2026.11571650","relation":{},"subject":[],"published":{"date-parts":[[2026,5,18]]}}}