Towards a Flexible and High-Fidelity Approach to Distributed DNN Training Emulation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914785311850496 |
|---|---|
| author | Liu, Banruo Ojewale, Mubarak Adetunji Ding, Yuhan Canini, Marco |
| author_facet | Liu, Banruo Ojewale, Mubarak Adetunji Ding, Yuhan Canini, Marco |
| contents | We propose NeuronaBox, a flexible, user-friendly, and high-fidelity approach to emulate DNN training workloads. We argue that to accurately observe performance, it is possible to execute the training workload on a subset of real nodes and emulate the networked execution environment along with the collective communication operations. Initial results from a proof-of-concept implementation show that NeuronaBox replicates the behavior of actual systems with high accuracy, with an error margin of less than 1% between the emulated measurements and the real system. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_02969 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards a Flexible and High-Fidelity Approach to Distributed DNN Training Emulation Liu, Banruo Ojewale, Mubarak Adetunji Ding, Yuhan Canini, Marco Machine Learning Distributed, Parallel, and Cluster Computing We propose NeuronaBox, a flexible, user-friendly, and high-fidelity approach to emulate DNN training workloads. We argue that to accurately observe performance, it is possible to execute the training workload on a subset of real nodes and emulate the networked execution environment along with the collective communication operations. Initial results from a proof-of-concept implementation show that NeuronaBox replicates the behavior of actual systems with high accuracy, with an error margin of less than 1% between the emulated measurements and the real system. |
| title | Towards a Flexible and High-Fidelity Approach to Distributed DNN Training Emulation |
| topic | Machine Learning Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2405.02969 |