yexinru/doc-exports
1
0
Fork 0
forked from docs/doc-exports
Code Pull Requests Activity
Files
fix_python
doc-exports/docs/modelarts/umn/modelarts_05_3147.html
Lai, Weijian 6aa966a79a ModelArts UMN 24.3.0 version 
Reviewed-by: Pruthi, Vineet <vineet.pruthi@t-systems.com>
Co-authored-by: Lai, Weijian <laiweijian4@huawei.com>
Co-committed-by: Lai, Weijian <laiweijian4@huawei.com>
2024-11-02 09:04:52 +00:00

1.1 KiB
Raw Permalink Blame History

Why Is the Training Speed Similar When Different Notebook Flavors Are Used?

If your training job is single-process in code, the training speed is basically the same no matter when the notebook flavor of 8 vCPUs and 64 GB of memory or the flavor of 72 vCPUs and 512 GB of memory is used. For example, if your training job uses 2 vCPUs and 4 GB of memory, the training speed is similar no matter when you use the notebook flavor of 4 vCPUs and 8 GB of memory or the flavor of 8 vCPUs and 64 GB of memory.

If your training job is multi-process in code, the training speed backed by the notebook flavor of 72 vCPUs and 512 GB of memory is higher than that backed by the notebook flavor of 8 vCPUs and 64 GB of memory.

Parent topic: Others