1 個回答
- 最新
- 最多得票
- 最多評論
0
At the top right of a notebook (near the kernel ) there will be a resource configurations button, you'll be able to choose the instance you want to run the notebook on.
A nice feature of that is that all the instances shares the same EFS mount (SageMaker studio uses EFS for notebook storage) if you save a dataframe to the local disk (EFS) you can change instance type during your work and continue from the place you've been in (Move from a GPU instance to a CPU instance for cost effectiveness / back to GPU for performance)
已回答 4 年前
相關內容
- 已提問 9 個月前
- AWS 官方已更新 1 年前
- AWS 官方已更新 1 年前
- AWS 官方已更新 1 年前
- AWS 官方已更新 6 個月前