Amazon SageMaker Built-in algorithms and Spot checkpointing

0

Which Amazon SageMaker built-in algorithms support checkpointing? In the documentation it says that:

SageMaker built-in algorithms and marketplace algorithms that do not checkpoint are currently limited to a MaxWaitTimeInSeconds of 3600 seconds (60 minutes).

However, in the algorithms I don't find any pointer to "checkpoint" or "spot". Can you help me out?

AWS
已提问 4 年前351 查看次数
1 回答
0
已接受的回答

This is the best resource that I've found to clarify this:

https://aws.amazon.com/blogs/aws/managed-spot-training-save-up-to-90-on-your-amazon-sagemaker-training-jobs/

Built-in algorithms: computer vision algorithms support checkpointing (Object Detection, Semantic Segmentation, and very soon Image Classification). As they tend to train on large data sets and run for longer than other algorithms, they have a higher likelihood of being interrupted. Other built-in algorithms do not support checkpointing for now.

Also:

Please note that TensorFlow uses checkpoints by default. For other frameworks, you’ll find examples in our sample notebooks and in the documentation.

已回答 4 年前

您未登录。 登录 发布回答。

一个好的回答可以清楚地解答问题和提供建设性反馈,并能促进提问者的职业发展。

回答问题的准则