How to check smdistributed-modelparallel version?

0

According to the doc ( https://sagemaker.readthedocs.io/en/stable/api/training/smd_model_parallel_general.html ), there are different parameters depending on the version of smdistributed-modelparallel module / package. However, I am unable to find a way to check the version (e.g. via sagemaker python SDK) or just from the training container documentation (e.g. https://github.com/aws/deep-learning-containers/blob/master/available_images.md#huggingface-training-containers ).

Any idea?

Thanks!

asked 2 years ago406 views
1 Answer
0
Accepted Answer

Have not yet found a programmatic way to check the version.

However, for each DLC (Deep Learning Container) available at https://github.com/aws/deep-learning-containers/blob/master/available_images.md , we can look at the corresponding docker build files.

E.g. for PyTorch 1.10.2 with HuggingFace transformers DLC, the corresponding dockerfile is here: https://github.com/aws/deep-learning-containers/blob/master/huggingface/pytorch/training/docker/1.10/py3/cu113/Dockerfile.gpu

And we can see that the version: smdistributed_modelparallel-1.8.1-cp38-cp38-linux_x86_64.whl.

answered 2 years ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions