/sagemaker-benchmarks

Benchmark scripts on SageMaker for various frameworks

Primary LanguagePython

Sage Maker Containers Benchmarks

Manual steps to start

Raise SM limits

Limits for P3.16 and C5.18 must be raised.

Create a bucket with datasources

Something like:

  • mxnet-bln-data-sagemaker/imagenet/raw/train-480px
  • mxnet-bln-data-sagemaker/imagenet/raw/validation-480px - imagenet validation set
  • mxnet-bln-data-sagemaker/imagenet/processed/validation-480px - idx/rec files with imagenet
  • mxnet-bln-data-sagemaker/small - small file just to make SM happy for the synthetic training

To stay out of troubles - sagemaker must be in the name.

Steps to get the results

The results are saved to AWS Cloud Watch We can use for example awslogs tool https://github.com/jorgebastida/awslogs