Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

NVIDIA/benchmarks/unet3d/implementations/mxnet-preview does not work #8

Open
kcoms555 opened this issue Jun 15, 2023 · 1 comment · May be fixed by #9
Open

NVIDIA/benchmarks/unet3d/implementations/mxnet-preview does not work #8

kcoms555 opened this issue Jun 15, 2023 · 1 comment · May be fixed by #9

Comments

@kcoms555
Copy link

kcoms555 commented Jun 15, 2023

Problems

At NVIDIA/benchmarks/unet3d/implementations/mxnet-preview,

  1. mlperf_logger.py file is missing - so an error "mlperf_logger is not found" is produced.
  2. when you build a Dockerfile and run a container, it will produce an error message such as "np.int is deprecated" or "np.bool" is deprecated. The version of numpy should be lower than 1.24.0(not exact. or maybe 1.23.3) at which not supporting such aliases of builtin types.

Solutions

  1. Copy and paste mlperf_logger.py from NVIDIA/benchmarks/unet3d/implementations/mxnet-22.04/mlperf_logger.py
  2. Set numpy version as 1.23.1 at which it supports aliases of builtin types. It can be done by editting Dockerfile or reinstalling it in a container.
kcoms555 added a commit to kcoms555/training_results_v2.1 that referenced this issue Jun 15, 2023
@kcoms555 kcoms555 linked a pull request Jun 15, 2023 that will close this issue
@arjunsuresh
Copy link

Preview submissions are not required to have the code shared in the results repository. So, this missing issue might be intentional.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants