You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
Currently CPP GPU takes around 72 minutes to finish all its trainning tasks and it is not ideal for a PR CI test to handle that. Essentially the CI to verify the PR is to ensure the functionalities didn't have any problems. The performance test ideally should be placed under the nightly tests. As @leleamol recommended, some improvement can be taken in here such as the number of epoch to be set to 1. This would bring down the runtime for this test to under 30 minutes and save time.
URL:
http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-gpu/detail/PR-13914/3/pipeline/278
Currently CPP GPU takes around 72 minutes to finish all its trainning tasks and it is not ideal for a PR CI test to handle that. Essentially the CI to verify the PR is to ensure the functionalities didn't have any problems. The performance test ideally should be placed under the nightly tests. As @leleamol recommended, some improvement can be taken in here such as the number of epoch to be set to 1. This would bring down the runtime for this test to under 30 minutes and save time.
@marcoabreu
The text was updated successfully, but these errors were encountered: