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

Add more function to the TensorRT conversion CI/CD test #422

Open
binliunls opened this issue May 14, 2023 · 3 comments
Open

Add more function to the TensorRT conversion CI/CD test #422

binliunls opened this issue May 14, 2023 · 3 comments
Assignees

Comments

@binliunls
Copy link
Contributor

Is your feature request related to a problem? Please describe.
Currently the CI/CD test of TensorRT conversion only contains limited input parameters. And the test performs on a MONAI docker instead of the TensorRT docker, which cannot track the newest version of TensorRT.
Describe the solution you'd like

  1. It would be better to add all input parameters to the test and assign a None as the default value.
  2. It would be better to find a way to test the given models on the newest TensorRT docker.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

@yiheng-wang-nv
Copy link
Collaborator

Hi @binliunls , thanks for posting this ticket, we do need a more powerful ci test on TensorRT. Let's discuss first.

@yiheng-wang-nv
Copy link
Collaborator

I'm working on this PR: #400 which I expect for each bundle, all config files and commands (include TensorRT test) should be tested in the future.

cc @Nic-Ma

@yiheng-wang-nv
Copy link
Collaborator

yiheng-wang-nv commented May 30, 2023

I'm working on this PR: #400 which I expect for each bundle, all config files and commands (include TensorRT test) should be tested in the future.

cc @Nic-Ma

Update: the unit tests will focus on runnable configs first (to ensure tests can run quickly). As for the TensorRT test, I will submit a separate PR to enhance the current weekly TensorRT tests and address the following feature request:

It would be better to add all input parameters to the test and assign a None as the default value.

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

No branches or pull requests

2 participants