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

[Question] A question about epoch visualization during the training process #339

Open
2 of 3 tasks
CrazyTrumpet10 opened this issue May 29, 2024 · 1 comment
Open
2 of 3 tasks
Labels
question Further information is requested

Comments

@CrazyTrumpet10
Copy link

CrazyTrumpet10 commented May 29, 2024

Required prerequisites

Questions

epoch_1
1716970569230
Hello, my research direction is path planning, I would like to ask how to render the visualization of each epoch trajectory image, check the description of the document does not seem to find the relevant instructions, look forward to your reply, thank you very much!
Every time I render the graph, it is the initialized environment with no trajectories, I also try to get the coordinates of the intelligences for each epoch, and I can get them, but I can't render them in the environment graph, and I would like to ask how to render the trajectories properly?

@CrazyTrumpet10 CrazyTrumpet10 added the question Further information is requested label May 29, 2024
@CrazyTrumpet10 CrazyTrumpet10 changed the title [Question] Hello, my research direction is path planning, I would like to ask how to render the visualization of each epoch trajectory image, check the description of the document does not seem to find the relevant instructions, look forward to your reply, thank you very much! [Question] A question about epoch visualization during the training process May 29, 2024
@Gaiejj
Copy link
Member

Gaiejj commented Jun 4, 2024

I've grasped your requirement, but currently, OmniSafe doesn't offer an interface for visualizing during the training process or saving rendered videos after each epoch. We acknowledge this as a valuable feature and intend to support it in future updates.

In the meantime, to address your current needs, you might consider saving the model weights at the end of each epoch.
By setting save_model_freq in the algorithm yaml file.

Post-training, you can individually load these saved weights and proceed with visualizations. This approach allows you to retrospectively analyze the progress and performance of the model at every stage.

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

No branches or pull requests

2 participants