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

Isolate moving average object into its own class #40

Open
p-ferreira opened this issue Jun 12, 2023 · 1 comment
Open

Isolate moving average object into its own class #40

p-ferreira opened this issue Jun 12, 2023 · 1 comment

Comments

@p-ferreira
Copy link
Contributor

The current implementation of the moving average is set as a tensor of the main self. Ideally, we would have the moving average isolated in an object with a resync function, similar to the dendrite_pool and gating_model approach.

@steffencruz
Copy link
Contributor

This would also potentially fix 2 other issues:

  1. Periodic logging of weights (default every 10 steps) actually logs the weights n_concurrent_foward times
  2. Thread-unsafe attribute setting in concurrent forwards

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