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

Figure out exact PES matching #11

Closed
3 tasks
tbekolay opened this issue Aug 20, 2018 · 2 comments
Closed
3 tasks

Figure out exact PES matching #11

tbekolay opened this issue Aug 20, 2018 · 2 comments

Comments

@tbekolay
Copy link
Member

The PES learning rule in the emulator and hardware both differ from the Nengo core implementation in some ways. It would be good to get a better understanding of what is different so that we can emulate it better, and choose better constants during the build process.

Specifically, we should:

  • Figure out the exact Nengo core -> hardware learning rate scaling
  • Figure out the exact hardware -> emulator learning rate scaling
  • Vary other parameters (e.g. pre_synapse) and ensure that they work on Loihi the same as they work in Nengo core
@tbekolay
Copy link
Member Author

Also a note that in order to match the PES rule in all cases, it may be necessary to generate microcode values on the fly. We currently hardcode them.

@drasmuss
Copy link
Member

This was largely resolved in #139, I opened an issue to look at the pre_synapse behaviour specifically (#179).

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

No branches or pull requests

2 participants