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

Integration of new deconvolution implementation based on the REVD2 algorithm (Ongoing) #1263

Draft
wants to merge 13 commits into
base: feature/deconv
Choose a base branch
from

Conversation

hleblevec
Copy link
Contributor

Integration of the new deconvolution implementation (see https://github.com/Xilinx/finn-hlslib/blob/feature/deconv/deconv.hpp for the HLS part).

Current known issues:

  • Several failing cppsim testcases, which seem to be linked to padding when P < K-S
  • rtlsim currently not working, probably due to issues in the templates
  • Some testcases might fail due to loop timeout value being too small (need to identify if there are still cases where that happens)

@auphelia
Copy link
Collaborator

Hi @hleblevec ,

Thank you for starting with this PR! I wanted to quickly point you to this new template which has an easier setup to use the time out simulation: https://github.com/Xilinx/finn/blob/feature/deconv/src/finn/custom_op/fpgadataflow/templates.py#L63

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

Successfully merging this pull request may close these issues.

2 participants