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

Fix bug of empty input in gelu.cu #800

Merged
merged 6 commits into from
Jun 27, 2021
Merged

Conversation

iProzd
Copy link
Collaborator

@iProzd iProzd commented Jun 26, 2021

#675 and #742 maybe the same bugs when:

  1. using gelu as activation function (GPU environment);
  2. there's one type (or types) of atom not showing in the system but in the type_map;
    this causes the empty input in gelu.cu and thus breaks.

@codecov-commenter
Copy link

codecov-commenter commented Jun 26, 2021

Codecov Report

Merging #800 (a958bcb) into devel (b145ca3) will not change coverage.
The diff coverage is n/a.

Impacted file tree graph

@@           Coverage Diff           @@
##            devel     #800   +/-   ##
=======================================
  Coverage   73.96%   73.96%           
=======================================
  Files          84       84           
  Lines        6588     6588           
=======================================
  Hits         4873     4873           
  Misses       1715     1715           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b145ca3...a958bcb. Read the comment docs.

@iProzd iProzd closed this Jun 26, 2021
@iProzd iProzd reopened this Jun 26, 2021
@njzjz
Copy link
Member

njzjz commented Jun 26, 2021

Is our gelu much faster than TensorFlow's (tensorflow/tensorflow#41178)? It seems that they implement gelu only using some python codes. See https://github.com/tensorflow/tensorflow/blob/v2.5.0/tensorflow/python/ops/nn_ops.py#L3505-L3548

@amcadmus
Copy link
Member

Is our gelu much faster than TensorFlow's (tensorflow/tensorflow#41178)? It seems that they implement gelu only using some python codes. See https://github.com/tensorflow/tensorflow/blob/v2.5.0/tensorflow/python/ops/nn_ops.py#L3505-L3548

Yes, that's why we implement gelu by ourselves

@amcadmus amcadmus merged commit 4db15c8 into deepmodeling:devel Jun 27, 2021
gzq942560379 pushed a commit to HPC-AI-Team/deepmd-kit that referenced this pull request Sep 2, 2021
* add gpu op unittest in source/tests

* fix bug in deepmodeling#713

* Fix bug of empty input in gelu.cu
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.

4 participants