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.

Sign up for GitHub

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

Jump to bottom

Pytorch models needs two warmup runs #114

Closed
kkontny opened this issue Mar 18, 2022 · 1 comment
Closed

Pytorch models needs two warmup runs #114

kkontny opened this issue Mar 18, 2022 · 1 comment
Assignees
@jan-grzybek-ampere
Labels
enhancement New feature or request

Comments

@kkontny
Copy link
Contributor

kkontny commented Mar 18, 2022

Pytorch on first run of the model gathers profiling data for graph optimization which occurs on second run. Each of this runs can take several seconds which can bias the performance results. First two runs in Pytorch should be dropped from performance calculations.
@jan-grzybek-ampere jan-grzybek-ampere self-assigned this May 11, 2022
@jan-grzybek-ampere jan-grzybek-ampere added the enhancement New feature or request label May 11, 2022
@kkontny kkontny mentioned this issue May 27, 2022
Closed
@kkontny
Copy link
Contributor Author

kkontny commented May 27, 2022

I think this is implemented right now, closing.

@kkontny kkontny closed this as completed May 27, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@jan-grzybek-ampere jan-grzybek-ampere

Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Pytorch should use two warmup runs AmpereComputingAI/ampere_model_library
2 participants
@jan-grzybek-ampere @kkontny