-
Notifications
You must be signed in to change notification settings - Fork 24.4k
[BE] Import CUDATemplateCaller non-lazily in select_algorithm.py #153765
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
Conversation
Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/153765
Note: Links to docs will display an error until the docs builds have been completed. ❌ 4 New FailuresAs of commit ed685eb with merge base c45515c ( NEW FAILURES - The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) ghstack-source-id: 284551003 Pull Request resolved: #153765
This pull request was exported from Phabricator. Differential Revision: D74911280 |
…thm.py" Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov [ghstack-poisoned]
Pull Request resolved: #153765 Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) ghstack-source-id: 284881312
This pull request was exported from Phabricator. Differential Revision: D74911280 |
…thm.py" Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov [ghstack-poisoned]
Pull Request resolved: #153765 ghstack-source-id: 284892017 Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/)
This pull request was exported from Phabricator. Differential Revision: D74911280 |
…thm.py" Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/) cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov [ghstack-poisoned]
Pull Request resolved: #153765 ghstack-source-id: 284959940 Differential Revision: [D74911280](https://our.internmc.facebook.com/intern/diff/D74911280/)
This pull request was exported from Phabricator. Differential Revision: D74911280 |
Just to be sure, this doesn't do any poisoning of the CUDA context when torch is imported, right? |
Actually, double check CUDA context poisoning
good question. What do you mean by poisoning the cuda context? |
Accidentally initializing CUDA when it's not necessary (and loading memory onto the GPU etc). IE. Imagine a dataloader process that only create CPU tensor wasitng GPU memory for every worker. |
I think it shouldn't, but unless I know what to look for, it is hard to test. I am fine with either of the following options:
also @ColinPeppler who proposed this change. |
Stack from ghstack (oldest at bottom):
Differential Revision: D74911280
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @chenyang78 @kadeng @muchulee8 @amjames @chauhang @aakhundov