-
Notifications
You must be signed in to change notification settings - Fork 24.3k
fix dynamo nn module stack fqn #142823
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 dynamo nn module stack fqn #142823
Conversation
Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Here we address them explicitly. Fixes #141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/142823
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 6bb19ec with merge base 2b105de ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This pull request was exported from Phabricator. Differential Revision: D67064189 |
Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Here we address them explicitly. Fixes #141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) ghstack-source-id: 257572754 Pull Request resolved: #142823
Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Added a test that used to have `._modules` inside nn_module_stack fqns, now doesn't. (Unfortunately couldn't repro a case mentioned in the GH issue where `.slice(...)` is claimed to appear as well.) Fixes #141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) [ghstack-poisoned]
This pull request was exported from Phabricator. Differential Revision: D67064189 |
Pull Request resolved: #142823 Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Here we address them explicitly. Fixes #141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) ghstack-source-id: 257599607
Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Added a test that used to have `._modules` inside nn_module_stack fqns, now doesn't. (Unfortunately couldn't repro a case mentioned in the GH issue where `.slice(...)` is claimed to appear as well.) Fixes #141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) [ghstack-poisoned]
This pull request was exported from Phabricator. Differential Revision: D67064189 |
Pull Request resolved: #142823 Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Here we address them explicitly. Fixes #141939 ghstack-source-id: 257610691 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is _modules the only case we need to handle?
As I said in the summary, there's another case that was reported (slice call) but I can't seem to repro it. |
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Dynamo can produce sources that have funny patterns in their `.name()` that break `nn_module_stack` fqns. Added a test that used to have `._modules` inside nn_module_stack fqns, now doesn't. (Unfortunately couldn't repro a case mentioned in the GH issue where `.slice(...)` is claimed to appear as well.) Fixes pytorch#141939 Differential Revision: [D67064189](https://our.internmc.facebook.com/intern/diff/D67064189/) Pull Request resolved: pytorch#142823 Approved by: https://github.com/pianpwk, https://github.com/zhxchen17
Stack from ghstack (oldest at bottom):
Dynamo can produce sources that have funny patterns in their
.name()
that breaknn_module_stack
fqns. Added a test that used to have._modules
inside nn_module_stack fqns, now doesn't. (Unfortunately couldn't repro a case mentioned in the GH issue where.slice(...)
is claimed to appear as well.)Fixes #141939
Differential Revision: D67064189