-
Notifications
You must be signed in to change notification settings - Fork 526
Fix static attention mask update #9101
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/9101
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit dd04ea9 with merge base d71f54a ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This pull request was exported from Phabricator. Differential Revision: D70914174 |
This pull request was exported from Phabricator. Differential Revision: D70914174 |
Summary: The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Differential Revision: D70914174
Summary: The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Differential Revision: D70914174
This pull request was exported from Phabricator. Differential Revision: D70914174 |
1 similar comment
This pull request was exported from Phabricator. Differential Revision: D70914174 |
Summary: Pull Request resolved: pytorch#9101 The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Differential Revision: D70914174
Summary: The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Reviewed By: billmguo Differential Revision: D70914174
Summary: The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Reviewed By: billmguo Differential Revision: D70914174
This pull request was exported from Phabricator. Differential Revision: D70914174 |
Summary: Pull Request resolved: pytorch#9101 The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor. Reviewed By: billmguo Differential Revision: D70914174
This pull request was exported from Phabricator. Differential Revision: D70914174 |
Summary: The range based for loop was making a copy of the mask, and thus the updates did not take effect. Remove the copy and move constructors of StaticKVCache and StaticAttention as they are not needed. Also add the missing deallocate call in mask's destructor.
Differential Revision: D70914174