[Profiler] Use parameter as key for optimizer state recording. #86753
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/86753
Note: Links to docs will display an error until the docs builds have been completed. ✅ No Failures, 8 PendingAs of commit 017d779: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This was referenced Oct 12, 2022
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
This was referenced Oct 12, 2022
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
slgong-fb
approved these changes
Oct 17, 2022
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
aaronenyeshi
approved these changes
Oct 20, 2022
Member
aaronenyeshi
left a comment
aaronenyeshi
left a comment
There was a problem hiding this comment.
This looks great! The code looks more extensible and clean!
added 2 commits
October 21, 2022 11:42
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
…ing." While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) [ghstack-poisoned]
This was referenced Oct 23, 2022
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Nov 5, 2022
…ch#86753) While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) Pull Request resolved: pytorch#86753 Approved by: https://github.com/slgong-fb, https://github.com/aaronenyeshi
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Dec 10, 2022
…ch#86753) While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all `torch.optim` optimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises. One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the `check_and_store` logic could be simplified and inlined by this joining of collected optimizer state. Differential Revision: [D40210703](https://our.internmc.facebook.com/intern/diff/D40210703/) Pull Request resolved: pytorch#86753 Approved by: https://github.com/slgong-fb, https://github.com/aaronenyeshi
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
AccumulateGradname #86909While optimizer can store state however it likes, in practice most optimizer state corresponds to a particular parameter. (This is the case for all
torch.optimoptimizers.) Thus, it turns out to be ergonomic to collect using that structure. Note that this doesn't lock us into anything; we can always collect state with non Tensor keys if the use case arises.One simplification that arises is that Module and Optimizer collection has very similar structure. So similar, in fact, that it is possible to use a common template for config. I also found that a lot of the
check_and_storelogic could be simplified and inlined by this joining of collected optimizer state.Differential Revision: D40210703