[JIT] clear GraphFunction.optimized_graphs_ after freezing a module #68316
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
Consider the following:
```
class Mod(nn.Module):
def __init__(self, val):
super().__init__()
self.param = nn.Parameter(val)
def forward(self, x):
# this method will change during freezing
return x + self.param
@torch.jit.export
def make_prediction(self, x):
y = x + x
return self.forward(y)
param = torch.rand([2, 2])
unscripted_mod = Mod(param)
mod = torch.jit.script(unscripted_mod)
mod.eval()
mod = torch.jit.freeze(mod, preserved_attrs=["make_prediction"])`
```
During freezing the following will occur:
1. do some pre-freezing, including inlining; in particular, forward will be inlined into make_prediction. During inlining, forward.optimized_graph() is called, and the result is cached
2. freeze some methods. While freezing forward, the graph associated with the function will get updated. The cached optimized_graphs_ are not updated.
Previously, a call to `mod.forward(x)` would return an exectutor that would run on the old cached optimized_graph(). This would mean that the freezing optimizations would not apply, and potentially that the execution would fail because of parameters removed from the module.
This change clears the optimized_graphs_ cache after running freezing to prevent executing an old version of the graph.
[ghstack-poisoned]
CI Flow Status⚛️ CI FlowRuleset - Version:
You can add a comment to the PR and tag @pytorchbot with the following commands: # ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun
# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slowFor more information, please take a look at the CI Flow Wiki. |
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 76f04cc (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
davidberard98
added a commit
that referenced
this pull request
Nov 13, 2021
Consider the following:
```
class Mod(nn.Module):
def __init__(self, val):
super().__init__()
self.param = nn.Parameter(val)
def forward(self, x):
# this method will change during freezing
return x + self.param
torch.jit.export
def make_prediction(self, x):
y = x + x
return self.forward(y)
param = torch.rand([2, 2])
unscripted_mod = Mod(param)
mod = torch.jit.script(unscripted_mod)
mod.eval()
mod = torch.jit.freeze(mod, preserved_attrs=["make_prediction"])`
```
During freezing the following will occur:
1. do some pre-freezing, including inlining; in particular, forward will be inlined into make_prediction. During inlining, forward.optimized_graph() is called, and the result is cached
2. freeze some methods. While freezing forward, the graph associated with the function will get updated. The cached optimized_graphs_ are not updated.
Previously, a call to `mod.forward(x)` would return an exectutor that would run on the old cached optimized_graph(). This would mean that the freezing optimizations would not apply, and potentially that the execution would fail because of parameters removed from the module.
This change clears the optimized_graphs_ cache after running freezing to prevent executing an old version of the graph.
ghstack-source-id: f5d821c
Pull Request resolved: #68316
facebook-github-bot
added
the
oncall: jit
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
eellison
approved these changes
Nov 15, 2021
| return true; | ||
| } | ||
|
|
||
| void clear_optimized_graphs() { |
There was a problem hiding this comment.
is this the right way of clearing out an array ? im not that well-versed on this c++ idiom cc @jjsjann123
There was a problem hiding this comment.
I suppose std::fill might be better?
There was a problem hiding this comment.
std::array::fill. So I suppose something like
optimized_graphs_.fill(c10::nullopt);
… a module"
Consider the following:
```
class Mod(nn.Module):
def __init__(self, val):
super().__init__()
self.param = nn.Parameter(val)
def forward(self, x):
# this method will change during freezing
return x + self.param
torch.jit.export
def make_prediction(self, x):
y = x + x
return self.forward(y)
param = torch.rand([2, 2])
unscripted_mod = Mod(param)
mod = torch.jit.script(unscripted_mod)
mod.eval()
mod = torch.jit.freeze(mod, preserved_attrs=["make_prediction"])`
```
During freezing the following will occur:
1. do some pre-freezing, including inlining; in particular, forward will be inlined into make_prediction. During inlining, forward.optimized_graph() is called, and the result is cached
2. freeze some methods. While freezing forward, the graph associated with the function will get updated. The cached optimized_graphs_ are not updated.
Previously, a call to `mod.forward(x)` would return an exectutor that would run on the old cached optimized_graph(). This would mean that the freezing optimizations would not apply, and potentially that the execution would fail because of parameters removed from the module.
This change clears the optimized_graphs_ cache after running freezing to prevent executing an old version of the graph.
Differential Revision: [D32410862](https://our.internmc.facebook.com/intern/diff/D32410862)
[ghstack-poisoned]
davidberard98
added a commit
that referenced
this pull request
Nov 15, 2021
Consider the following:
```
class Mod(nn.Module):
def __init__(self, val):
super().__init__()
self.param = nn.Parameter(val)
def forward(self, x):
# this method will change during freezing
return x + self.param
torch.jit.export
def make_prediction(self, x):
y = x + x
return self.forward(y)
param = torch.rand([2, 2])
unscripted_mod = Mod(param)
mod = torch.jit.script(unscripted_mod)
mod.eval()
mod = torch.jit.freeze(mod, preserved_attrs=["make_prediction"])`
```
During freezing the following will occur:
1. do some pre-freezing, including inlining; in particular, forward will be inlined into make_prediction. During inlining, forward.optimized_graph() is called, and the result is cached
2. freeze some methods. While freezing forward, the graph associated with the function will get updated. The cached optimized_graphs_ are not updated.
Previously, a call to `mod.forward(x)` would return an exectutor that would run on the old cached optimized_graph(). This would mean that the freezing optimizations would not apply, and potentially that the execution would fail because of parameters removed from the module.
This change clears the optimized_graphs_ cache after running freezing to prevent executing an old version of the graph.
ghstack-source-id: 283191a
Pull Request resolved: #68316
|
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
jjsjann123
approved these changes
Nov 16, 2021
|
@davidberard98 merged this pull request in 5cfca55. |
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:
Consider the following:
During freezing the following will occur:
Previously, a call to
mod.forward(x)would return an exectutor that would run on the old cached optimized_graph(). This would mean that the freezing optimizations would not apply, and potentially that the execution would fail because of parameters removed from the module.This change clears the optimized_graphs_ cache after running freezing to prevent executing an old version of the graph.
Differential Revision: D32410862