Skip to content

Add peephole optimization for type_as operators. #9316

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.

Sign up for GitHub

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

Jump to bottom
Closed
runtian-zhou wants to merge 5 commits into from
Closed

Add peephole optimization for type_as operators. #9316

runtian-zhou wants to merge 5 commits into from

Conversation

@runtian-zhou
Copy link

@runtian-zhou runtian-zhou commented Jul 10, 2018

If the type_as operator takes in two values with the same type, remove that operator.

zou3519
zou3519 reviewed Jul 10, 2018
View reviewed changes
torch/csrc/jit/passes/peephole.cpp Outdated
}
} break;
case aten::type_as: {
if (n->inputs().size() != 2) {

This comment was marked as off-topic.

Sign in to view

zou3519
zou3519 reviewed Jul 10, 2018
View reviewed changes
test/expect/TestJit.test_peephole-different_size.expect Outdated
@@ -0,0 +1,5 @@
graph(%0 : Double(1)
%1 : Double(2)) {
%2 : Double(1) = aten::type_as(%0, %1)

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

zdevito
zdevito reviewed Jul 10, 2018
View reviewed changes
Copy link
Contributor

@zdevito zdevito left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The tests look good. We should make sure type_as is removed even if sizes are different.

torch/csrc/jit/passes/peephole.cpp Outdated
Value* LHS(n->input(0));
Value* RHS(n->input(1));
// If LHS and RHS have the same static type, remove the type_as operator.
if ((RHS->type()->kind() != TypeKind::DynamicType &&

This comment was marked as off-topic.

Sign in to view

torch/csrc/jit/passes/peephole.cpp Outdated
} break;
case aten::type_as: {
JIT_ASSERT(n->inputs().size() == 2);
Value* LHS(n->input(0));

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

torch/csrc/jit/passes/peephole.cpp Outdated
Value* LHS(n->input(0));
Value* RHS(n->input(1));
// If LHS and RHS have the same static type, remove the type_as operator.
if ((RHS->type()->kind() != TypeKind::DynamicType &&

This comment was marked as off-topic.

Sign in to view

@ssnl
Copy link
Collaborator

ssnl commented Jul 10, 2018

@pytorchbot retest this please

zdevito
zdevito reviewed Jul 11, 2018
View reviewed changes
torch/csrc/jit/passes/peephole.cpp Outdated
Value* RHS(n->input(1));
// If LHS and RHS have the same static type, remove the type_as operator.
if (RHS->type()->kind() == TypeKind::TensorType) {
auto LType = (*LHS->type()).cast<TensorType>();

This comment was marked as off-topic.

Sign in to view

apaszke
apaszke reviewed Jul 11, 2018
View reviewed changes
torch/csrc/jit/passes/peephole.cpp Outdated
} break;
case aten::type_as: {
JIT_ASSERT(n->inputs().size() == 2);
Value* LHS(n->input(0));

This comment was marked as off-topic.

Sign in to view

test/test_jit.py Outdated
self.assertExpectedGraph(trace, subname="same_size")
trace, z = torch.jit.get_trace_graph(f, (a, c))
self.run_pass('peephole', trace)
self.assertExpectedGraph(trace, subname="different_size")

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

test/test_jit.py Outdated

fn = torch.jit.script(f)
torch._C._jit_pass_peephole(fn.graph)
self.assertExpectedGraph(fn.graph)

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

test/test_jit.py Outdated

trace, z = torch.jit.get_trace_graph(f, (a, c))
self.run_pass('peephole', trace)
self.assertExpectedGraph(trace, subname="different_device")

This comment was marked as off-topic.

Sign in to view

test/test_jit.py Outdated
self.assertExpectedGraph(trace, subname="different_size")
trace, z = torch.jit.get_trace_graph(f, (a, d))
self.run_pass('peephole', trace)
self.assertExpectedGraph(trace, subname="different_type")

This comment was marked as off-topic.

Sign in to view

facebook-github-bot
facebook-github-bot reviewed Jul 11, 2018
View reviewed changes
Copy link
Contributor

@facebook-github-bot facebook-github-bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ezyang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

zdevito
zdevito approved these changes Jul 13, 2018
View reviewed changes
Copy link
Contributor

@zdevito zdevito left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good now!

facebook-github-bot
facebook-github-bot reviewed Jul 13, 2018
View reviewed changes
Copy link
Contributor

@facebook-github-bot facebook-github-bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@zdevito has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@ssnl
Copy link
Collaborator

ssnl commented Jul 13, 2018

@pytorchbot retest this please

petrex pushed a commit to petrex/pytorch that referenced this pull request Jul 16, 2018
Merge remote-tracking branch 'upstream/master' into hipify_c2
986eeb7 
* upstream/master: (24 commits)
  Implement tensor weak references (pytorch#9363)
  Nuke TestCollectEnv (pytorch#9459)
  Add test case for segmentation fault fix in grad_fn (pytorch#9457)
  Add peephole optimization for type_as operators. (pytorch#9316)
  Fix out-of-range error for test_neg (pytorch#9431)
  add depthwise conv support for mkldnn (pytorch#8782)
  Refactor `_log_sum_exp` (pytorch#9173)
  Add ModuleDict and ParameterDict containers (pytorch#8463)
  Introduce SupervisedPtr, delete THAllocator and THCDeviceAllocator (pytorch#9358)
  Introducing IsInf (pytorch#9169)
  add device to CUDAEvent (pytorch#9415)
  Make localScalar error message more intuitive (pytorch#9443)
  Only accept continguous tensors in TopK for cuda (pytorch#9441)
  Add support for .norm() pytorch onnx export and ReduceL1/ReduceL2 caffe2 operators (pytorch#9299)
  Only view() rhs of index_put if we need to (pytorch#9424)
  Add BatchBucketizeOp in caffe2 (pytorch#9385)
  Implementation of Wngrad optimizer caffe2 python wrapper and unit test on least square regression (pytorch#9001)
  Implementation and operator test for Wngrad optimizer (pytorch#8999)
  Fix segmentation fault in grad_fn (pytorch#9292)
  update docs (pytorch#9423)
  ...
goldsborough pushed a commit to goldsborough/pytorch that referenced this pull request Jul 20, 2018
@goldsborough
Add peephole optimization for type_as operators. (pytorch#9316)
bfd82d0 
Summary:
If the type_as operator takes in two values with the same type, remove that operator.
Pull Request resolved: pytorch#9316

Reviewed By: zdevito

Differential Revision: D8808355

fbshipit-source-id: 2d5710a6380b22f4568fc38a439061b5340c4eb1
jramseyer pushed a commit to jramseyer/pytorch that referenced this pull request Jul 30, 2018
@jramseyer
Add peephole optimization for type_as operators. (pytorch#9316)
15ad321 
Summary:
If the type_as operator takes in two values with the same type, remove that operator.
Pull Request resolved: pytorch#9316

Reviewed By: zdevito

Differential Revision: D8808355

fbshipit-source-id: 2d5710a6380b22f4568fc38a439061b5340c4eb1
goodlux pushed a commit to goodlux/pytorch that referenced this pull request Aug 15, 2018
Add peephole optimization for type_as operators. (pytorch#9316)
61276ea 
Summary:
If the type_as operator takes in two values with the same type, remove that operator.
Pull Request resolved: pytorch#9316

Reviewed By: zdevito

Differential Revision: D8808355

fbshipit-source-id: 2d5710a6380b22f4568fc38a439061b5340c4eb1
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@zou3519 zou3519 zou3519 left review comments

@facebook-github-bot facebook-github-bot facebook-github-bot left review comments

@zdevito zdevito zdevito approved these changes

@colesbury colesbury Awaiting requested review from colesbury

@ezyang ezyang Awaiting requested review from ezyang

@gchanan gchanan Awaiting requested review from gchanan

@soumith soumith Awaiting requested review from soumith

+1 more reviewer

@apaszke apaszke apaszke left review comments

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

open source

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

7 participants

@runtian-zhou @ssnl @zdevito @apaszke @zou3519 @facebook-github-bot @ezyang