Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/73253
This PR allows TS schema_matching to match input arg with self for aten operators. This is because, operators in their functional form have input as paremeter instead of self.
fixes: https://github.com/pytorch/pytorch/issues/71994
Test Plan: Imported from OSS
Reviewed By: gmagogsfm
Differential Revision: D34427556
Pulled By: tugsbayasgalan
fbshipit-source-id: 96c2340d605c59634bf6e37db1db6025d93a933a
(cherry picked from commit 45a593d73bc5e6308dd80a4a29afed8e318a0a1c)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/72596
debug_pkl file inside of pytorch's .pt file consists of a list of SourceRanges. Each SourceRange points to a Source which is a stack track, filename, and start, end numbers. Those are emitted in debug_pkl file as strings.
Since many SourceRange shares the same source, the string for trace can be deduped.
The newer format saves a set of unique traces in a tuple, then each SourceRange will save the offset of it's trace w.r.t. position in that tuple. (i.e. manually applying dictionary compression).
The above helps with smaller file size. On loading, if we copy each trace to Source as string the runtime memory would still blowup.
To mitigate this, we use SourceView directly instead of source which will take the reference of string inside of Deserializer and make that into string_view. This is safe because Deserializer is hold by Unpickler by shared_ptr, and Unpickler is also hold by shared_ptr by another Source object. That Source object will be alive during the model construction.
Test Plan:
unit test
Took original file (312271638_930.predictor.disagg.local); loaded with `torch.jit.load` save again with `torch.jit.save`. Unzip both, look at contents:
```
[qihan@devvm5585.vll0 ~]$ du archive -h
4.0K archive/xl_model_weights
3.7M archive/extra
8.0K archive/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K archive/code/__torch__/caffe2/torch/fb/model_transform
8.0K archive/code/__torch__/caffe2/torch/fb
8.0K archive/code/__torch__/caffe2/torch
8.0K archive/code/__torch__/caffe2
20M archive/code/__torch__/torch/fx/graph_module
20M archive/code/__torch__/torch/fx
8.0K archive/code/__torch__/torch/classes
20M archive/code/__torch__/torch
20M archive/code/__torch__
20M archive/code
2.7M archive/constants
35M archive
[qihan@devvm5585.vll0 ~]$ du resaved -h
4.0K resaved/extra
8.0K resaved/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K resaved/code/__torch__/caffe2/torch/fb/model_transform
8.0K resaved/code/__torch__/caffe2/torch/fb
8.0K resaved/code/__torch__/caffe2/torch
8.0K resaved/code/__torch__/caffe2
1.3M resaved/code/__torch__/torch/fx/graph_module
1.3M resaved/code/__torch__/torch/fx
8.0K resaved/code/__torch__/torch/classes
1.4M resaved/code/__torch__/torch
1.4M resaved/code/__torch__
1.4M resaved/code
2.7M resaved/constants
13M resaved
[qihan@devvm5585.vll0 ~]$
```
Reviewed By: JasonHanwen
Differential Revision: D33994011
fbshipit-source-id: 8e6224c6e942e91c3403f686c8f0937d1002ed41
(cherry picked from commit a7014dd4029308c95007f362a57c31796d686647)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/72899
Reland D33282878 (911d527b87). This is the frontend change.
ghstack-source-id: 149204031
Test Plan: Refer to D33282878 (911d527b87). Also check CI
Reviewed By: gmagogsfm
Differential Revision: D34252127
fbshipit-source-id: 27b17ddd4d05d904eb91fd9ee094d9121f00e388
(cherry picked from commit 1d276baca3)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/70471
Reland D33282878 (911d527b87). This is the frontend change.
ghstack-source-id: 149114933
Test Plan: Refer to D33282878 (911d527b87). Also check CI
Reviewed By: gmagogsfm
Differential Revision: D33342569
fbshipit-source-id: 57984ac67ae2c56c38f72d3b1fb69105901fb472
(cherry picked from commit b47cc935ee)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/69645
As noted in code comment:
existing device operator is registered with input name `a`, which prevents torch.device(type="cuda") from working. add shim-layer here
Test Plan: Imported from OSS
Reviewed By: jbschlosser
Differential Revision: D33515231
Pulled By: eellison
fbshipit-source-id: c04af8158a9568a20cd5fbbbd573f6efab98fd60
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/65645
This is a retry of PR: https://github.com/pytorch/pytorch/pull/59492
Latest Changes: Added more tests, added the getOrCreateDB pattern, updated logic to remove unnecessary checks
addressed all comments.
Adding code to find common expressions from the two subblocks of an if
operation and hoist them before the if block.
This also allows Dead Code Elimination to
then eliminate some if blocks.
Test Plan: python test_jit.py TestIfHoisting
Reviewed By: eellison
Differential Revision: D33302065
Pulled By: Gamrix
fbshipit-source-id: a5a184a480cf07354359aaca344c6e27b687a3c2
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/70339
When a python program is translated to TorchScript, the python exception type is dropped. This makes users's life hard when they need to categorize errors based more than only exception message.
Here we make the change so when we raise a python exception, we record the fully qualified class name for the exception. Later on when the TorchScript is interpreted, a special exception CustomJITException is thrown. User can get the python class name from CustomJITException::getPythonClassName .
Note that, this diff does not customize the mapping from C++ exception to Python exception. It's left to the users to do whatever mapping they want.
Code under scripts/shunting are just my own experimental code. I can split them out if requested.
ghstack-source-id: 146221879
Test Plan: buck test mode/opt //caffe2/test:jit
Reviewed By: gmagogsfm
Differential Revision: D33282878
fbshipit-source-id: 910f67a764519f1053a48589d1a34df69001525d
Summary:
1. Split the test `test_save_load.py` to two files. Basically move the operator versioning related changes to `test_save_load_for_op_versions.py`.
2. Add mobile module related test to `test_save_load_for_op_versions.py`
How to run:
```
buck test mode/opt //caffe2/test:jit
or
python test/test_jit.py TestSaveLoadForOpVersion
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/70090
ghstack-source-id: 146103547
Test Plan:
```
buck test mode/opt //caffe2/test:jit
python test/test_jit.py TestSaveLoadForOpVersion
```
Reviewed By: tugsbayasgalan
Differential Revision: D33180767
fbshipit-source-id: dd31e313c81e90b598ea9dd5ad04a853c017f994
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/69614
Previously sparse COO tensors were ignored during freezing, because
`tryInsertConstant` would fail during `freeze_module.cpp`, and because
hashes weren't implemented for COO tensor IValues.
Test Plan: Imported from OSS
Reviewed By: mrshenli
Differential Revision: D32954620
Pulled By: davidberard98
fbshipit-source-id: a91f97fdfc2152b417f43a6948100c94970c0831
Summary:
As per title. This in particular allows to more easily override backward function for which the underlying backend returns `None`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/67793
Reviewed By: zou3519
Differential Revision: D32242962
Pulled By: albanD
fbshipit-source-id: 6e114def90ee9499161e1303d301ba7fd003ff89
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/68672
This PR adds `python_module: sparse` to `native_function.yaml`.
These functions would appear in `torch._C._sparse` namespace instead of
just `torch`.
Test Plan: Imported from OSS
Reviewed By: mruberry
Differential Revision: D32517813
fbshipit-source-id: 7c3d6df57a24d7c7354d0fefe1b628dc89be9431
Summary:
Adds native_dropout to have a reasonable target for torchscript in auto diff. native_dropout has scale and train as arguments in its signature, this makes native_dropout more consistent with other operators and removes conditionals in the autodiff definition.
cc gmagogsfm
Pull Request resolved: https://github.com/pytorch/pytorch/pull/63937
Reviewed By: mruberry
Differential Revision: D32477657
Pulled By: ngimel
fbshipit-source-id: d37b137a37acafa50990f60c77f5cea2818454e4
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/68367
- bmm_test.py was using syntax not allowed in 3.6
- Some suppressions were not placed on the correct line.
With this file,
```
lintrunner --paths-cmd='git grep -Il .'
```
passes successfully.
Test Plan: Imported from OSS
Reviewed By: janeyx99, mrshenli
Differential Revision: D32436644
Pulled By: suo
fbshipit-source-id: ae9300c6593d8564fb326822de157d00f4aaa3c2
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/68311
prim::SetAttr is listed as an op with side effects, but in AliasDb, `analyzeSetAttr` already accounts for its behavior. By removing it from the list of ops with side effects, dead code elimination will work in a few other scenarios.
Test Plan: Imported from OSS
Reviewed By: mrshenli
Differential Revision: D32409510
fbshipit-source-id: 52ed9e19f92afb95c669ad3c2440f72f9515ba4c
Summary:
CAFFE2 has been deprecated for a while, but still included in every PyTorch build.
We should stop building it by default, although CI should still validate that caffe2 code is buildable.
Build even fewer dependencies when compiling mobile builds without Caffe2
Introduce `TEST_CAFFE2` in torch.common.utils
Skip `TestQuantizedEmbeddingOps` and `TestJit.test_old_models_bc` is code is compiled without Caffe2
Should be landed after https://github.com/pytorch/builder/pull/864
Pull Request resolved: https://github.com/pytorch/pytorch/pull/66658
Reviewed By: driazati, seemethere, janeyx99
Differential Revision: D31669156
Pulled By: malfet
fbshipit-source-id: 1cc45e2d402daf913a4685eb9f841cc3863e458d
Summary:
All of the pooling modules except MaxUnpool and LPPool return either a
Tensor or [Tensor, Tensor]. The current type annotations are inaccurate,
and prevent scripting the module if return_indices is set as True in the
module.
There's not a great way to make this agree with mypy because the
overload is dependent on the value of return_indices, an attribute.
I tried changing the annotations from `Tensor` to
`Union[Tensor, Tuple[Tensor, Tensor]]`, but that breaks a bunch of uses
that have return_indices=False.
For example, this breaks:
4e94e84f65/torch/nn/modules/container.py (L139)
Also clean up how test names were being constructed in test_jit, since
otherwise we were getting name collisions when there were two tests on
the same nn.Module.
Fixes https://github.com/pytorch/pytorch/issues/45904
Pull Request resolved: https://github.com/pytorch/pytorch/pull/65847
Reviewed By: ZolotukhinM
Differential Revision: D31462517
Pulled By: eellison
fbshipit-source-id: 6f9e8df1be6c75e5e1e9bae07cf3ad3603ba59bd
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/65097
Previously, BatchMM would skip any block containing any mutable
operators. Now it will avoid batching any operation whose inputs or
outputs are ever mutated. Specifically: consider a tree of ADD, T,
and MM nodes rooted at an ADD node. If any input or output to any
node in the tree is ever mutated, then the entire tree will be ignored
by BatchMM.
Test Plan: python test/test_jit.py TestBatchMM
Reviewed By: eellison
Differential Revision: D30973515
Pulled By: davidberard98
fbshipit-source-id: 9d836faa1ef0c9e3fefe0ffc0bd265f275471f48
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/63552
In this PR, we want to exclude these 2 cases in the `Autocast` weight cache usages:
- Using `torch.jit.trace` under the `Autocast`
As report in https://github.com/pytorch/pytorch/issues/50231 and several other discussions, using `torch.jit.trace` under the `Autocast`, the trace process would hit Autocast's weight cache and fails. So we should disable weight cache under the trace process.
- Using `Autocast` with `Grad mode`
- Usually we are using `Grad mode` for training. Since in the training phase, the weight will change in every step. So we doesn't need to cache the weight.
- For the recommended `Autocast` training case in the [doc](https://pytorch.org/docs/stable/amp.html), `Autocast` will clear the cache every step leaving the context. We should disable it to save the clear operations.
```
model = Net().cuda()
optimizer = optim.SGD(model.parameters(), ...)
for input, target in data:
optimizer.zero_grad()
with autocast():
output = model(input)
loss = loss_fn(output, target)
loss.backward()
optimizer.step()
```
Test Plan: Imported from OSS
Reviewed By: mrshenli
Differential Revision: D30644913
Pulled By: ezyang
fbshipit-source-id: ad7bc87372e554e7aa1aa0795e9676871b3974e7
Summary:
This PR is created to replace https://github.com/pytorch/pytorch/pull/53180 PR stack, which has all the review discussions. Reason for needing a replacement is due to a messy Sandcastle issue.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/64234
Reviewed By: gmagogsfm
Differential Revision: D30656444
Pulled By: ansley
fbshipit-source-id: 77536c8bcc88162e2c72636026ca3c16891d669a
Summary:
RecursiveScriptModule has its customized `__copy__` and `__deepcopy__` defined. The warning/error that says it is not copiable is outdated
Pull Request resolved: https://github.com/pytorch/pytorch/pull/64085
Reviewed By: rohan-varma
Differential Revision: D30598623
Pulled By: gmagogsfm
fbshipit-source-id: 0701d8617f42d818bc7b88244caee4cd47fbe976
Summary:
Turns on BN in autodiff:
1. outputs an empty tensor for running stats to by pass autodiff issue on None;
2. fixing BN inference backward in cudnn & miopen, where backward falls back to native batchnorm kernel instead;
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57321
Reviewed By: albanD, ngimel
Differential Revision: D30250419
Pulled By: jansel
fbshipit-source-id: a62553789c20fb50a820003a056f40d9d642dfaa
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59492
Adding code to find common expressions from the two subblocks of an if
operation and hoist them before the if block.
This also allows Dead Code Elimination to
then eliminate some if blocks.
Also eliminated some dead code in the codebase.
Test Plan:
python test_jit.py TestIfHoisting
Imported from OSS
Reviewed By: ngimel
Differential Revision: D29399533
fbshipit-source-id: 9336b9dc48c02c38862f98f98cd72fc1767a1802
Summary:
Fixes https://github.com/pytorch/pytorch/issues/54420
When I tested on master, with the testing code, there were multiple objects on the garbage collector that cannot be printed.
Testing code:
```
import torch
import gc
import os
import sys
print(torch.__version__)
a = torch.rand(10)
print(a)
objects = gc.get_objects()
for i in range(len(objects)):
print(objects[i])
```
### 1
```
print(torch.classes)
```
Like SplitInfinity has mentioned in the GitHub issue, the solution here is to set `__file__` for `torch.classes` to something. Similar to [_ops.py](https://github.com/pytorch/pytorch/blob/master/torch/_ops.py#L69), where `__file__` is set to `_ops.py`, we could set `__file__` for torch.classes to `_classes.py`.
### 2
```
print(torch._ops.ops.quantized)
print(torch._ops.ops.atan)
```
When we try to print these two modules, it will call `_OpNamespace::__getattr__`, but the `op_name` is `__file__`. This becomes a problem when `torch._C._jit_get_operation(qualified_op_name)` [(link)](https://github.com/pytorch/pytorch/blob/master/torch/_ops.py#L60) tries to look for an actual op on the native C++ side.
Only when we get the attribute for an actual op, e.g. `print(torch._ops.ops.quantized.elu)`, the `op_name` becomes proper (e.g. `elu`).
My current solution is to return a hardcoded string (i.e. “torch.ops”) if `op_name` is `"__file__"`.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/62447
Reviewed By: saketh-are
Differential Revision: D30234654
Pulled By: yidawang-oss
fbshipit-source-id: de43a8f599739c749fb3307eea015cc61f1da60e
Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/62842
Test Plan:
Wrote unit test TestAtenPow to test behavior of aten::pow when:
1. base is int, exponent is int
2. base is int, exponent is float
3. base is float, exponent is int
4. base is float, exponent is float
Specifically, we test that when base is zero and exponent is negative, we raise error. In all other cases, we expect behavior to be the same as the result returned by Python.
It is because the cpp code relies on overloading, we need to make sure all combinations of types give us the expected result.
Reviewed By: zhxchen17
Differential Revision: D30146115
Pulled By: szewaiyuen7
fbshipit-source-id: dc661897ad38da286ee454120fbe41314b7f2995
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59956
Issue #50175. Basically two things need to be checked and are lacking currently:
1. Overload declarations should always have a single `pass` statement as the body.
2. There should be always an implementation provided for decls which doesn't
have the torch.jit._overload decorator. So in this case we need to check
whether we are actually compiling a function body with decorator ahead.
Test Plan:
python test/test_jit.py TestScript.test_function_overloads
Imported from OSS
Reviewed By: gmagogsfm
Differential Revision: D29106555
fbshipit-source-id: 2d9d7df2fb51ab6db0e1b726f9644e4cfbf733d6
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/62703
Re-enable test on Windows
Test Plan: CI
Reviewed By: ezyang
Differential Revision: D30094460
Pulled By: ilia-cher
fbshipit-source-id: 80521f6bc1365d2c252f20b5d0485fc062c8d9c3
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/61931
This PR consolidates the profiling code around a new C++ implementation
(profiler_kineto.h/cpp) and uses it unconditionally from
torch.autograd.profiler/torch.profiler:
1. Always use profiler_kineto.h/cpp as the C++ implementation
2. Simplify profiler.py to remove unneeded parts depending on legacy
impl
3. Move some of the legacy logic into profiler_legacy.py (to be fully
deleted later)
Test Plan:
USE_KINETO=1 USE_CUDA=1 USE_MKLDNN=1 BLAS=MKL BUILD_BINARY=1 python setup.py develop install --cmake
python test/test_profiler.py -v
USE_KINETO=0 USE_CUDA=1 USE_MKLDNN=1 BLAS=MKL BUILD_BINARY=1 python setup.py develop install --cmake
python test/test_profiler.py -v
Imported from OSS
Reviewed By: gdankel
Differential Revision: D29801599
fbshipit-source-id: 9794d29f2af38dddbcd90dbce4481fc8575fa29e
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/61878
CMakeLists.txt
Android NNAPI delegate library was moved from test/cpp/jit/CMakeLists.txt to torch/CMakeLists.txt. This resolves the issue the original PR had, where the NNAPI delegate library was added to builds without Python (when it depends on Python).
Original PR: https://github.com/pytorch/pytorch/pull/61594
There's an error where the library cannot be built on MacOS. This problem existed in the original PR as well, but now an issue has been created: https://github.com/pytorch/pytorch/issues/61930
test_backend_nnapi.py
Also changed the skip unit test headers so that it's a little cleaner. Now the unit tests are skipped if the Nnapi delegate library file is not found. Previously, the skip was based on the platform (only allowing Linux).
Test Plan:
To run NNAPI delegate unit tests: `python test/test_jit.py TestNnapiBackend`
Imported from OSS
Reviewed By: iseeyuan
Differential Revision: D29799895
fbshipit-source-id: b69a767b5cde3814b0853cfbc84d61ab4155f619
Summary:
Issue: https://github.com/pytorch/pytorch/issues/56909
Note the emitted code for such a call will either be a) getattr() call with first two args if the
attribute name (which must be a string literal) is determined to be valid based on the hasAttr() result,
or b) just the AST node for the default value (the 3rd arg) alone with no getattr call at all.
Test code:
```
import torch
import numpy as np
class Shape:
def __init__(self):
self.center = 1.0
def f(x):
s = Shape()
return getattr(s, "missing", [])
y = torch.jit.script(f)
print(y.graph)
```
Output:
```
graph(%x : Tensor):
%s.1 : __torch__.Shape = prim::CreateObject()
%2 : NoneType = prim::CallMethod[name="__init__"](%s.1) # ts.py:10:8
%4 : Tensor[] = prim::ListConstruct()
return (%4)
```
Another example:
```
import torch
class Shape:
def __init__(self):
self.center = 1.0
def f(x):
s = Shape()
y = getattr(s, "center")
w : list[float] = [1.0]
z = getattr(s, "missing", w)
z.append(y)
return z
y = torch.jit.script(f)
print(y.graph)
--- output ---
graph(%x : Tensor):
%5 : float = prim::Constant[value=1.]() # ts.py:12:23
%s.1 : __torch__.Shape = prim::CreateObject()
%2 : NoneType = prim::CallMethod[name="__init__"](%s.1) # ts.py:10:8
%center : float = prim::GetAttr[name="center"](%s.1)
%w.1 : float[] = prim::ListConstruct(%5)
%11 : float[] = aten::append(%w.1, %center) # ts.py:14:4
return (%w.1)
```
Fixes #{56969}
Pull Request resolved: https://github.com/pytorch/pytorch/pull/61599
Reviewed By: ZolotukhinM
Differential Revision: D29776058
Pulled By: jerryzhenleicai
fbshipit-source-id: 76333bd54002e08a064677c1f287115a80cc7c8e
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/61594
### Summary:
Added a unit test for the Nnapi delegate's preprocess() function. The
function was previously tested locally, but now a basic test is
added for OSS.
See https://github.com/pytorch/pytorch/pull/61499 for preprocess
implementation. See D29647123 for local testing.
**TODO:**
Add more comprehensive tests.
Add tests for model execution, after the Nnapi delegate's initialization
and execution is implemented T91991928.
**CMakeLists.txt:**
Added a library for the Nnapi delegate
- Explicit linking of torch_python is necessary for the Nnapi delegate's use of pybind
**test_backends.py:**
Added a test for lowering to Nnapi
- Based off https://github.com/pytorch/pytorch/blob/master/test/test_nnapi.py
- Only differences are the loading of the nnapi backend library and the need to change dtype from float64 to float32
### Test Plan:
Running `python test/test_jit.py TestBackendsWithCompiler -v` succeeds. Also saved and examined the model file locally.
Test Plan: Imported from OSS
Reviewed By: iseeyuan
Differential Revision: D29687143
fbshipit-source-id: 9ba9e57f7f856e5ac15e13527f6178d613b32802
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57137
This PR corrects and expands our typing algorithm for unannotated, non-empty dicts and lists. Previously, to verify type correctness for an unannotated, non-empty container, we had gotten the type of the first element in the container, then checked if each following element was a subtype of the first type. That's too restrictive--what if the first element were a subtype of the second element? Instead, we should type the container by getting the smallest common supertype of all the given elements.
We need slightly different rules for keys and values in dicts, though: because the set of key types is restricted, finding two key types that cannot be unified should cause an error. On the other hand, the set of value types is not restricted, so we should be able to use `Any` as a valid supertype. We need to keep the set of keys restricted since the keys are used to generate and match schemas.
This does not break backwards compatibility, because the default element type is the smallest supertype of all the given types. So, if someone creates an unannotated dict where the keys are all `str` and the values are all `torch.Tensor`, the dict will be inferred to `Dict[str, Tensor]` just like it was before. Empty lists are still typed as `List[torch.Tensor],` and empty dicts are still typed as `Dict[str, Tensor]`.
This PR unblocks three engineers on an FB-internal team and improves FX-TorchScript compatibility.
Test Plan: Imported from OSS
Reviewed By: gmagogsfm
Differential Revision: D28231839
Pulled By: ansley
fbshipit-source-id: 7297bf239749daa54895add708185c75e6ca5999
Summary:
* Minor: spelling, grammar.
* Add calls to `GRAPH_DUMP()` where they were missing.
* Add or expand a few comments.
* Move a few comments to seemingly more appropriate spots.
* In canonicalize_graph_fuser_ops.cpp inline `runnableInputs()` since it
was only called in one place and had a misleading comment and
confusing name.
* In `PeepholeOptimizeImpl::optimizeBlock()`, set `changed = true;` when
removing `aten::is_complex`. Pretty sure its absence was a bug.
* Delete unused `_jit_pass_remove_inplace_ops` and and its
implementation `RemoveInplaceOps()`.
* In `preprocessCaffe2Ops()`, remove redundant check for nested optional
types. It was already checked in `checkONNXCompatibility()`.
* In `EncoderBase::AddAttribute`, log the unexpected attribute kind.
I don't remember the repro case now but I did hit this error at some
point and this additional logging made it easier to understand.
* In `fuseConvBatchNorm()` in eval_peephole.cpp, consistently use
camelCase instead of snake_case for local variables.
* Add curly braces around the bodies of if and loops.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/60390
Reviewed By: Krovatkin
Differential Revision: D29523283
Pulled By: SplitInfinity
fbshipit-source-id: 4e16c5648616f53da07d68dab7fdf252e06a0752
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/52832
**Summary**
This commit adds `torch._C.ScriptList`, a list type that has reference
semantics across the Python/TorchScript boundary. That is, modifications
made in TorchScript to instances of `torch._C.ScriptList`
are visible in Python even when it is not returned from the function.
`torch._C.ScriptList` is implemented using a modified version of pybind's
`stl_bind.h`-style bindings attached to `ScriptList` and `ScriptListIterator`,
wrapper classes around `c10::impl::GenericList` and
`c10::impl::GenericList::iterator`. These bindings allow instances of
`torch._C.ScriptList` to be used as if it were a
regular `list` in Python. Reference semantics are achieved by simply
retrieving the `IValue` contained in `ScriptList` in `toIValue` (invoked
when converting Python arguments to `IValues` before calling TorchScript
code).
**Test Plan**
This commit adds `TestScriptList` to `test_list_dict.py`, a set of tests
that check that all of the common list operations are supported
and that instances have reference semantics across the
Python/TorchScript boundary.
Test Plan: Imported from OSS
Reviewed By: gmagogsfm
Differential Revision: D29478121
Pulled By: SplitInfinity
fbshipit-source-id: 652cc25cfa37debe28db9527504846f22abd8b54
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/60969
This PR fixes the tracing in the parametrizations.
The current resolution is that when tracing is performed while caching is enabled, we throw an error.
Without caching, the tracing should work properly (tests added).
Currently, the parametrizations don't support scripting.
This PR introduces the same logic as with the tracing (throw error if caching).
However, the scripting itself cannot enabled due to the use of the generator expressions in the parametrizations.
Added TODO to fix it.
Test Plan: Imported from OSS
Reviewed By: albanD
Differential Revision: D29462887
Pulled By: z-a-f
fbshipit-source-id: 49721d3059be58f36055d1c374080df41a748d66
Summary:
Added JIT support for the vararg version of `torch.einsum`. Note that JIT does not support the Python's Ellipsis object (`...`)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59265
Reviewed By: VitalyFedyunin
Differential Revision: D29328469
Pulled By: heitorschueroff
fbshipit-source-id: 5e4b177fda93255251f45d735b00c08220f0f124
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/60468
Added a unit test for the execution of a basic module with a compiler
ghstack-source-id: 132307488
Test Plan:
Running python test/test_jit.py TestBackendsWithCompiler -v returns a successful test
Imported from OSS
Reviewed By: iseeyuan
Differential Revision: D29306225
fbshipit-source-id: bf1ff075ebc63acbbe46d6ea030086405e29d7d3
Summary:
The grad() function needs to return the updated values, and hence
needs a non-empty inputs to populate.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/52016
Test Plan:
Passes Python and C++ unit tests, and added new tests to catch this behavior.
Fixes https://github.com/pytorch/pytorch/issues/47061
Reviewed By: albanD
Differential Revision: D26406444
Pulled By: dagitses
fbshipit-source-id: 023aeca9a40cd765c5bad6a1a2f8767a33b75a1a
Summary:
Fixes https://github.com/pytorch/pytorch/issues/27655
This PR adds a C++ and Python version of ReflectionPad3d with structured kernels. The implementation uses lambdas extensively to better share code from the backward and forward pass.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59791
Reviewed By: gchanan
Differential Revision: D29242015
Pulled By: jbschlosser
fbshipit-source-id: 18e692d3b49b74082be09f373fc95fb7891e1b56
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/60003
**Summary**
`infer_concrete_type_builder` in `_recursive.py` assumes `__constants__`
is a `set` if it exists as an attribute on the module being scripted.
Instead, it should create a set out of whatever `__constants__` is.
**Test Plan**
Ran code from the issue.
**Fixes**
This commit fixes#59947.
Test Plan: Imported from OSS
Reviewed By: pbelevich
Differential Revision: D29174243
Pulled By: SplitInfinity
fbshipit-source-id: aeb8bded80038da35478714b6a697a766ac447f5
Summary:
Also fix `TestProducerVersion` by removing assumption that major and minor are single digit
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59345
Reviewed By: robieta
Differential Revision: D28853720
Pulled By: malfet
fbshipit-source-id: 4b6d03c6b0c9d652a5aef792aaa84eaa522d10e8
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/52659
**Summary**
This commit adds `torch._C.ScriptDict`, a dictionary type that has reference
semantics across the Python/TorchScript boundary. That is, modifications
made to instances of `torch._C.ScriptDict` in TorchScript are visible in
Python even when it is not returned from the function. Instances can be
constructed by passing an instance of a Python dictionary to
`torch.jit.script`. In the case of an empty dictionary, its type is
assumed to be `Dict[str, Tensor]` to be consistent with the handling of
empty dictionaries in TorchScript source code.
`torch._C.ScriptDict` is implemented using a modified version of pybind's `stl_bind.h`-style bindings attached to `ScriptDict`, `ScriptDictIterator` and `ScriptDictKeyIterator`, wrapper classes around `c10::impl::GenericDict` and `c10::impl::GenericDict::iterator`. These bindings allow instances of `torch._C.ScriptDict` to be used as if it were a regular `dict` Python. Reference semantics are achieved by simply retrieving the `IValue` contained in `ScriptDict` in `toIValue` (invoked when converting Python arguments to `IValues` before calling TorchScript code).
**Test Plan**
This commit adds `TestScriptDict` to `test_list_dict.py`, a set of tests
that check that all of the common dictionary operations are supported
and that instances have reference semantics across the
Python/TorchScript boundary.
Differential Revision:
D27211605
D27211605
Test Plan: Imported from OSS
Reviewed By: gmagogsfm
Pulled By: SplitInfinity
fbshipit-source-id: 446d4e5328375791aa73eb9e8b04dfe3465af960
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/58492
Update graph rewrite to specify how values in replacement pattern should
map to values in original pattern for fuse_linear pass
(Note: this ignores all push blocking failures!)
Test Plan:
python test/test_quantization.py TestQuantizeJitPasses.test_fuse_linear
Imported from OSS
Reviewed By: jerryzh168
Differential Revision: D28512464
fbshipit-source-id: 250a69cebc11eb4328a34c8f685b36e337439aae
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/58300
Current state: During graph rewriting that can fuse nodes or add nodes
result in new nodes without debug information that was available in
original node. Thus we lose this information during graph rewrite.
This PR changes graph rewriting API to let user specify how the values
in the replacement pattern map to values in the pattern to be matched.
Then the graph rewriting will copy source range and inlined callstack
from the matched nodes onto the nodes being inserted.
(Note: this ignores all push blocking failures!)
Test Plan:
python test/test_jit.py
TestJit.test_pattern_based_rewrite_with_source_range_preserved
Imported from OSS
Reviewed By: malfet
Differential Revision: D28512465
fbshipit-source-id: 863173c29de726be85b3acbd3ddf3257eea36d13
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/54809
I'm going to post on dev-discuss soon with a more thorough explanation of the design and advantages of this shape analysis, so I'm leaving out that for now.
There is still a ton left to do, I'm posting this initial version so we can get something on master multiple can work on. List of many remaining steps to do:
- [ ] Add symbolic shapes support
- [ ] Bind shape functions for operators in C++
- [ ] Make classes of operators share the same shape function (e.g. pointwise, broadcast two inputs)
- [ ] Refactor APIs
- [ ] Only iteratively optimize shape function while a change has been made
- [ ] Expand coverage of coverage to common ops
- [ ] Add shape analysis pass on Graph that handles Ifs and Loops
- [ ] Allow concurrent reads to the operator map
- [ ] Successive applications of same inputs to same shape function (e.g. series of pointwise ops)
For this review, I am mostly looking for comments related to the implementation of symolic_shape_analysis.cpp, with the caveats listed above. I am not really looking for comments related to api/registration/graph level analysis as those are all planned to be changed. I am fine landing this as is or waiting until necessary components of the TODOs above are finished.
Test Plan: Imported from OSS
Reviewed By: pbelevich
Differential Revision: D27750998
Pulled By: eellison
fbshipit-source-id: 4338b99e8651df076291c6b781c0e36a1bcbec03
Summary:
This PR adds an alternative way of calling `torch.einsum`. Instead of specifying the subscripts as letters in the `equation` parameter, one can now specify the subscripts as a list of integers as in `torch.einsum(operand1, subscripts1, operand2, subscripts2, ..., [subscripts_out])`. This would be equivalent to `torch.einsum('<subscripts1>,<subscripts2>,...,->[<subscript_out>]', operand1, operand2, ...)`
TODO
- [x] Update documentation
- [x] Add more error checking
- [x] Update tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56625
Reviewed By: zou3519
Differential Revision: D28062616
Pulled By: heitorschueroff
fbshipit-source-id: ec50ad34f127210696e7c545e4c0675166f127dc
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/55172
Description:
This is part 1 of series of PRs for supporting torch.jit.ignore as context manager. Following features are implemented in this PR:
- Unique name for the registered function under torch.jit.frontend module. The unique name is generated based on the file name and line number of context manager
- Forcing user to explicitly annotate the input and outputs.
- No side effects are considered.
Test Plan: Imported from OSS
Reviewed By: gmagogsfm
Differential Revision: D27895283
Pulled By: tugsbayasgalan
fbshipit-source-id: 5d36d9aa5d457055a6bb1676f264647a745ec36a
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/49352
In this PR, we replace all definitions of slice to take None parameters for the start, end, and step. This will simplify the compiler logic
Test Plan:
test_jit test cases
Imported from OSS
Reviewed By: jamesr66a, nikithamalgifb
Differential Revision: D25929903
fbshipit-source-id: 5bfc6bad514a8aafbef2dacc706f95f867fe85f1
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/54284
In order to bring mobile deployment, via lite interpreter, on feature
parity with JIT, with respect model level debug information we must make
model level debug information available to mobile runtime.
At the moment, model level debug information is stored in SourceRange
which associates node's of graph to where the come from in original
python source code.
This information is serialized as part of debug_pkl and deserialized
when JIT loads the model and reads the model code.
On lite interpreter, we do not have access to all the functionality of
JIT and hence we cannot load model in the same way as JIT, by reading
code, constructing module hierarchy and graph corresponding module
methods etc. Instead in, lite interpreter, only bytecode corresonding to
the compiled graph, Code, is saved.
Thus in order to annotate OPs in the bytecode with equivalent
SourceRange information we do the following:
1. During model serialization, we create a unique tag for each source
range of the model.
2. Create a map of <SourceRange, tag>
3. During debug_pkl serialization we save tag along with SourceRange, on
top of byte offset.
4. During bytecode generation, the methods of the top module are
lowered. During this process methods are inlined. In the inlined graph,
when the node of a graph is lowered to bytecode, we query node's source
range and look it up against the map.
5. Resulting source range tag is serialized in module_debug_info.
6. During model deserialization, we read all the debug_pkl records in
the archieve and create a map of <tag, SourceRange>
7. This map can be used to find source code information.
During mobile runtime:
1. We read all the debug_pkl records and create <tag=debug_handle,
SourceRange> map.
1.1 This map, MobileDebugInfo, is a member of mobile Module.
2. Interpreter catches appropriate exceptions and sets the thread local
debug handle and rethrows the exception.
3. In Function's run method we catch exception and query current debug
handle where the exception happened.
4. Query MobileDebugInfo with debug handle to retrieve source range and
augment error with source range info.
This information is still incomplete as it does not contain entire
callstack.
In the following diffs we will serialize InlinedCallStack directly.
Note that compilation is gated by SYMBOLICATE_MOBILE_DEBUG_HANDLE macro,
so that mobile builds can avoid building MobileDebugInfo, source range
and source range pickler/unpickler. Later we will add path where, if
building without debug support stack trace will contain only debug
handles. They can be symbolicated later.
Test Plan:
Ported bunch of source range tests from test_jit.py. Added on more test
in test_lite_interpreter.py
Imported from OSS
Reviewed By: raziel
Differential Revision: D27174722
fbshipit-source-id: a7b7c6088ce16dec37e823c7fefa4f0b61047e12
Summary:
Adds support for type inference of nn.Module methods using monkeytype in JIT
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57165
Reviewed By: gmagogsfm
Differential Revision: D28064983
Pulled By: nikithamalgifb
fbshipit-source-id: 303eaf8d7a27e74be09874f70f519b4c1081645b
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56147
This is support of #55686, you can see the broader context of the metaclass in
a more complete PR #56017. The short story is that in the future I want to
give Tensor a non-trivial metaclass, so to derisk the change first I give it a
trivial metaclass to shake out any bugs that might be caused by it. The
metaclass shouldn't have any performance impact on Tensor as it only gets
invoked upon subclass creation.
By the way, it was totally not documented how to create metaclasses in the Python
C API, and it took a good bit of trial error to figure it out (and the answer is
now immortalized in https://stackoverflow.com/q/67077317/23845 -- the things
that I got wrong in earlier versions of the PR included setting tp_basicsize
incorrectly, incorrectly setting Py_TPFLAGS_HAVE_GC on the metaclass--you want
to leave it unset so that it inherits, and determining that tp_init is what
actually gets called when you construct a class, not tp_call as another
not-to-be-named StackOverflow question suggests).
Aside: Ordinarily, adding a metaclass to a class is a user visible change, as
it means that it is no longer valid to mixin another class with a different
metaclass. However, because _C._TensorBase is a C extension object, it will
typically conflict with most other metaclasses, so this is not BC breaking.
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
Test Plan: Imported from OSS
Reviewed By: H-Huang
Differential Revision: D28028747
Pulled By: ezyang
fbshipit-source-id: c1e35a986aeb3db540c73d188f53dce951eeed33
Summary:
As this diff shows, currently there are a couple hundred instances of raw `noqa` in the codebase, which just ignore all errors on a given line. That isn't great, so this PR changes all existing instances of that antipattern to qualify the `noqa` with respect to a specific error code, and adds a lint to prevent more of this from happening in the future.
Interestingly, some of the examples the `noqa` lint catches are genuine attempts to qualify the `noqa` with a specific error code, such as these two:
```
test/jit/test_misc.py:27: print(f"{hello + ' ' + test}, I'm a {test}") # noqa E999
test/jit/test_misc.py:28: print(f"format blank") # noqa F541
```
However, those are still wrong because they are [missing a colon](https://flake8.pycqa.org/en/3.9.1/user/violations.html#in-line-ignoring-errors), which actually causes the error code to be completely ignored:
- If you change them to anything else, the warnings will still be suppressed.
- If you add the necessary colons then it is revealed that `E261` was also being suppressed, unintentionally:
```
test/jit/test_misc.py:27:57: E261 at least two spaces before inline comment
test/jit/test_misc.py:28:35: E261 at least two spaces before inline comment
```
I did try using [flake8-noqa](https://pypi.org/project/flake8-noqa/) instead of a custom `git grep` lint, but it didn't seem to work. This PR is definitely missing some of the functionality that flake8-noqa is supposed to provide, though, so if someone can figure out how to use it, we should do that instead.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56272
Test Plan:
CI should pass on the tip of this PR, and we know that the lint works because the following CI run (before this PR was finished) failed:
- https://github.com/pytorch/pytorch/runs/2365189927
Reviewed By: janeyx99
Differential Revision: D27830127
Pulled By: samestep
fbshipit-source-id: d6dcf4f945ebd18cd76c46a07f3b408296864fcb
Summary:
1. move module related stuff to test_module_container
2. created test_types for types and annotation
3. created test_misc for the rest
Pull Request resolved: https://github.com/pytorch/pytorch/pull/55560
Reviewed By: VitalyFedyunin
Differential Revision: D27650911
Pulled By: walterddr
fbshipit-source-id: d895a7da9e9c3d25a662a37faf4daabc276b9c1a
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/52881
**This PR adds:**
1. logic to parse complex constants (complex literals of the form `bj`)
2. logic to parse complex lists
3. support for complex constructors: `complex(tensor/int/float/bool, tensor/int/float/bool)`
4. Limited operator support
- `add`, `sub`, `mul`, `torch.tensor`, `torch.as_tensor`
**Follow-up work:**
1. Add complex support for unary and other registered ops.
2. support complex constructor with string as input (this is supported in Python eager mode).
3. Test all emitXYZ for all XYZ in `ir_emitter.cpp` (currently only emitConst, emitValueToTensor are tested). e.g., test loops etc.
4. onnx doesn't support complex tensors, so we should error out with a clear and descriptive error message.
Test Plan: Imported from OSS
Reviewed By: bdhirsh
Differential Revision: D27245059
Pulled By: anjali411
fbshipit-source-id: af043b5159ae99a9cc8691b5a8401503fa8d6f05
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/53908
This adds reinplacing to MKLDNN Subgraphs so that we replace `aten::add` with `aten::add_`. Normally you would have to prove device and dtype, but we know that already, and because we have explicit broadcast nodes for other reasons we dont have to prove that the output shape of add is the same as inputs.
Ive tested correctness on resnet, I'm going to do more extensive testing as well. When I benchmarked the "unsafe" version (always inplace) I saw average speedups of ~16% for both Single threaded and Multithreaded. I dont think the "safe" version will be far beyond; when I looked at resnet for example every `add` and `relu` were reinplaced.
Theres some question of reusing other alias / liveness / inplacing passes in SR. I thought about it, however I didnt want to add a cross-dependency between very different parts of the code base with a bunch of different assumptions. The logic here is also covering a simpler case and does not add much complexity IMO.
Test Plan: Imported from OSS
Reviewed By: Krovatkin
Differential Revision: D27132969
Pulled By: eellison
fbshipit-source-id: 121a38daaedf01363f6b66a814beaaa72a0ab0dc
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/53682
With this, under the meta device, 101 tests passed and 16953 skipped.
It ain't much, but it's a start.
Some various bits and bobs:
- NotImplementedError suppression at test level is implemented
in the same way as CUDA memory leak check, i.e., by wrapping
test methods and monkeypatching them back in.
- I had to reimplement assertRaises/assertRaisesRegex from scratch to
ignore NotImplementedError when _ignore_not_implemented_error is True.
The implementation relies on a small amount of private API that hasn't
changed since 2010
- expectedAlertNondeterministic doesn't really work so I skipped them
all; there's probably a way to do it better
I tested this using `pytest --disable-warnings --tb=native -k meta --sw
test/*.py` and a pile of extra patches to make collection actually work
(lol).
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
Test Plan: Imported from OSS
Reviewed By: mruberry
Differential Revision: D26955539
Pulled By: ezyang
fbshipit-source-id: ac21c8734562497fdcca3b614a28010bc4c03d74
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/53424
Fixes https://github.com/pytorch/pytorch/issues/24807 and supersedes the stale https://github.com/pytorch/pytorch/issues/25093 (Cc Microsheep). If you now run the reproduction
```python
import torch
if __name__ == "__main__":
t = torch.tensor([1, 2, 3], dtype=torch.float64)
```
with `pylint==2.6.0`, you get the following output
```
test_pylint.py:1:0: C0114: Missing module docstring (missing-module-docstring)
test_pylint.py:4:8: E1101: Module 'torch' has no 'tensor' member; maybe 'Tensor'? (no-
member)
test_pylint.py:4:38: E1101: Module 'torch' has no 'float64' member (no-member)
```
Now `pylint` doesn't recognize `torch.tensor` at all, but it is promoted in the stub. Given that it also doesn't recognize `torch.float64`, I think fixing this is out of scope of this PR.
---
## TL;DR
This BC-breaking only for users that rely on unintended behavior. Since `torch/__init__.py` loaded `torch/tensor.py` it was populated in `sys.modules`. `torch/__init__.py` then overwrote `torch.tensor` with the actual function. With this `import torch.tensor as tensor` does not fail, but returns the function rather than the module. Users that rely on this import need to change it to `from torch import tensor`.
Reviewed By: zou3519
Differential Revision: D26223815
Pulled By: bdhirsh
fbshipit-source-id: 125b9ff3d276e84a645cd7521e8d6160b1ca1c21
Summary:
Previously `torch.jit.trace` relies on AutoGrad hooks to infer name of tensors in computation, including those of function/method arguments. This often doesn't work out because:
- These names often do not exist
- Tracer uses argument name of first tensor operation on each tensor as inferred argument names. These tensor operations have programmatically-generated names like `argument_1`
This PR extracts argument names directly from Python functions and pass them down to tracer, which then assigns them to correct graph inputs. This way, we always have the correct argument names captured in IR.
This is useful for both debugging and supporting using `InterfaceType` to represent traced modules.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/51775
Reviewed By: izdeby
Differential Revision: D26273105
Pulled By: gmagogsfm
fbshipit-source-id: 934a385041137dc3731bb6fa8657b11532fed9e5
Summary:
Previously TorchScript allows a ignore-all type check suppression rule that looks like
```
code code code # type: ignore
```
But a more common use case is
```
code code code # type: ignore[specific-rule]
```
This PR allows the more common use case
Fixes https://github.com/pytorch/pytorch/issues/48643
Pull Request resolved: https://github.com/pytorch/pytorch/pull/51675
Reviewed By: ansley
Differential Revision: D26304870
Pulled By: gmagogsfm
fbshipit-source-id: 0ac9ee34f0219c86e428318a69484d5aa3ec433f
