Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/49765
Some PyTorch module can have None as submodule, which causes the following error in JIT-tracing:
Repro script:
```
import torch
class TestModule(torch.nn.Module):
def __init__(self):
super().__init__()
self.submod = torch.nn.Linear(3, 4)
self.submod = None
def forward(self, inputs):
return inputs
m = TestModule()
tm = torch.jit.trace(m, torch.tensor(1.))
```
Error:
```
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 742, in trace
_module_class,
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 928, in trace_module
module = make_module(mod, _module_class, _compilation_unit)
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 560, in make_module
return _module_class(mod, _compilation_unit=_compilation_unit)
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 1039, in __init__
submodule, TracedModule, _compilation_unit=None
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 560, in make_module
return _module_class(mod, _compilation_unit=_compilation_unit)
File "/data/miniconda3/envs/master_nightly/lib/python3.7/site-packages/torch/jit/_trace.py", line 988, in __init__
assert isinstance(orig, torch.nn.Module)
AssertionError
```
This pull request changes the JIT-tracing logic to skip the None submodule when tracing.
Test Plan: `buck test mode/dev //caffe2/test:jit -- test_trace_skip_none_submodule`
Reviewed By: wanchaol
Differential Revision: D25670948
fbshipit-source-id: 468f42f5ddbb8fd3de06d0bc224dc67bd7172358
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/47211
The attribute is getting shadowed by the default one set on all modules,
and the __setattr__ on the TracedModule object prevents setting it correctly.
import torch
inp = torch.zeros(1, 3, 224, 224)
model = torch.hub.load('pytorch/vision:v0.6.0', 'mobilenet_v2', pretrained=True)
model.eval()
print(model.training)
with torch.no_grad():
traced = torch.jit.trace(model, inp)
print(traced.training)
traced.eval()
print(traced.training)
traced.training = False
print(traced.training)
torch.jit.freeze(traced)
Test Plan: Imported from OSS
Reviewed By: suo
Differential Revision: D24686690
Pulled By: zdevito
fbshipit-source-id: 9c1678dc68e9bf83176e9f5a20fa8f6bff5d69a0
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/43872
This PR allows the recursive scripting to have a separate
submodule_stubs_fn to create its submodule with specific user provided
rules.
Fixes https://github.com/pytorch/pytorch/issues/43729
Test Plan: Imported from OSS
Reviewed By: suo
Differential Revision: D23430176
Pulled By: wanchaol
fbshipit-source-id: 20530d7891ac3345b36f1ed813dc9c650b28d27a
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/42486
**Summary**
This commit fixes a small bug in which `torch.jit.is_tracing()` returns
`torch._C.is_tracing`, the function object, instead of calling the
function and returning the result.
**Test Plan**
Continuous integration?
**Fixes**
This commit fixes#42448.
Test Plan: Imported from OSS
Reviewed By: bertmaher
Differential Revision: D22911062
Pulled By: SplitInfinity
fbshipit-source-id: b94eca0c1c65ca6f22acc6c5542af397f2dc37f0
Summary:
solves most of gh-38011 in the framework of solving gh-32703.
These should only be formatting fixes, I did not try to fix grammer and syntax.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/41068
Differential Revision: D22411919
Pulled By: zou3519
fbshipit-source-id: 25780316b6da2cfb4028ea8a6f649bb18b746440
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/40807
We pack a lot of logic into `jit/__init__.py`, making it unclear to
developers and users which parts of our API are public vs. internal. This
is one in a series of PRs intended to pull implementation out into
separate files, and leave `__init__.py` as a place to register the
public API.
This PR moves all the tracing-related stuff out, and fixes other spots up
as necessary. Followups will move other core APIs out.
The desired end-state is that we conform to the relevant rules in [PEP 8](https://www.python.org/dev/peps/pep-0008/#public-and-internal-interfaces). In particular:
- Internal implementation goes in modules prefixed by `_`.
- `__init__.py` exposes a public API from these private modules, and nothing more.
- We set `__all__` appropriately to declare our public API.
- All use of JIT-internal functionality outside the JIT are removed (in particular, ONNX is relying on a number internal APIs). Since they will need to be imported explicitly, it will be easier to catch new uses of internal APIs in review.
Test Plan: Imported from OSS
Reviewed By: eellison
Differential Revision: D22320645
Pulled By: suo
fbshipit-source-id: 0720ea9976240e09837d76695207e89afcc58270
