[Feature] Add is_tracing to wrap torch.jit.is_tracing in different versions. (#1187)

* Add `is_tracing` to wrap `torch.jit.is_tracing` in different versions.

* Remame `is_tracing` to `is_jit_tracing`

* Ignore `is_jit_tracing` tests in CI.

.github/workflows/build.yml

@@ -48,7 +48,7 @@ jobs:
       - name: Run unittests and generate coverage report
         run: |
           pip install -r requirements/test.txt
-          pytest tests/ --ignore=tests/test_runner --ignore=tests/test_optimizer.py --ignore=tests/test_cnn --ignore=tests/test_parallel.py --ignore=tests/test_ops --ignore=tests/test_load_model_zoo.py --ignore=tests/test_utils/test_logging.py --ignore=tests/test_image/test_io.py --ignore=tests/test_utils/test_registry.py --ignore=tests/test_utils/test_parrots_jit.py
+          pytest tests/ --ignore=tests/test_runner --ignore=tests/test_optimizer.py --ignore=tests/test_cnn --ignore=tests/test_parallel.py --ignore=tests/test_ops --ignore=tests/test_load_model_zoo.py --ignore=tests/test_utils/test_logging.py --ignore=tests/test_image/test_io.py --ignore=tests/test_utils/test_registry.py --ignore=tests/test_utils/test_parrots_jit.py --ignore=tests/test_utils/test_trace.py
 
   build_without_ops:
     runs-on: ubuntu-18.04

mmcv/utils/__init__.py

@@ -45,6 +45,7 @@ else:
         _AdaptiveMaxPoolNd, _AvgPoolNd, _BatchNorm, _ConvNd,
         _ConvTransposeMixin, _InstanceNorm, _MaxPoolNd, get_build_config)
     from .registry import Registry, build_from_cfg
+    from .trace import is_jit_tracing
     __all__ = [
         'Config', 'ConfigDict', 'DictAction', 'collect_env', 'get_logger',
         'print_log', 'is_str', 'iter_cast', 'list_cast', 'tuple_cast',
@@ -63,5 +64,5 @@ else:
         'assert_dict_contains_subset', 'assert_attrs_equal',
         'assert_dict_has_keys', 'assert_keys_equal', 'assert_is_norm_layer',
         'assert_params_all_zeros', 'check_python_script',
-        'is_method_overridden'
+        'is_method_overridden', 'is_jit_tracing'
     ]

mmcv/utils/trace.py

@@ -0,0 +1,21 @@
+import warnings
+from distutils.version import LooseVersion
+
+import torch
+
+
+def is_jit_tracing() -> bool:
+    if LooseVersion(torch.__version__) >= LooseVersion('1.6.0'):
+        on_trace = torch.jit.is_tracing()
+        # In PyTorch 1.6, torch.jit.is_tracing has a bug.
+        # Refers to https://github.com/pytorch/pytorch/issues/42448
+        if isinstance(on_trace, bool):
+            return on_trace
+        else:
+            return torch._C._is_tracing()
+    else:
+        warnings.warn(
+            'torch.jit.is_tracing is only supported after v1.6.0. '
+            'Therefore is_tracing returns False automatically. Please '
+            'set on_trace manually if you are using trace.', UserWarning)
+        return False

tests/test_utils/test_trace.py

@@ -0,0 +1,26 @@
+from distutils.version import LooseVersion
+
+import pytest
+import torch
+
+from mmcv.utils import is_jit_tracing
+
+
+@pytest.mark.skipif(
+    LooseVersion(torch.__version__) < LooseVersion('1.6.0'),
+    reason='torch.jit.is_tracing is not available before 1.6.0')
+def test_is_jit_tracing():
+
+    def foo(x):
+        if is_jit_tracing():
+            return x
+        else:
+            return x.tolist()
+
+    x = torch.rand(3)
+    # test without trace
+    assert isinstance(foo(x), list)
+
+    # test with trace
+    traced_foo = torch.jit.trace(foo, (torch.rand(1), ))
+    assert isinstance(traced_foo(x), torch.Tensor)