Update autocast.py to fix attribute creation error

[justinchuby]

This change should fix the type of errors like below (reported in pytorch/pytorch#153214 (comment)):

  File "/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/python_message.py", line 589, in init
    _ReraiseTypeErrorWithFieldName(message_descriptor.name, field_name)
  File "/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/python_message.py", line 478, in _ReraiseTypeErrorWithFieldName
    raise exc.with_traceback(sys.exc_info()[2])
  File "/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/python_message.py", line 587, in init
    setattr(self, field_name, field_value)
  File "/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/python_message.py", line 730, in field_setter
    raise TypeError(
TypeError: Cannot set onnx.AttributeProto.type to <AttrType.TENSOR: 4>: <AttrType.TENSOR: 4> has type <class 'onnx.onnx_cpp2py_export.defs.OpSchema.AttrType'>, but expected one of: (<class 'int'>,) for field AttributeProto.type

[codecov]

❌ 4 Tests Failed:

Tests completed	Failed	Passed	Skipped
14865	4	14861	2354

View the top 3 failed test(s) by shortest run time

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0118_test_bitwise_and_i32_2d

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_bitwise_and_i32_2d'

The above exception was the direct cause of the following exception:
.nox\test_onnx_ir_git\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_bitwise_and_i32_2d' (e=No module named 'tests.onnx_backend_test_code.test_bitwise_and_i32_2d') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_bitwise_and_i32_2d.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_bitwise_and_i32_2d.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import INT32
E   from onnxscript.onnx_opset import opset18
E   
E   @script()
E   def bck_test_bitwise_and_i32_2d(x: INT32[3,4], y: INT32[3,4]) -> (INT32[3,4]):
E       bitwiseand = opset18.BitwiseAnd(x, y)
E       return bitwiseand

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0257_test_col2im_pads

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_col2im_pads'

The above exception was the direct cause of the following exception:
.nox\test_onnx_ir_git\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_col2im_pads' (e=No module named 'tests.onnx_backend_test_code.test_col2im_pads') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_col2im_pads.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_col2im_pads.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import FLOAT, INT64
E   from onnxscript.onnx_opset import opset18
E   
E   @script()
E   def bck_test_col2im_pads(input: FLOAT[1,5,15], image_shape: INT64[2], block_shape: INT64[2]) -> (FLOAT[1,1,5,5]):
E       output = opset18.Col2Im(input, image_shape, block_shape, pads=[0, 1, 0, 1])
E       return output

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0756_test_or_bcast3v2d

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_or_bcast3v2d'

The above exception was the direct cause of the following exception:
.nox\test_onnx_ir_git\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_or_bcast3v2d' (e=No module named 'tests.onnx_backend_test_code.test_or_bcast3v2d') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_or_bcast3v2d.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_or_bcast3v2d.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import BOOL
E   from onnxscript.onnx_opset import opset7
E   
E   @script()
E   def bck_test_or_bcast3v2d(x: BOOL[3,4,5], y: BOOL[4,5]) -> (BOOL[3,4,5]):
E       r_or = opset7.Or(x, y)
E       return r_or

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[Copilot]

Pull Request Overview

This PR updates the internal helper for creating ONNX AttributeProto objects so that the type field is passed as an integer rather than an enum, fixing type errors when setting attributes.

• Changed the attr_type annotation to use Python’s union syntax.
• Wrapped attr_type in int() for all onnx.AttributeProto constructor calls.

Comments suppressed due to low confidence (1)

onnxscript/_internal/autocast.py:66

• Add unit tests covering cases where attr_type is a tensor and where value is an empty list to ensure the int(attr_type) conversion is applied correctly.

return onnx.AttributeProto(name=key, type=int(attr_type))

[justinchuby]

Updated to fix the callers instead

[titaiwangms]

Are failing checks more than usual?

[justinchuby]

Are failing checks more than usual?

They seem like the usual windows flakiness & unresolved issues with the latest ort. We should fix them though at some point.