[torchlib] Set allowzero=True on Reshape where appropriate

[justinchuby]

When we reshape from a dynamic shape, the shape can contain zeros. This change accounts for those cases.

[Copilot]

Pull Request Overview

This PR updates multiple calls to op.Reshape across the torch library to include allowzero=True, ensuring that reshaping operations correctly handle dynamic shapes that may contain zeros.

• Updated op.Reshape invocations in several functions to pass allowzero=True.
• Adjusted functions in instance normalization, pixel shuffle/unshuffle, unique operations, and view operations.

Comments suppressed due to low confidence (1)

onnxscript/function_libs/torch_lib/ops/core.py:4534

• Consider adding unit tests that cover scenarios where dynamic shapes include zeros to verify that op.Reshape correctly handles these cases when allowzero=True is set.

allowzero=True

[codecov]

❌ 14 Tests Failed:

Tests completed	Failed	Passed	Skipped
16742	14	16728	2523

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

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0458_test_hardmax_axis_1

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\3.12.10\x64\Lib\importlib\__init__.py:90: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_hardmax_axis_1'

The above exception was the direct cause of the following exception:
.nox\test\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_hardmax_axis_1' (e=No module named 'tests.onnx_backend_test_code.test_hardmax_axis_1') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_hardmax_axis_1.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_hardmax_axis_1.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
E   from onnxscript.onnx_opset import opset13
E   
E   @script()
E   def bck_test_hardmax_axis_1(x: FLOAT[3,4,5]) -> (FLOAT[3,4,5]):
E       y = opset13.Hardmax(x, axis=1)
E       return y

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_1127_test_slice_neg

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\3.12.10\x64\Lib\importlib\__init__.py:90: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_slice_neg'

The above exception was the direct cause of the following exception:
.nox\test\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_slice_neg' (e=No module named 'tests.onnx_backend_test_code.test_slice_neg') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_slice_neg.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_slice_neg.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 opset13
E   
E   @script()
E   def bck_test_slice_neg(x: FLOAT[20,10,5], starts: INT64[1], ends: INT64[1], axes: INT64[1], steps: INT64[1]) -> (FLOAT[20,9,5]):
E       y = opset13.Slice(x, starts, ends, axes, steps)
E       return y

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0919_test_reduce_sum_square_do_not_keepdims_example_expanded

Stack Traces | 0.004s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
C:\hostedtoolcache\windows\Python\3.12.10\x64\Lib\importlib\__init__.py:90: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_reduce_sum_square_do_not_keepdims_example_expanded'

The above exception was the direct cause of the following exception:
.nox\test\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_reduce_sum_square_do_not_keepdims_example_expanded' (e=No module named 'tests.onnx_backend_test_code.test_reduce_sum_square_do_not_keepdims_example_expanded') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_reduce_sum_square_do_not_keepdims_example_expanded.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_reduce_sum_square_do_not_keepdims_example_expanded.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_reduce_sum_square_do_not_keepdims_example_expanded(data: FLOAT[3,2,2], axes: INT64[1]) -> (FLOAT[3,2]):
E       ReduceSumSquare_test_reduce_sum_square_do_not_keepdims_example_expanded_function_data_square = opset18.Mul(data, data)
E       reduced = opset18.ReduceSum(ReduceSumSquare_test_reduce_sum_square_do_not_keepdims_example_expanded_function_data_square, axes, keepdims=0)
E       return reduced

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

[justinchuby]

@gramalingam could you reapprove? I fixed the unflatten test case