Skip to content

Improve redundant slice removal #2441

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jul 9, 2025
Merged

Improve redundant slice removal #2441

merged 2 commits into from
Jul 9, 2025

Conversation

gramalingam
Copy link
Collaborator

Improve the optimization for removal of redundant slices. (It doesn't currently handle dynamic shapes.) The optimization is fairly simple, and eliminates the slice when the input and output shapes are same.

Signed-off-by: Ganesan Ramalingam <[email protected]>
Copy link

codecov bot commented Jul 9, 2025

❌ 4 Tests Failed:

Tests completed Failed Passed Skipped
14955 4 14951 2348
View the top 3 failed test(s) by shortest run time
onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0212_test_cast_STRING_to_FLOAT
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_cast_STRING_to_FLOAT'

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_cast_STRING_to_FLOAT' (e=No module named 'tests.onnx_backend_test_code.test_cast_STRING_to_FLOAT') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_cast_STRING_to_FLOAT.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_cast_STRING_to_FLOAT.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, STRING
E   from onnxscript.onnx_opset import opset21
E   
E   @script()
E   def bck_test_cast_STRING_to_FLOAT(input: STRING[3,4]) -> (FLOAT[3,4]):
E       output = opset21.Cast(input, to=1)
E       return output
onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0492_test_isinf_positive
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_isinf_positive'

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_isinf_positive' (e=No module named 'tests.onnx_backend_test_code.test_isinf_positive') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_isinf_positive.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_isinf_positive.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, FLOAT
E   from onnxscript.onnx_opset import opset20
E   
E   @script()
E   def bck_test_isinf_positive(x: FLOAT[6]) -> (BOOL[6]):
E       y = opset20.IsInf(x, detect_negative=0)
E       return y
onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0606_test_matmul_3d
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_matmul_3d'

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_matmul_3d' (e=No module named 'tests.onnx_backend_test_code.test_matmul_3d') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_matmul_3d.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_matmul_3d.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_matmul_3d(a: FLOAT[2,3,4], b: FLOAT[2,4,3]) -> (FLOAT[2,3,3]):
E       c = opset13.MatMul(a, b)
E       return c

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

@github-project-automation github-project-automation bot moved this from Todo to Done in ONNX Script Review Board Jul 9, 2025
@justinchuby
Copy link
Collaborator

Maybe need to update the test: test_slice_pattern_is_not_matched_when_input_is_dynamic

Signed-off-by: Ganesan Ramalingam <[email protected]>
@gramalingam gramalingam enabled auto-merge (squash) July 9, 2025 22:24
@gramalingam gramalingam merged commit f42c2bb into main Jul 9, 2025
23 of 29 checks passed
@gramalingam gramalingam deleted the rama/slice-opt branch July 9, 2025 22:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
Development

Successfully merging this pull request may close these issues.

2 participants