Skip to content

Extend basic matcher to handle multiple-output-nodes #7643

Extend basic matcher to handle multiple-output-nodes

Extend basic matcher to handle multiple-output-nodes #7643

GitHub Actions / Test Results failed Jul 24, 2024 in 0s

30 fail, 1 538 skipped, 11 818 pass in 3h 26m 42s

     24 files  ±      0      24 suites  ±0   3h 26m 42s ⏱️ + 1h 26m 10s
 13 386 tests +    438  11 818 ✅ + 2 005    1 538 💤  -   1 596   30 ❌ + 29 
477 208 runs  +165 760  99 962 ✅ +25 778  376 978 💤 +139 723  268 ❌ +259 

Results for commit ee10468. ± Comparison against earlier commit 65f647d.

Annotations

Check warning on line 0 in tests.function_libs.torch_lib.quantization_test.QuantizedModelExportTest

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 21 runs failed: test_simple_quantized_model (tests.function_libs.torch_lib.quantization_test.QuantizedModelExportTest)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 1s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 3s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 2s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 4s]
Raw output
torch.onnx.OnnxExporterError: Failed to export the model to ONNX. Generating SARIF report at 'report_dynamo_export.sarif'. SARIF is a standard format for the output of static analysis tools. SARIF logs can be loaded in VS Code SARIF viewer extension, or SARIF web viewer (https://microsoft.github.io/sarif-web-component/). Please report a bug on PyTorch Github: https://github.com/pytorch/pytorch/issues
.nox\test_torch_nightly\Lib\site-packages\torch\onnx\_internal\exporter.py:1476: in dynamo_export
    ).export()
.nox\test_torch_nightly\Lib\site-packages\torch\onnx\_internal\exporter.py:1202: in export
    graph_module = self.options.fx_tracer.generate_fx(
.nox\test_torch_nightly\Lib\site-packages\torch\onnx\_internal\fx\dynamo_graph_extractor.py:196: in generate_fx
    graph_module, graph_guard = torch._dynamo.export(
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\eval_frame.py:1399: in inner
    result_traced = opt_f(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\eval_frame.py:448: in _fn
    return fn(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\external_utils.py:38: in inner
    return fn(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\onnx\_internal\fx\dynamo_graph_extractor.py:151: in wrapped
    return output_adapter.apply(model_func(*args, **kwargs), model=model)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:1170: in __call__
    return self._torchdynamo_orig_callable(
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:499: in __call__
    return _compile(
C:\hostedtoolcache\windows\Python\3.12.4\x64\Lib\contextlib.py:81: in inner
    return func(*args, **kwds)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:850: in _compile
    guarded_code = compile_inner(code, one_graph, hooks, transform)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\utils.py:246: in time_wrapper
    r = func(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_utils_internal.py:85: in wrapper_function
    return StrobelightCompileTimeProfiler.profile_compile_time(
.nox\test_torch_nightly\Lib\site-packages\torch\_strobelight\compile_time_profiler.py:129: in profile_compile_time
    return func(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:668: in compile_inner
    out_code = transform_code_object(code, transform)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\bytecode_transformation.py:1322: in transform_code_object
    transformations(instructions, code_options)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:194: in _fn
    return fn(*args, **kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\convert_frame.py:610: in transform
    tracer.run()
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:2613: in run
    super().run()
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:910: in run
    while self.step():
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:822: in step
    self.dispatch_table[inst.opcode](self, inst)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:514: in wrapper
    return inner_fn(self, inst)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:2153: in CALL
    self._call(inst)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:2148: in _call
    self.call_function(fn, args, kwargs)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\symbolic_convert.py:760: in call_function
    self.push(fn.call_function(self, args, kwargs))
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\variables\functions.py:687: in call_function
    unimplemented(msg)
.nox\test_torch_nightly\Lib\site-packages\torch\_dynamo\exc.py:224: in unimplemented
    raise Unsupported(msg)
E   torch._dynamo.exc.Unsupported: 'skip function tree_flatten_spec in file D:\a\onnxscript\onnxscript\.nox\test_torch_nightly\Lib\site-packages\torch\fx\_pytree.py'
E   
E   from user code:
E      File "<eval_with_key>.12", line 5, in forward
E       arg0, = fx_pytree.tree_flatten_spec(([x], {}), self._in_spec)
E   
E   Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\quantization_test.py:49: in test_simple_quantized_model
    program = torch.onnx.dynamo_export(pt2e_torch_model, *example_inputs)
.nox\test_torch_nightly\Lib\site-packages\torch\onnx\_internal\exporter.py:1487: in dynamo_export
    raise OnnxExporterError(
E   torch.onnx.OnnxExporterError: Failed to export the model to ONNX. Generating SARIF report at 'report_dynamo_export.sarif'. SARIF is a standard format for the output of static analysis tools. SARIF logs can be loaded in VS Code SARIF viewer extension, or SARIF web viewer (https://microsoft.github.io/sarif-web-component/). Please report a bug on PyTorch Github: https://github.com/pytorch/pytorch/issues

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

4 out of 24 runs failed: test_output_match_opinfo__matmul_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 25 (4.0%)
Greatest absolute difference: 0.03515625 at index (4, 4) (up to 0.02 allowed)
Greatest relative difference: 0.0084075927734375 at index (4, 4) (up to 0.002 allowed)
AssertionError: Tensor-likes are not close!

Mismatched elements: 2 / 250 (0.8%)
Greatest absolute difference: 0.029296875 at index (0, 1, 7) (up to 0.02 allowed)
Greatest relative difference: 0.019683837890625 at index (4, 0, 7) (up to 0.002 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 25 (4.0%)
E   Greatest absolute difference: 0.03515625 at index (4, 4) (up to 0.02 allowed)
E   Greatest relative difference: 0.0084075927734375 at index (4, 4) (up to 0.002 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 2 / 250 (0.8%)
E   Greatest absolute difference: 0.029296875 at index (0, 1, 7) (up to 0.02 allowed)
E   Greatest relative difference: 0.019683837890625 at index (4, 0, 7) (up to 0.002 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__clamp_cpu_float32 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 2s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 1s]
Raw output
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1653: in aten_clamp
    min_clamp = op.CastLike(min, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x116d0edaca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x116d0eda7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x116d0ed9b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1657: in aten_clamp
    max_clamp = op.CastLike(max, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x116d0edaca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x116d0eda7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x116d0ed9b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_float32 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 10 / 25 (40.0%)
Greatest absolute difference: 12.9130220413208 at index (1, 1) (up to 1e-05 allowed)
Greatest relative difference: 3.1326284408569336 at index (1, 0) (up to 1.3e-06 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 10 / 25 (40.0%)
E   Greatest absolute difference: 12.9130220413208 at index (1, 1) (up to 1e-05 allowed)
E   Greatest relative difference: 3.1326284408569336 at index (1, 0) (up to 1.3e-06 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__sub_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 2 / 25 (8.0%)
Greatest absolute difference: 0.001953125 at index (3, 3) (up to 1e-05 allowed)
Greatest relative difference: 0.0028400421142578125 at index (3, 3) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 2 / 25 (8.0%)
E   Greatest absolute difference: 0.001953125 at index (3, 3) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0028400421142578125 at index (3, 3) (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

2 out of 24 runs failed: test_output_match_opinfo__topk_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
Raw output
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 10 / 25 (40.0%)
Greatest absolute difference: 13.84375 at index (3, 0) (up to 1e-05 allowed)
Greatest relative difference: 7.5 at index (1, 4) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 10 / 25 (40.0%)
E   Greatest absolute difference: 13.84375 at index (3, 0) (up to 1e-05 allowed)
E   Greatest relative difference: 7.5 at index (1, 4) (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__rsub_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 2 / 25 (8.0%)
Greatest absolute difference: 0.00390625 at index (0, 2) (up to 1e-05 allowed)
Greatest relative difference: 0.0017986297607421875 at index (0, 4) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 2 / 25 (8.0%)
E   Greatest absolute difference: 0.00390625 at index (0, 2) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0017986297607421875 at index (0, 4) (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__clamp_cpu_int64 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 2s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 1s]
Raw output
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1653: in aten_clamp
    min_clamp = op.CastLike(min, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x2b3c962aca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x2b3c962a7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x2b3c9629b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1657: in aten_clamp
    max_clamp = op.CastLike(max, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x2b3c962aca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x2b3c962a7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x2b3c9629b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_int32 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not equal!

Mismatched elements: 9 / 25 (36.0%)
Greatest absolute difference: 15 at index (1, 0)
Greatest relative difference: 6.0 at index (3, 4)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 9 / 25 (36.0%)
E   Greatest absolute difference: 15 at index (1, 0)
E   Greatest relative difference: 6.0 at index (3, 4)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__clamp_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 1s]
Raw output
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1653: in aten_clamp
    min_clamp = op.CastLike(min, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x2b3c962aca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x2b3c962a7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x2b3c9629b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'
tests\function_libs\torch_lib\ops_test.py:215: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests\function_libs\torch_lib\ops_test_common.py:515: in _capture_graph_and_evaluate_torch_script_evaluator
    symbolic_outputs = function(*onnxscript_args, **onnxscript_kwargs)
onnxscript\values.py:583: in __call__
    return self.func(*args, **kwargs)
onnxscript\function_libs\torch_lib\ops\core.py:1657: in aten_clamp
    max_clamp = op.CastLike(max, self)
onnxscript\onnx_opset\_impl\opset15.py:257: in CastLike
    return op(*self._prepare_inputs(schema, input, target_type))
onnxscript\values.py:301: in __call__
    return evaluator.default().eval(schema, args, kwargs)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:341: in eval
    return self._graph.add_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph.add_op_call) at 0x2b3c962aca0>:111: in add_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:878: in add_op_call
    result = self._add_torchscript_op_call(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch.TorchScriptGraph._add_torchscript_op_call) at 0x2b3c962a7a0>:129: in _add_torchscript_op_call
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:758: in _add_torchscript_op_call
    result = _create_op_call_in_torch_graph(
<@beartype(onnxscript.function_libs.torch_lib.graph_building._graph_building_torch._create_op_call_in_torch_graph) at 0x2b3c9629b20>:113: in _create_op_call_in_torch_graph
    ???
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:498: in _create_op_call_in_torch_graph
    _add_attribute_to_torchscript_node(node, key, value)
onnxscript\function_libs\torch_lib\graph_building\_graph_building_torch.py:463: in _add_attribute_to_torchscript_node
    raise TypeError(f"Unsupported attribute type '{type(value)}' for attribute '{key}'")
E   TypeError: Unsupported attribute type '<class 'torch._C._onnx.TensorProtoDataType'>' for attribute 'to'

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__addmv_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 0.046875 at index (2,) (up to 0.01 allowed)
Greatest relative difference: 0.002635955810546875 at index (2,) (up to 0.001 allowed)
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 0.0234375 at index (2,) (up to 0.01 allowed)
Greatest relative difference: 0.0024318695068359375 at index (2,) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 5 (20.0%)
E   Greatest absolute difference: 0.046875 at index (2,) (up to 0.01 allowed)
E   Greatest relative difference: 0.002635955810546875 at index (2,) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 5 (20.0%)
E   Greatest absolute difference: 0.0234375 at index (2,) (up to 0.01 allowed)
E   Greatest relative difference: 0.0024318695068359375 at index (2,) (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__floor_divide_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
Failed: Unexpected success
Unexpected success

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__erfc_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 20 (5.0%)
Greatest absolute difference: 0.00023031234741210938 at index (2,) (up to 0.0002 allowed)
Greatest relative difference: 0.892578125 at index (2,) (up to 0.01 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 20 (5.0%)
E   Greatest absolute difference: 0.00023031234741210938 at index (2,) (up to 0.0002 allowed)
E   Greatest relative difference: 0.892578125 at index (2,) (up to 0.01 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__var_mean_dim_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 1s]
Raw output
AssertionError: Output 0 mismatch
AssertionError: Output 0 mismatch
AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 25 (4.0%)
E   Greatest absolute difference: 0.0078125 at index (3, 3) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0010890960693359375 at index (3, 3) (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 25 (4.0%)
E   Greatest absolute difference: 0.0078125 at index (3, 0, 3) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0010890960693359375 at index (3, 0, 3) (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 4 / 20 (20.0%)
E   Greatest absolute difference: 0.0625 at index (2, 0) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0014753341674804688 at index (0, 1) (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_int64 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyFullGraphCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not equal!

Mismatched elements: 9 / 25 (36.0%)
Greatest absolute difference: 15 at index (1, 0)
Greatest relative difference: 6.0 at index (3, 4)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 9 / 25 (36.0%)
E   Greatest absolute difference: 15 at index (1, 0)
E   Greatest relative difference: 6.0 at index (3, 4)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__var_mean_unbiased_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Scalars are not close!
E   
E   Expected 25.765625 but got 25.796875.
E   Absolute difference: 0.03125 (up to 1e-05 allowed)
E   Relative difference: 0.001212856276531231 (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_float32 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 10 / 25 (40.0%)
Greatest absolute difference: 12.9130220413208 at index (1, 1) (up to 1e-05 allowed)
Greatest relative difference: 3.1326284408569336 at index (1, 0) (up to 1.3e-06 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 10 / 25 (40.0%)
E   Greatest absolute difference: 12.9130220413208 at index (1, 1) (up to 1e-05 allowed)
E   Greatest relative difference: 3.1326284408569336 at index (1, 0) (up to 1.3e-06 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__ops_aten_embedding_bag_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 2s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 6s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 25s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 1s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 6s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 26s]
Raw output
AssertionError: Output 0 mismatch
AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 15 (6.7%)
E   Greatest absolute difference: 0.0546875 at index (2, 0) (up to 0.01 allowed)
E   Greatest relative difference: 0.0124359130859375 at index (2, 0) (up to 0.01 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 15 (6.7%)
E   Greatest absolute difference: 0.0546875 at index (2, 0) (up to 0.01 allowed)
E   Greatest relative difference: 0.0124359130859375 at index (2, 0) (up to 0.01 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__var_unbiased_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Scalars are not close!

Expected 25.765625 but got 25.796875.
Absolute difference: 0.03125 (up to 1e-05 allowed)
Relative difference: 0.001212856276531231 (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Scalars are not close!
E   
E   Expected 25.765625 but got 25.796875.
E   Absolute difference: 0.03125 (up to 1e-05 allowed)
E   Relative difference: 0.001212856276531231 (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__div_mode_floor_rounding_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
Failed: Unexpected success
Unexpected success

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

2 out of 24 runs failed: test_output_match_opinfo__topk_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch
tests/function_libs/torch_lib/ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 3 / 75 (4.0%)
E   Greatest absolute difference: 8 at index (4, 2, 0)
E   Greatest relative difference: 3.0 at index (4, 1, 4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 1 mismatch

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__addmv_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 0.046875 at index (2,) (up to 0.01 allowed)
Greatest relative difference: 0.002635955810546875 at index (2,) (up to 0.001 allowed)
AssertionError: Tensor-likes are not close!

Mismatched elements: 1 / 5 (20.0%)
Greatest absolute difference: 0.0234375 at index (2,) (up to 0.01 allowed)
Greatest relative difference: 0.0024318695068359375 at index (2,) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 5 (20.0%)
E   Greatest absolute difference: 0.046875 at index (2,) (up to 0.01 allowed)
E   Greatest relative difference: 0.002635955810546875 at index (2,) (up to 0.001 allowed)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 1 / 5 (20.0%)
E   Greatest absolute difference: 0.0234375 at index (2,) (up to 0.01 allowed)
E   Greatest relative difference: 0.0024318695068359375 at index (2,) (up to 0.001 allowed)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__index_put_bool_cpu_int32 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Tensor-likes are not equal!

Mismatched elements: 9 / 25 (36.0%)
Greatest absolute difference: 15 at index (1, 0)
Greatest relative difference: 6.0 at index (3, 4)
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not equal!
E   
E   Mismatched elements: 9 / 25 (36.0%)
E   Greatest absolute difference: 15 at index (1, 0)
E   Greatest relative difference: 6.0 at index (3, 4)

Check warning on line 0 in tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU

See this annotation in the file changed.

@github-actions github-actions / Test Results

6 out of 24 runs failed: test_output_match_opinfo__var_mean_correction_cpu_float16 (tests.function_libs.torch_lib.ops_test.TestOutputConsistencyEagerCPU)

artifacts/Test Results (py311-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py311-torch-nightly-windows-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-macos-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-ubuntu-latest)/pytest.xml [took 0s]
artifacts/Test Results (py312-torch-nightly-windows-latest)/pytest.xml [took 0s]
Raw output
AssertionError: Output 0 mismatch
AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 2 / 25 (8.0%)
E   Greatest absolute difference: 0.03125 at index (2, 3) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0017795562744140625 at index (2, 3) (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch
tests\function_libs\torch_lib\ops_test.py:252: in run_test_output_match
    torch.testing.assert_close(
E   AssertionError: Tensor-likes are not close!
E   
E   Mismatched elements: 2 / 25 (8.0%)
E   Greatest absolute difference: 0.0625 at index (2, 2) (up to 1e-05 allowed)
E   Greatest relative difference: 0.0014429092407226562 at index (2, 3) (up to 0.001 allowed)

The above exception was the direct cause of the following exception:
tests\function_libs\torch_lib\ops_test.py:266: in run_test_output_match
    raise AssertionError(f"Output {j} mismatch") from e
E   AssertionError: Output 0 mismatch