AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,10] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <int64 _val_3, float16[5,10] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <int64 _val_3, float16[1] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,10] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <float _val_3, float16[5,10] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <float _val_3, float16[1] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <int64 _val_3, float16 _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
   <float _val_3, float16 _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,10] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <int64 _val_3, float16[5,10] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <int64 _val_3, float16[1] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,10] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <float _val_3, float16[5,10] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <float _val_3, float16[1] _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <int64 _val_3, float16 _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, int64 _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,5] input_1, float16[5,5,10] input_2) => (float16[5,10] _val_10) 
E      <float _val_3, float16 _val_4, int64[1] _val_5, float16[5,5,10] _val_6, float16[5,10] _val_7, float _val_8, float16[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <int64 _val_3, int32[5] _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <float _val_3, int32[5] _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <int64 _val_3, int32[1] _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <float _val_3, int32[1] _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <int64 _val_3, int32 _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
   <float _val_3, int32 _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <int64 _val_3, int32[5] _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <float _val_3, int32[5] _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <int64 _val_3, int32[1] _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <float _val_3, int32[1] _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <int64 _val_3, int32 _val_4, int32[5] _val_5, int64 _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,10] input_1, int32[10] input_2) => (int32[5] _val_8) 
E      <float _val_3, int32 _val_4, int32[5] _val_5, float _val_6, int32[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: 
Not equal to tolerance rtol=1e-07, atol=0

Mismatched elements: 1 / 16 (6.25%)
Max absolute difference: 2.3841858e-07
Max relative difference: 1.5581162e-07
 x: array([[[1.608876, 1.258398, 1.868434, 1.530172, 1.502594, 1.577003,
         0.930287, 1.438904],
        [2.2129  , 1.367099, 2.42691 , 2.158696, 1.992608, 2.096078,
         1.297177, 2.084625]]], dtype=float32)
 y: array([[[1.608876, 1.258398, 1.868434, 1.530172, 1.502594, 1.577003,
         0.930287, 1.438904],
        [2.2129  , 1.367099, 2.42691 , 2.158696, 1.992608, 2.096078,
         1.297177, 2.084625]]], dtype=float32)
tests/functions/gemmgelu_test.py:62: in test_gemmgelu
    self.run_converter_test(case)
tests/common/onnx_script_test_case.py:255: in run_converter_test
    np.testing.assert_allclose(actual, param.output, rtol=self.rtol)
/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/contextlib.py:81: in inner
    return func(*args, **kwds)
E   AssertionError: 
E   Not equal to tolerance rtol=1e-07, atol=0
E   
E   Mismatched elements: 1 / 16 (6.25%)
E   Max absolute difference: 2.3841858e-07
E   Max relative difference: 1.5581162e-07
E    x: array([[[1.608876, 1.258398, 1.868434, 1.530172, 1.502594, 1.577003,
E            0.930287, 1.438904],
E           [2.2129  , 1.367099, 2.42691 , 2.158696, 1.992608, 2.096078,
E            1.297177, 2.084625]]], dtype=float32)
E    y: array([[[1.608876, 1.258398, 1.868434, 1.530172, 1.502594, 1.577003,
E            0.930287, 1.438904],
E           [2.2129  , 1.367099, 2.42691 , 2.158696, 1.992608, 2.096078,
E            1.297177, 2.084625]]], dtype=float32)

RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([ 7.4700713, -2.1084523], dtype=float32),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[0] input_1) => (float[2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 8.267502 , -1.9719322],
       [ 1.8161162, -4.3816953],
       [ 5.285544 ,  7.9338865]], dtype=float32),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[0] input_1) => (float[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[2] input_1) => (float[0,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 8.127981 , -7.645214 ],
       [ 6.9482465,  1.4977723],
       [-2.9223409,  5.5615497]], dtype=float32),
 'input_1': array([0, 2])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[2] input_1) => (float[0,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[2] input_1) => (float[1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-2.8383586,  2.418153 ],
       [-2.440615 ,  3.7877178],
       [ 8.035399 ,  5.202536 ]], dtype=float32),
 'input_1': array([1, 1])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[2] input_1) => (float[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[2] input_1) => (float[2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 4.570515 , -5.4855456],
       [-8.909176 , -3.4772449],
       [-6.903206 ,  7.38485  ]], dtype=float32),
 'input_1': array([2, 3])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[2] input_1) => (float[6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[3] input_1) => (float[2,3,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[3] input_1) => (float[2,9,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[3] input_1) => (float[0,2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[3] input_1) => (float[0,6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[2] input_0, int64[4] input_1) => (float[2,1,1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[3,2] input_0, int64[4] input_1) => (float[2,1,3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_0' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([ 7.4700713, -2.1084523], dtype=float32),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[0] input_1) => (float[2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 8.267502 , -1.9719322],
E          [ 1.8161162, -4.3816953],
E          [ 5.285544 ,  7.9338865]], dtype=float32),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[0] input_1) => (float[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[2] input_1) => (float[0,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 8.127981 , -7.645214 ],
E          [ 6.9482465,  1.4977723],
E          [-2.9223409,  5.5615497]], dtype=float32),
E    'input_1': array([0, 2])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[2] input_1) => (float[0,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[2] input_1) => (float[1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-2.8383586,  2.418153 ],
E          [-2.440615 ,  3.7877178],
E          [ 8.035399 ,  5.202536 ]], dtype=float32),
E    'input_1': array([1, 1])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[2] input_1) => (float[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[2] input_1) => (float[2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 4.570515 , -5.4855456],
E          [-8.909176 , -3.4772449],
E          [-6.903206 ,  7.38485  ]], dtype=float32),
E    'input_1': array([2, 3])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[2] input_1) => (float[6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[3] input_1) => (float[2,3,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[3] input_1) => (float[2,9,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[3] input_1) => (float[0,2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[3] input_1) => (float[0,6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[2] input_0, int64[4] input_1) => (float[2,1,1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[3,2] input_0, int64[4] input_1) => (float[2,1,3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64 input_0, int64 input_1) => (bool _val_9) 
   <int64 _val_2, int64 _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5] input_0, int64 input_1) => (bool[5] _val_9) 
   <int64[5] _val_2, int64[5] _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,1] input_0, int64[5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[10,5] input_0, int64 input_1) => (bool[10,5] _val_9) 
   <int64[10,5] _val_2, int64[10,5] _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,10,5] input_0, int64[10,5] input_1) => (bool[5,10,5] _val_9) 
   <int64[5,10,5] _val_2, int64[5,10,5] _val_3, int64[10,5] _val_4, float _val_5, float[10,5] _val_6, float _val_7, float[10,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,10,5] input_0, int64[5,10,5] input_1) => (bool[5,10,5] _val_9) 
   <int64[5,10,5] _val_2, int64[5,10,5] _val_3, int64[5,10,5] _val_4, float _val_5, float[5,10,5] _val_6, float _val_7, float[5,10,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[10,1,5] input_0, int64[10,5] input_1) => (bool[10,10,5] _val_9) 
   <int64[10,10,5] _val_2, int64[10,10,5] _val_3, int64[10,5] _val_4, float _val_5, float[10,5] _val_6, float _val_7, float[10,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[10,1,5] input_0, int64[1,10,5] input_1) => (bool[10,10,5] _val_9) 
   <int64[10,10,5] _val_2, int64[10,10,5] _val_3, int64[1,10,5] _val_4, float _val_5, float[1,10,5] _val_6, float _val_7, float[1,10,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[0,1,3] input_0, int64[0,10,3] input_1) => (bool[0,10,3] _val_9) 
   <int64[0,10,3] _val_2, int64[0,10,3] _val_3, int64[0,10,3] _val_4, float _val_5, float[0,10,3] _val_6, float _val_7, float[0,10,3] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-05}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-08}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-07}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-07}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-07}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-07}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-07}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-07}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
   <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {1e-07}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-07}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64 input_0, int64 input_1) => (bool _val_9) 
E      <int64 _val_2, int64 _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5] input_0, int64 input_1) => (bool[5] _val_9) 
E      <int64[5] _val_2, int64[5] _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,1] input_0, int64[5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[10,5] input_0, int64 input_1) => (bool[10,5] _val_9) 
E      <int64[10,5] _val_2, int64[10,5] _val_3, int64 _val_4, float _val_5, float _val_6, float _val_7, float _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,10,5] input_0, int64[10,5] input_1) => (bool[5,10,5] _val_9) 
E      <int64[5,10,5] _val_2, int64[5,10,5] _val_3, int64[10,5] _val_4, float _val_5, float[10,5] _val_6, float _val_7, float[10,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,10,5] input_0, int64[5,10,5] input_1) => (bool[5,10,5] _val_9) 
E      <int64[5,10,5] _val_2, int64[5,10,5] _val_3, int64[5,10,5] _val_4, float _val_5, float[5,10,5] _val_6, float _val_7, float[5,10,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[10,1,5] input_0, int64[10,5] input_1) => (bool[10,10,5] _val_9) 
E      <int64[10,10,5] _val_2, int64[10,10,5] _val_3, int64[10,5] _val_4, float _val_5, float[10,5] _val_6, float _val_7, float[10,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[10,1,5] input_0, int64[1,10,5] input_1) => (bool[10,10,5] _val_9) 
E      <int64[10,10,5] _val_2, int64[10,10,5] _val_3, int64[1,10,5] _val_4, float _val_5, float[1,10,5] _val_6, float _val_7, float[1,10,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[0,1,3] input_0, int64[0,10,3] input_1) => (bool[0,10,3] _val_9) 
E      <int64[0,10,3] _val_2, int64[0,10,3] _val_3, int64[0,10,3] _val_4, float _val_5, float[0,10,3] _val_6, float _val_7, float[0,10,3] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-05}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-08}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-07}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-07}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-07}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-07}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-07}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-07}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,5] input_0, int64[5,5] input_1) => (bool[5,5] _val_9) 
E      <int64[5,5] _val_2, int64[5,5] _val_3, int64[5,5] _val_4, float _val_5, float[5,5] _val_6, float _val_7, float[5,5] _val_8>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {1e-07}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-07}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([4, 2], dtype=int32), 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[0] input_1) => (int32[2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 5, -7],
       [ 5,  6],
       [-3, -4]], dtype=int32),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[0] input_1) => (int32[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[2] input_1) => (int32[0,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-1,  7],
       [ 1, -6],
       [-6,  0]], dtype=int32),
 'input_1': array([0, 2])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[0,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[2] input_1) => (int32[1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-9,  1],
       [-9,  2],
       [ 6,  1]], dtype=int32),
 'input_1': array([1, 1])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[2] input_1) => (int32[2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 0, -5],
       [ 1,  1],
       [ 1, -9]], dtype=int32),
 'input_1': array([2, 3])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[3] input_1) => (int32[2,3,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[3] input_1) => (int32[2,9,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[3] input_1) => (int32[0,2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[3] input_1) => (int32[0,6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2] input_0, int64[4] input_1) => (int32[2,1,1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0, int64[4] input_1) => (int32[2,1,3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_0' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([4, 2], dtype=int32), 'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[0] input_1) => (int32[2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 5, -7],
E          [ 5,  6],
E          [-3, -4]], dtype=int32),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[0] input_1) => (int32[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[2] input_1) => (int32[0,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-1,  7],
E          [ 1, -6],
E          [-6,  0]], dtype=int32),
E    'input_1': array([0, 2])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[0,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[2] input_1) => (int32[1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-9,  1],
E          [-9,  2],
E          [ 6,  1]], dtype=int32),
E    'input_1': array([1, 1])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[2] input_1) => (int32[2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 0, -5],
E          [ 1,  1],
E          [ 1, -9]], dtype=int32),
E    'input_1': array([2, 3])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[2] input_1) => (int32[6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[3] input_1) => (int32[2,3,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[3] input_1) => (int32[2,9,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[3] input_1) => (int32[0,2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[3] input_1) => (int32[0,6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2] input_0, int64[4] input_1) => (int32[2,1,1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0, int64[4] input_1) => (int32[2,1,3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,10] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <int64 _val_3, int32[5,10] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <int64 _val_3, int32[1] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,10] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <float _val_3, int32[5,10] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <float _val_3, int32[1] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <int64 _val_3, int32 _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = int64 {1}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
   <float _val_3, int32 _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,10] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <int64 _val_3, int32[5,10] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <int64 _val_3, int32[1] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,10] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <float _val_3, int32[5,10] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <float _val_3, int32[1] _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <int64 _val_3, int32 _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, int64 _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = int64 {1}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,5,5] input_1, int32[5,5,10] input_2) => (int32[5,10] _val_10) 
E      <float _val_3, int32 _val_4, int64[1] _val_5, int32[5,5,10] _val_6, int32[5,10] _val_7, float _val_8, int32[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1) => (bool _val_12) 
   <float16 _val_2, float16 _val_3, float16 _val_4, float _val_5, float _val_6, float _val_7, float _val_8, bool _val_9, int8 _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
   <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.1}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {0.01}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
   <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
{
   _val_2 = Sub (input_0, input_1)
   _val_3 = Abs (_val_2)
   _val_4 = Abs (input_1)
   _val_5 = Constant <value: tensor = float {0.5}> ()
   _val_6 = Mul (_val_5, _val_4)
   _val_7 = Constant <value: tensor = float {1e-16}> ()
   _val_8 = Add (_val_7, _val_6)
   _val_9 = LessOrEqual (_val_3, _val_8)
   _val_10 = Cast <to: int = 3> (_val_9)
   _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
   _val_12 = Cast <to: int = 9> (_val_11)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/funct…ut) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5] input_0, float16[5] input_1) => (bool _val_12) 
E      <float16[5] _val_2, float16[5] _val_3, float16[5] _val_4, float _val_5, float[5] _val_6, float _val_7, float[5] _val_8, bool[5] _val_9, int8[5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.1}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {0.01}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_6): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5,5] input_0, float16[5,5,5] input_1) => (bool _val_12) 
E      <float16[5,5,5] _val_2, float16[5,5,5] _val_3, float16[5,5,5] _val_4, float _val_5, float[5,5,5] _val_6, float _val_7, float[5,5,5] _val_8, bool[5,5,5] _val_9, int8[5,5,5] _val_10, int8 _val_11>
E   {
E      _val_2 = Sub (input_0, input_1)
E      _val_3 = Abs (_val_2)
E      _val_4 = Abs (input_1)
E      _val_5 = Constant <value: tensor = float {0.5}> ()
E      _val_6 = Mul (_val_5, _val_4)
E      _val_7 = Constant <value: tensor = float {1e-16}> ()
E      _val_8 = Add (_val_7, _val_6)
E      _val_9 = LessOrEqual (_val_3, _val_8)
E      _val_10 = Cast <to: int = 3> (_val_9)
E      _val_11 = ReduceMin <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_10)
E      _val_12 = Cast <to: int = 9> (_val_11)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5] input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
   <float _val_3, int64[5] _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[1] input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
   <float _val_3, int64[1] _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64 input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
   <float _val_3, int64 _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
{
   _val_3 = Constant <value: tensor = float {0.2}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = float {0.6}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5] input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
E      <float _val_3, int64[5] _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[1] input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
E      <float _val_3, int64[1] _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64 input_0, int64[5,10] input_1, int64[10] input_2) => (int64[5] _val_8) 
E      <float _val_3, int64 _val_4, int64[5] _val_5, float _val_6, int64[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = float {0.2}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = float {0.6}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[5] input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
   <int64 _val_3, float[5] _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float[1] input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
   <int64 _val_3, float[1] _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
   <int64 _val_3, float _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
{
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = MatMul (input_1, input_2)
   _val_6 = Constant <value: tensor = int64 {1}> ()
   _val_7 = Mul (_val_5, _val_6)
   _val_8 = Add (_val_4, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[5] input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
E      <int64 _val_3, float[5] _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float[1] input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
E      <int64 _val_3, float[1] _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float input_0, float[5,10] input_1, float[10] input_2) => (float[5] _val_8) 
E      <int64 _val_3, float _val_4, float[5] _val_5, int64 _val_6, float[5] _val_7>
E   {
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = MatMul (input_1, input_2)
E      _val_6 = Constant <value: tensor = int64 {1}> ()
E      _val_7 = Mul (_val_5, _val_6)
E      _val_8 = Add (_val_4, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([4, 2]), 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[0] input_1) => (int64[2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 5, -7],
       [ 5,  6],
       [-3, -4]]),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[0] input_1) => (int64[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[2] input_1) => (int64[0,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-1,  7],
       [ 1, -6],
       [-6,  0]]),
 'input_1': array([0, 2])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[0,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[2] input_1) => (int64[1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-9,  1],
       [-9,  2],
       [ 6,  1]]),
 'input_1': array([1, 1])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[2] input_1) => (int64[2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ 0, -5],
       [ 1,  1],
       [ 1, -9]]),
 'input_1': array([2, 3])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[3] input_1) => (int64[2,3,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[3] input_1) => (int64[2,9,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[3] input_1) => (int64[0,2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[3] input_1) => (int64[0,6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[2] input_0, int64[4] input_1) => (int64[2,1,1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[3,2] input_0, int64[4] input_1) => (int64[2,1,3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_0' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([4, 2]), 'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[0] input_1) => (int64[2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 5, -7],
E          [ 5,  6],
E          [-3, -4]]),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[0] input_1) => (int64[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[2] input_1) => (int64[0,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-1,  7],
E          [ 1, -6],
E          [-6,  0]]),
E    'input_1': array([0, 2])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[0,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[2] input_1) => (int64[1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-9,  1],
E          [-9,  2],
E          [ 6,  1]]),
E    'input_1': array([1, 1])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[2] input_1) => (int64[2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ 0, -5],
E          [ 1,  1],
E          [ 1, -9]]),
E    'input_1': array([2, 3])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[2] input_1) => (int64[6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[3] input_1) => (int64[2,3,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[3] input_1) => (int64[2,9,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[3] input_1) => (int64[0,2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[3] input_1) => (int64[0,6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[2] input_0, int64[4] input_1) => (int64[2,1,1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[3,2] input_0, int64[4] input_1) => (int64[2,1,3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([ True, False]), 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[0] input_1) => (bool[2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[False, False],
       [ True, False],
       [False, False]]),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[0] input_1) => (bool[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[2] input_1) => (bool[0,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[False, False],
       [ True,  True],
       [ True, False]]),
 'input_1': array([0, 2])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[0,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[2] input_1) => (bool[1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[False, False],
       [False,  True],
       [ True,  True]]),
 'input_1': array([1, 1])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[2] input_1) => (bool[2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[ True,  True],
       [False,  True],
       [False,  True]]),
 'input_1': array([2, 3])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[3] input_1) => (bool[2,3,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[3] input_1) => (bool[2,9,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[3] input_1) => (bool[0,2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[3] input_1) => (bool[0,6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[2] input_0, int64[4] input_1) => (bool[2,1,1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[3,2] input_0, int64[4] input_1) => (bool[2,1,3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_0' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([ True, False]), 'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[0] input_1) => (bool[2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[False, False],
E          [ True, False],
E          [False, False]]),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[0] input_1) => (bool[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[2] input_1) => (bool[0,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[False, False],
E          [ True,  True],
E          [ True, False]]),
E    'input_1': array([0, 2])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[0,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[2] input_1) => (bool[1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[False, False],
E          [False,  True],
E          [ True,  True]]),
E    'input_1': array([1, 1])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[2] input_1) => (bool[2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[ True,  True],
E          [False,  True],
E          [False,  True]]),
E    'input_1': array([2, 3])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[2] input_1) => (bool[6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[3] input_1) => (bool[2,3,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[3] input_1) => (bool[2,9,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[3] input_1) => (bool[0,2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[3] input_1) => (bool[0,6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[2] input_0, int64[4] input_1) => (bool[2,1,1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[3,2] input_0, int64[4] input_1) => (bool[2,1,3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2,3] input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
   _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
   _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
   _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.checker.ValidationError: Mismatched attribute type in 'Gemm_0 : alpha'. Expected: 'FLOAT', actual: 'INT'
E   
E   ==> Context: Bad node spec for node. Name: Gemm_0 OpType: Gemm

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2,3] input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
E      _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.checker.ValidationError: Mismatched attribute type in 'Gemm_0 : alpha'. Expected: 'FLOAT', actual: 'INT'
E   
E   ==> Context: Bad node spec for node. Name: Gemm_0 OpType: Gemm

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
E      _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.checker.ValidationError: Mismatched attribute type in 'Gemm_0 : alpha'. Expected: 'FLOAT', actual: 'INT'
E   
E   ==> Context: Bad node spec for node. Name: Gemm_0 OpType: Gemm

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[2,2] input_1, float16[2,3] input_2) => (float16[2,3] _val_3) {
E      _val_3 = Gemm <alpha: int = 1, beta: int = 1, transA: int = 0, transB: int = 0> (input_1, input_2, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[20,20] input_0) => (float16[20,20] _val_3) 
   <float _val_1, float16[20,20] _val_2>
{
   _val_1 = Constant <value: tensor = float {1}> ()
   _val_2 = Add (input_0, _val_1)
   _val_3 = Log (_val_2)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Add, node name: Add_2): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[20,20] input_0) => (float16[20,20] _val_3) 
E      <float _val_1, float16[20,20] _val_2>
E   {
E      _val_1 = Constant <value: tensor = float {1}> ()
E      _val_2 = Add (input_0, _val_1)
E      _val_3 = Log (_val_2)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3,3] input_0, int64[2] input_1) => (float16[1,3,7] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {0}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3] input_0, int64[2] input_1) => (float16[1,6] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {2}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3] input_0, int64[2] input_1) => (float16[1,4] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {2}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3] input_0, int64[4] input_1) => (float16[2,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {2}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3] input_0, int64[2] input_1) => (float16[0,3,6] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3] input_0, int64[2] input_1) => (float16[0,3,4] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3] input_0, int64[4] input_1) => (float16[0,4,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3] input_0, int64[6] input_1) => (float16[2,5,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3,3] input_0, int64[2] input_1) => (float16[1,3,6] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3,3] input_0, int64[2] input_1) => (float16[1,3,4] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3,3] input_0, int64[4] input_1) => (float16[1,4,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1,3,3] input_0, int64[6] input_1) => (float16[3,5,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3,3] input_0, int64[2] input_1) => (float16[0,3,3,6] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {4}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3,3] input_0, int64[2] input_1) => (float16[0,3,3,4] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {4}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3,3] input_0, int64[4] input_1) => (float16[0,3,4,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {4}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[0,3,3,3] input_0, int64[6] input_1) => (float16[0,5,5,5] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {4}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,3,5,5] input_0, int64[2] input_1) => (float16[3,3,5,8] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {4}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {1}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onn…E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[0,3,3,3] input_0, int64[4] input_1) => (float16[0,3,4,5] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[0,3,3,3] input_0, int64[6] input_1) => (float16[0,5,5,5] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,3,5,5] input_0, int64[2] input_1) => (float16[3,3,5,8] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,3,5,5] input_0, int64[2] input_1) => (float16[3,3,5,6] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,3,5,5] input_0, int64[4] input_1) => (float16[3,3,6,7] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,3,5,5] input_0, int64[6] input_1) => (float16[3,5,7,7] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1,3,3,3,3] input_0, int64[2] input_1) => (float16[1,3,3,3,6] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {5}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1,3,3,3,3] input_0, int64[2] input_1) => (float16[1,3,3,3,4] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {5}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1,3,3,3,3] input_0, int64[4] input_1) => (float16[1,3,3,4,5] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {5}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1,3,3,3,3] input_0, int64[6] input_1) => (float16[1,3,5,5,5] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {5}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1,3,4,4] input_0, int64[4] input_1) => (float16[1,3,3,4] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {4}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {2}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32 input_1) => (int32 _val_14) 
   <int64 _val_2, bool _val_3, uint32 _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32 _val_10, uint32 _val_11, int32 _val_12, int32 _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5] input_0, int32 input_1) => (int32[5] _val_14) 
   <int64 _val_2, bool[5] _val_3, uint32[5] _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32[5] _val_10, uint32[5] _val_11, int32[5] _val_12, int32[5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,1] input_0, int32[5] input_1) => (int32[5,5] _val_14) 
   <int64 _val_2, bool[5,1] _val_3, uint32[5,1] _val_4, uint32[5] _val_5, int64 _val_6, uint32 _val_7, uint32[5] _val_8, uint32[5] _val_9, uint32[5,5] _val_10, uint32[5,5] _val_11, int32[5,5] _val_12, int32[5,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[10,5] input_0, int32 input_1) => (int32[10,5] _val_14) 
   <int64 _val_2, bool[10,5] _val_3, uint32[10,5] _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32[10,5] _val_10, uint32[10,5] _val_11, int32[10,5] _val_12, int32[10,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,10,5] input_0, int32[10,5] input_1) => (int32[5,10,5] _val_14) 
   <int64 _val_2, bool[5,10,5] _val_3, uint32[5,10,5] _val_4, uint32[10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[10,5] _val_8, uint32[10,5] _val_9, uint32[5,10,5] _val_10, uint32[5,10,5] _val_11, int32[5,10,5] _val_12, int32[5,10,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,10,5] input_0, int32[5,10,5] input_1) => (int32[5,10,5] _val_14) 
   <int64 _val_2, bool[5,10,5] _val_3, uint32[5,10,5] _val_4, uint32[5,10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[5,10,5] _val_8, uint32[5,10,5] _val_9, uint32[5,10,5] _val_10, uint32[5,10,5] _val_11, int32[5,10,5] _val_12, int32[5,10,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[10,1,5] input_0, int32[10,5] input_1) => (int32[10,10,5] _val_14) 
   <int64 _val_2, bool[10,1,5] _val_3, uint32[10,1,5] _val_4, uint32[10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[10,5] _val_8, uint32[10,5] _val_9, uint32[10,10,5] _val_10, uint32[10,10,5] _val_11, int32[10,10,5] _val_12, int32[10,10,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[10,1,5] input_0, int32[1,10,5] input_1) => (int32[10,10,5] _val_14) 
   <int64 _val_2, bool[10,1,5] _val_3, uint32[10,1,5] _val_4, uint32[1,10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[1,10,5] _val_8, uint32[1,10,5] _val_9, uint32[10,10,5] _val_10, uint32[10,10,5] _val_11, int32[10,10,5] _val_12, int32[10,10,5] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[0,1,3] input_0, int32[0,10,3] input_1) => (int32[0,10,3] _val_14) 
   <int64 _val_2, bool[0,1,3] _val_3, uint32[0,1,3] _val_4, uint32[0,10,3] _val_5, int64 _val_6, uint32 _val_7, uint32[0,10,3] _val_8, uint32[0,10,3] _val_9, uint32[0,10,3] _val_10, uint32[0,10,3] _val_11, int32[0,10,3] _val_12, int32[0,10,3] _val_13>
{
   _val_2 = Constant <value: tensor = int64 {0}> ()
   _val_3 = Less (input_0, _val_2)
   _val_4 = Cast <to: int = 12> (input_0)
   _val_5 = Cast <to: int = 12> (input_1)
   _val_6 = Constant <value_int: int = 4294967295> ()
   _val_7 = Cast <to: int = 12> (_val_6)
   _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
   _val_9 = BitwiseNot (_val_8)
   _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
   _val_11 = BitwiseOr (_val_10, _val_9)
   _val_12 = Cast <to: int = 6> (_val_11)
   _val_13 = Cast <to: int = 6> (_val_10)
   _val_14 = Where (_val_3, _val_12, _val_13)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32 input_1) => (int32 _val_14) 
E      <int64 _val_2, bool _val_3, uint32 _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32 _val_10, uint32 _val_11, int32 _val_12, int32 _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5] input_0, int32 input_1) => (int32[5] _val_14) 
E      <int64 _val_2, bool[5] _val_3, uint32[5] _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32[5] _val_10, uint32[5] _val_11, int32[5] _val_12, int32[5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,1] input_0, int32[5] input_1) => (int32[5,5] _val_14) 
E      <int64 _val_2, bool[5,1] _val_3, uint32[5,1] _val_4, uint32[5] _val_5, int64 _val_6, uint32 _val_7, uint32[5] _val_8, uint32[5] _val_9, uint32[5,5] _val_10, uint32[5,5] _val_11, int32[5,5] _val_12, int32[5,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[10,5] input_0, int32 input_1) => (int32[10,5] _val_14) 
E      <int64 _val_2, bool[10,5] _val_3, uint32[10,5] _val_4, uint32 _val_5, int64 _val_6, uint32 _val_7, uint32 _val_8, uint32 _val_9, uint32[10,5] _val_10, uint32[10,5] _val_11, int32[10,5] _val_12, int32[10,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,10,5] input_0, int32[10,5] input_1) => (int32[5,10,5] _val_14) 
E      <int64 _val_2, bool[5,10,5] _val_3, uint32[5,10,5] _val_4, uint32[10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[10,5] _val_8, uint32[10,5] _val_9, uint32[5,10,5] _val_10, uint32[5,10,5] _val_11, int32[5,10,5] _val_12, int32[5,10,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,10,5] input_0, int32[5,10,5] input_1) => (int32[5,10,5] _val_14) 
E      <int64 _val_2, bool[5,10,5] _val_3, uint32[5,10,5] _val_4, uint32[5,10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[5,10,5] _val_8, uint32[5,10,5] _val_9, uint32[5,10,5] _val_10, uint32[5,10,5] _val_11, int32[5,10,5] _val_12, int32[5,10,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[10,1,5] input_0, int32[10,5] input_1) => (int32[10,10,5] _val_14) 
E      <int64 _val_2, bool[10,1,5] _val_3, uint32[10,1,5] _val_4, uint32[10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[10,5] _val_8, uint32[10,5] _val_9, uint32[10,10,5] _val_10, uint32[10,10,5] _val_11, int32[10,10,5] _val_12, int32[10,10,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[10,1,5] input_0, int32[1,10,5] input_1) => (int32[10,10,5] _val_14) 
E      <int64 _val_2, bool[10,1,5] _val_3, uint32[10,1,5] _val_4, uint32[1,10,5] _val_5, int64 _val_6, uint32 _val_7, uint32[1,10,5] _val_8, uint32[1,10,5] _val_9, uint32[10,10,5] _val_10, uint32[10,10,5] _val_11, int32[10,10,5] _val_12, int32[10,10,5] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_2): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[0,1,3] input_0, int32[0,10,3] input_1) => (int32[0,10,3] _val_14) 
E      <int64 _val_2, bool[0,1,3] _val_3, uint32[0,1,3] _val_4, uint32[0,10,3] _val_5, int64 _val_6, uint32 _val_7, uint32[0,10,3] _val_8, uint32[0,10,3] _val_9, uint32[0,10,3] _val_10, uint32[0,10,3] _val_11, int32[0,10,3] _val_12, int32[0,10,3] _val_13>
E   {
E      _val_2 = Constant <value: tensor = int64 {0}> ()
E      _val_3 = Less (input_0, _val_2)
E      _val_4 = Cast <to: int = 12> (input_0)
E      _val_5 = Cast <to: int = 12> (input_1)
E      _val_6 = Constant <value_int: int = 4294967295> ()
E      _val_7 = Cast <to: int = 12> (_val_6)
E      _val_8 = BitShift <direction: string = "RIGHT"> (_val_7, _val_5)
E      _val_9 = BitwiseNot (_val_8)
E      _val_10 = BitShift <direction: string = "RIGHT"> (_val_4, _val_5)
E      _val_11 = BitwiseOr (_val_10, _val_9)
E      _val_12 = Cast <to: int = 6> (_val_11)
E      _val_13 = Cast <to: int = 6> (_val_10)
E      _val_14 = Where (_val_3, _val_12, _val_13)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[5,10] input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
   <float _val_3, int64[5,10] _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[1] input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
   <float _val_3, int64[1] _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64 input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
   <float _val_3, int64 _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
{
   _val_3 = Constant <value: tensor = float {0.6}> ()
   _val_4 = Mul (input_0, _val_3)
   _val_5 = Constant <value_ints: ints = [0]> ()
   _val_6 = MatMul (input_1, input_2)
   _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
   _val_8 = Constant <value: tensor = float {0.2}> ()
   _val_9 = Mul (_val_7, _val_8)
   _val_10 = Add (_val_4, _val_9)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[5,10] input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
E      <float _val_3, int64[5,10] _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[1] input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
E      <float _val_3, int64[1] _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64 input_0, int64[5,5,5] input_1, int64[5,5,10] input_2) => (int64[5,10] _val_10) 
E      <float _val_3, int64 _val_4, int64[1] _val_5, int64[5,5,10] _val_6, int64[5,10] _val_7, float _val_8, int64[5,10] _val_9>
E   {
E      _val_3 = Constant <value: tensor = float {0.6}> ()
E      _val_4 = Mul (input_0, _val_3)
E      _val_5 = Constant <value_ints: ints = [0]> ()
E      _val_6 = MatMul (input_1, input_2)
E      _val_7 = ReduceSum <keepdims: int = 0, noop_with_empty_axes: int = 0> (_val_6, _val_5)
E      _val_8 = Constant <value: tensor = float {0.2}> ()
E      _val_9 = Mul (_val_7, _val_8)
E      _val_10 = Add (_val_4, _val_9)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3] input_0) => (int32[3] _val_5) 
   <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0) => (int32 _val_5) 
   <float _val_1, float _val_2, int64 _val_3, bool _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
   <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {0}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
   <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
{
   _val_1 = Cast <to: int = 1> (input_0)
   _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Less (_val_2, _val_3)
   _val_5 = CastLike (_val_4, input_0)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3] input_0) => (int32[3] _val_5) 
E      <float[3] _val_1, float[3] _val_2, int64 _val_3, bool[3] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0) => (int32 _val_5) 
E      <float _val_1, float _val_2, int64 _val_3, bool _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[3,2] input_0) => (int32[3,2] _val_5) 
E      <float[3,2] _val_1, float[3,2] _val_2, int64 _val_3, bool[3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {0}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Less, node name: Less_4): B has inconsistent type tensor(int64)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[2,3,2] input_0) => (int32[2,3,2] _val_5) 
E      <float[2,3,2] _val_1, float[2,3,2] _val_2, int64 _val_3, bool[2,3,2] _val_4>
E   {
E      _val_1 = Cast <to: int = 1> (input_0)
E      _val_2 = RandomUniformLike <high: float = 1, low: float = 0> (_val_1)
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Less (_val_2, _val_3)
E      _val_5 = CastLike (_val_4, input_0)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32[5,5] input_1, int32[5,5] input_2) => (int32[5,5] _val_6) 
   <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32[5,5] input_1, int32[5,5] input_2) => (int32[5,5] _val_6) 
   <int64 _val_3, int64[5,5] _val_4, int64[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32[5,1] input_1, int32[1,5] input_2) => (int32[5,5] _val_6) 
   <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32[5,1] input_1, int32[1,5] input_2) => (int32[5,5] _val_6) 
   <int64 _val_3, int64[5,1] _val_4, int64[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[1] input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
   <int64 _val_3, int64[5,5,1] _val_4, int64[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32 input_1, int32 input_2) => (int32 _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32 input_1, int32 input_2) => (int32 _val_6) 
   <int64 _val_3, int64 _val_4, int64 _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32 input_1, int32 input_2) => (int32[5,5] _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32[5,5] input_0, int32 input_1, int32 input_2) => (int32[5,5] _val_6) 
   <int64 _val_3, int64 _val_4, int64 _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int32 input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
   <int64 _val_3, int64[5,5,1] _val_4, int64[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32[5,5] input_1, int32[5,5] input_2) => (int32[5,5] _val_6) 
E      <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32[5,5] input_1, int32[5,5] input_2) => (int32[5,5] _val_6) 
E      <int64 _val_3, int64[5,5] _val_4, int64[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32[5,1] input_1, int32[1,5] input_2) => (int32[5,5] _val_6) 
E      <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32[5,1] input_1, int32[1,5] input_2) => (int32[5,5] _val_6) 
E      <int64 _val_3, int64[5,1] _val_4, int64[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[1] input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
E      <int64 _val_3, int64[5,5,1] _val_4, int64[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32 input_1, int32 input_2) => (int32 _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32 input_1, int32 input_2) => (int32 _val_6) 
E      <int64 _val_3, int64 _val_4, int64 _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32 input_1, int32 input_2) => (int32[5,5] _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32[5,5] input_0, int32 input_1, int32 input_2) => (int32[5,5] _val_6) 
E      <int64 _val_3, int64 _val_4, int64 _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(int32)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int32 input_0, int32[5,5,1] input_1, int32[1,5] input_2) => (int32[5,5,5] _val_6) 
E      <int64 _val_3, int64[5,5,1] _val_4, int64[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([ 3.824, -1.441], dtype=float16),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[0] input_1) => (float16[2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[2.373 , 0.9316],
       [0.624 , 6.152 ],
       [0.1758, 5.4   ]], dtype=float16),
 'input_1': array([], dtype=int64)}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[0] input_1) => (float16[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[2] input_1) => (float16[0,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-6.574,  1.406],
       [-4.965,  2.75 ],
       [-8.81 , -6.785]], dtype=float16),
 'input_1': array([0, 2])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[0,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[2] input_1) => (float16[1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-4.836,  0.51 ],
       [-4.516,  5.984],
       [ 8.99 ,  4.176]], dtype=float16),
 'input_1': array([1, 1])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[2] input_1) => (float16[2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
RuntimeError: ONNX Runtime failed to evaluate:
Inputs:
{'input_0': array([[-7.34 , -0.589],
       [-2.97 , -0.589],
       [-4.062,  7.69 ]], dtype=float16),
 'input_1': array([2, 3])}
Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[3] input_1) => (float16[2,3,4] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[3] input_1) => (float16[2,9,4] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[3] input_1) => (float16[0,2,6] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[3] input_1) => (float16[0,6,6] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[2] input_0, int64[4] input_1) => (float16[2,1,1,2] _val_2) {
   _val_2 = Tile (input_0, input_1)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[3,2] input_0, int64[4] input_1) => (float16[2,1,3,2] _val_8) 
   <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
{
   _val_2 = Constant <value_ints: ints = [1]> ()
   _val_3 = Constant <value: tensor = int64 {1}> ()
   _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
   _val_5 = Constant <value_ints: ints = [1]> ()
   _val_6 = Expand (_val_5, _val_4)
   _val_7 = Concat <axis: int = 0> (_val_6, input_1)
   _val_8 = Tile (input_0, _val_7)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_0' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([ 3.824, -1.441], dtype=float16),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[0] input_1) => (float16[2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[2.373 , 0.9316],
E          [0.624 , 6.152 ],
E          [0.1758, 5.4   ]], dtype=float16),
E    'input_1': array([], dtype=int64)}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[0] input_1) => (float16[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[2] input_1) => (float16[0,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-6.574,  1.406],
E          [-4.965,  2.75 ],
E          [-8.81 , -6.785]], dtype=float16),
E    'input_1': array([0, 2])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[0,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[2] input_1) => (float16[1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-4.836,  0.51 ],
E          [-4.516,  5.984],
E          [ 8.99 ,  4.176]], dtype=float16),
E    'input_1': array([1, 1])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (2)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[2] input_1) => (float16[2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:550: in _capture_graph_and_evaluate_torch_script_evaluator
    return _ort_session_run(onnx_model.SerializeToString(), ort_inputs)
tests/function_libs/torch_lib/ops_test_common.py:325: in _ort_session_run
    return session.run(None, ort_inputs)
.nox/test/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py:220: in run
    return self._sess.run(output_names, input_feed, run_options)
E   onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Non-zero status code returned while running Tile node. Name:'Tile_7' Status Message: 'repeat' input tensor must have the same length as the 'input' tensor

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:564: in _capture_graph_and_evaluate_torch_script_evaluator
    raise RuntimeError(
E   RuntimeError: ONNX Runtime failed to evaluate:
E   Inputs:
E   {'input_0': array([[-7.34 , -0.589],
E          [-2.97 , -0.589],
E          [-4.062,  7.69 ]], dtype=float16),
E    'input_1': array([2, 3])}
E   Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[2] input_1) => (float16[6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[3] input_1) => (float16[2,3,4] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[3] input_1) => (float16[2,9,4] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[3] input_1) => (float16[0,2,6] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (3)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[3] input_1) => (float16[0,6,6] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_0): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (1) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[2] input_0, int64[4] input_1) => (float16[2,1,1,2] _val_2) {
E      _val_2 = Tile (input_0, input_1)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] Inference error(s): (op_type:Tile, node name: Tile_7): [ShapeInferenceError] Inferred shape and existing shape differ in rank: (2) vs (4)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[3,2] input_0, int64[4] input_1) => (float16[2,1,3,2] _val_8) 
E      <int64[1] _val_2, int64 _val_3, int64[1] _val_4, int64[1] _val_5, int64[unk__0] _val_6, int64[unk__1] _val_7>
E   {
E      _val_2 = Constant <value_ints: ints = [1]> ()
E      _val_3 = Constant <value: tensor = int64 {1}> ()
E      _val_4 = Reshape <allowzero: int = 0> (_val_3, _val_2)
E      _val_5 = Constant <value_ints: ints = [1]> ()
E      _val_6 = Expand (_val_5, _val_4)
E      _val_7 = Concat <axis: int = 0> (_val_6, input_1)
E      _val_8 = Tile (input_0, _val_7)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
   <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
   <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
   <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
   <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
   <float16 _val_3, float _val_4, float16 _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
   <float16 _val_3, float _val_4, float16 _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
   <float16 _val_3, float _val_4, float16 _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
   <float16 _val_3, float _val_4, float16 _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {1}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
{
   _val_3 = Div (input_1, input_2)
   _val_4 = Constant <value: tensor = float {3.14}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
E      <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
E      <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
E      <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
E      <float16[5,5] _val_3, float _val_4, float16[5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
E      <float16 _val_3, float _val_4, float16 _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
E      <float16 _val_3, float _val_4, float16 _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
E      <float16 _val_3, float _val_4, float16 _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
E      <float16 _val_3, float _val_4, float16 _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {1}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_3): B has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float16[5,5,5] _val_3, float _val_4, float16[5,5,5] _val_5>
E   {
E      _val_3 = Div (input_1, input_2)
E      _val_4 = Constant <value: tensor = float {3.14}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (int64[1,3,3] input_0, int64[2] input_1) => (int64[1,3,7] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {0}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (int64[1,3,3] input_0, int64[2] input_1) => (int64[1,3,7] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {0}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
   <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
   <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
   <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
   <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
   <float _val_3, float _val_4, float _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {1}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
   <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
{
   _val_3 = Constant <value: tensor = float {3.14}> ()
   _val_4 = Mul (_val_3, input_1)
   _val_5 = Mul (_val_4, input_2)
   _val_6 = Add (input_0, _val_5)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,5] input_1, float16[5,5] input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float[5,5] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16[5,1] input_1, float16[1,5] input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float[5,1] _val_4, float[5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[1] input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16 input_1, float16 input_2) => (float16 _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16[5,5] input_0, float16 input_1, float16 input_2) => (float16[5,5] _val_6) 
E      <float _val_3, float _val_4, float _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {1}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Mul, node name: Mul_2): B has inconsistent type tensor(float16)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (float16 input_0, float16[5,5,1] input_1, float16[1,5] input_2) => (float16[5,5,5] _val_6) 
E      <float _val_3, float[5,5,1] _val_4, float[5,5,5] _val_5>
E   {
E      _val_3 = Constant <value: tensor = float {3.14}> ()
E      _val_4 = Mul (_val_3, input_1)
E      _val_5 = Mul (_val_4, input_2)
E      _val_6 = Add (input_0, _val_5)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }

AssertionError: ONNX model is invalid. Model:
<
   ir_version: 8,
   opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
   producer_name: "pytorch",
   producer_version: "2.2.2"
>
main_graph (bool[1,3,3] input_0, int64[2] input_1) => (bool[1,3,7] _val_20) 
   <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
{
   _val_2 = Constant <value_ints: ints = [-1]> ()
   _val_3 = Constant <value: tensor = int64 {3}> ()
   _val_4 = Constant <value: tensor = int64 {2}> ()
   _val_5 = Mul (_val_3, _val_4)
   _val_6 = Size (input_1)
   _val_7 = Sub (_val_5, _val_6)
   _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
   _val_9 = Constant <value_ints: ints = [0]> ()
   _val_10 = Expand (_val_9, _val_8)
   _val_11 = Concat <axis: int = 0> (input_1, _val_10)
   _val_12 = Size (_val_11)
   _val_13 = Constant <value_ints: ints = [-2]> ()
   _val_14 = Sub (_val_13, _val_12)
   _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
   _val_16 = Sub (_val_2, _val_12)
   _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
   _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
   _val_19 = Constant <value: tensor = float {0}> ()
   _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
Rank (input) => (return_val)
{
   tmp = Shape (input)
   return_val = Size (tmp)
}
<
  domain: "pkg.onnxscript.torch_lib.common",
  opset_import: ["" : 18]
>
IsScalar (input) => (return_val)
{
   tmp = Shape (input)
   tmp_0 = Size (tmp)
   tmp_1 = Constant <value_int: int = 0> ()
   return_val = Equal (tmp_0, tmp_1)
}
tests/function_libs/torch_lib/ops_test_common.py:536: in _capture_graph_and_evaluate_torch_script_evaluator
    onnx.checker.check_model(onnx_model, full_check=True)
.nox/test/lib/python3.9/site-packages/onnx/checker.py:179: in check_model
    C.check_model(
E   onnx.onnx_cpp2py_export.shape_inference.InferenceError: [ShapeInferenceError] (op_type:Pad, node name: Pad_21): constant_value has inconsistent type tensor(float)

The above exception was the direct cause of the following exception:
tests/function_libs/torch_lib/ops_test.py:242: in run_test_output_match
    function_output = function_executor(test_name, reference_torch_outputs)(
tests/function_libs/torch_lib/ops_test_common.py:538: in _capture_graph_and_evaluate_torch_script_evaluator
    raise AssertionError(
E   AssertionError: ONNX model is invalid. Model:
E   <
E      ir_version: 8,
E      opset_import: ["" : 18, "pkg.onnxscript.torch_lib.common" : 1],
E      producer_name: "pytorch",
E      producer_version: "2.2.2"
E   >
E   main_graph (bool[1,3,3] input_0, int64[2] input_1) => (bool[1,3,7] _val_20) 
E      <int64[1] _val_2, int64 _val_3, int64 _val_4, int64 _val_5, int64 _val_6, int64 _val_7, int64[1] _val_8, int64[1] _val_9, int64[unk__0] _val_10, int64[unk__1] _val_11, int64 _val_12, int64[1] _val_13, int64[1] _val_14, int64[unk__2] _val_15, int64[1] _val_16, int64[unk__3] _val_17, int64[unk__4] _val_18, float _val_19>
E   {
E      _val_2 = Constant <value_ints: ints = [-1]> ()
E      _val_3 = Constant <value: tensor = int64 {3}> ()
E      _val_4 = Constant <value: tensor = int64 {2}> ()
E      _val_5 = Mul (_val_3, _val_4)
E      _val_6 = Size (input_1)
E      _val_7 = Sub (_val_5, _val_6)
E      _val_8 = Reshape <allowzero: int = 0> (_val_7, _val_2)
E      _val_9 = Constant <value_ints: ints = [0]> ()
E      _val_10 = Expand (_val_9, _val_8)
E      _val_11 = Concat <axis: int = 0> (input_1, _val_10)
E      _val_12 = Size (_val_11)
E      _val_13 = Constant <value_ints: ints = [-2]> ()
E      _val_14 = Sub (_val_13, _val_12)
E      _val_15 = Slice (_val_11, _val_13, _val_14, _val_9, _val_13)
E      _val_16 = Sub (_val_2, _val_12)
E      _val_17 = Slice (_val_11, _val_2, _val_16, _val_9, _val_13)
E      _val_18 = Concat <axis: int = 0> (_val_15, _val_17)
E      _val_19 = Constant <value: tensor = float {0}> ()
E      _val_20 = Pad <mode: string = "constant"> (input_0, _val_18, _val_19)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   Rank (input) => (return_val)
E   {
E      tmp = Shape (input)
E      return_val = Size (tmp)
E   }
E   <
E     domain: "pkg.onnxscript.torch_lib.common",
E     opset_import: ["" : 18]
E   >
E   IsScalar (input) => (return_val)
E   {
E      tmp = Shape (input)
E      tmp_0 = Size (tmp)
E      tmp_1 = Constant <value_int: int = 0> ()
E      return_val = Equal (tmp_0, tmp_1)
E   }