bitwise_and cannot have int argument

[wodesuck]

import torch
import torch.nn as nn


class Model(nn.Module):
    def forward(self, x):
        return x & 1


def main():
    model = Model()
    args = (torch.tensor([1, 2, 3]),)
    torch.onnx.export(
        model,
        args,
        "model_test.onnx",
        dynamo=True,
    )


if __name__ == '__main__':
    main()

[torch.onnx] Obtain model graph for `Model()` with `torch.export.export(..., strict=False)`...
[torch.onnx] Obtain model graph for `Model()` with `torch.export.export(..., strict=False)`... ✅
[torch.onnx] Run decomposition...
[torch.onnx] Run decomposition... ✅
[torch.onnx] Translate the graph into ONNX...
[torch.onnx] Translate the graph into ONNX... ❌
Traceback (most recent call last):
  File ".venv/lib/python3.13/site-packages/torch/onnx/_internal/exporter/_core.py", line 519, in _handle_call_function_node_with_lowering
    outputs = onnx_function(*onnx_args, **onnx_kwargs)
  File ".venv/lib/python3.13/site-packages/onnxscript/values.py", line 625, in __call__
    return self.func(*args, **kwargs)
           ~~~~~~~~~^^^^^^^^^^^^^^^^^
  File ".venv/lib/python3.13/site-packages/onnxscript/function_libs/torch_lib/ops/core.py", line 1249, in aten_bitwise_and
    assert self.dtype == other.dtype
                         ^^^^^^^^^^^
AttributeError: 'int' object has no attribute 'dtype'

Cause by #2582 , we should check inputs are int or not before checking dtype.

[justinchuby]

Strange: it should have been promoted: https://github.com/pytorch/pytorch/blob/724463d5a2fba369cd14e89215b84d1b01435df7/torch/onnx/_internal/fx/passes/type_promotion.py#L446

[justinchuby]

Thanks. We need to implement all the overloads: bitwise_and.Scalar(Tensor self, Scalar other) -> Tensor