ValueError: Op "complex_stft_0_promoted" (op_type: cast) Input x="complex_stft_0" expects tensor or scalar of dtype from type domain ['fp16', 'fp32', 'int32', 'bool'] but got tensor[1,1025,250,complex64]

[miku1958]

🐞Describing the bug

• Make sure to only create an issue here for bugs in the coremltools Python package. If this is a bug with the Core ML Framework or Xcode, please submit your bug here: https://developer.apple.com/bug-reporting/
• Provide a clear and consise description of the bug.

Stack Trace

    mlmodel = ct.convert(
  File "<python>/lib/python3.9/site-packages/coremltools/converters/_converters_entry.py", line 581, in convert
    mlmodel = mil_convert(
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/converter.py", line 188, in mil_convert
    return _mil_convert(model, convert_from, convert_to, ConverterRegistry, MLModel, compute_units, **kwargs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/converter.py", line 212, in _mil_convert
    proto, mil_program = mil_convert_to_proto(
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/converter.py", line 288, in mil_convert_to_proto
    prog = frontend_converter(model, **kwargs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/converter.py", line 108, in __call__
    return load(*args, **kwargs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 82, in load
    return _perform_torch_convert(converter, debug)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 116, in _perform_torch_convert
    prog = converter.convert()
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/converter.py", line 581, in convert
    convert_nodes(self.context, self.graph)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 86, in convert_nodes
    raise e     # re-raise exception
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 81, in convert_nodes
    convert_single_node(context, node)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 134, in convert_single_node
    add_op(context, node)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 1549, in pow
    x, y = promote_input_dtypes(inputs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/ops/defs/_utils.py", line 456, in promote_input_dtypes
    input_vars[i] = _promoted_var(var, promoted_dtype)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/ops/defs/_utils.py", line 441, in _promoted_var
    x = mb.cast(
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/ops/registry.py", line 182, in add_op
    return cls._add_op(op_cls_to_add, **kwargs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/builder.py", line 182, in _add_op
    new_op = op_cls(**kwargs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/operation.py", line 191, in __init__
    self._validate_and_set_inputs(input_kv)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/operation.py", line 504, in _validate_and_set_inputs
    self.input_spec.validate_inputs(self.name, self.op_type, input_kvs)
  File "<python>/lib/python3.9/site-packages/coremltools/converters/mil/mil/input_type.py", line 163, in validate_inputs
    raise ValueError(msg.format(name, var.name, input_type.type_str,
ValueError: Op "complex_stft_0_promoted" (op_type: cast) Input x="complex_stft_0" expects tensor or scalar of dtype from type domain ['fp16', 'fp32', 'int32', 'bool'] but got tensor[1,1025,250,complex64]

To Reproduce

I am modifying this script to convert model to MLModel
https://github.com/RVC-Boss/GPT-SoVITS/blob/main/GPT_SoVITS/onnx_export.py
And I have rewritten

        torch.onnx.export(
            self.vits,
            (text_seq, pred_semantic, ref_audio),
            f"onnx/{project_name}/{project_name}_vits.onnx",
            input_names=["text_seq", "pred_semantic", "ref_audio"],
            output_names=["audio"],
            dynamic_axes={
                "text_seq": {1 : "text_length"},
                "pred_semantic": {2 : "pred_length"},
                "ref_audio": {1 : "audio_length"},
            },
            opset_version=17,
            verbose=False
        )

to

        traced_model = torch.jit.trace(self.vits, (text_seq, pred_semantic, ref_audio))
        mlmodel = ct.convert(
            traced_model,
            inputs=[
                ct.TensorType(name="text_seq", shape=text_seq.shape),
                ct.TensorType(name="pred_semantic", shape=pred_semantic.shape),
                ct.TensorType(name="ref_audio", shape=ref_audio.shape),
            ],
        )
        mlmodel.save(f"onnx/{project_name}/{project_name}_vits.mlpackage")

• If the model conversion succeeds, but there is a numerical mismatch in predictions, please include the code used for comparisons.

System environment (please complete the following information):

• coremltools version: 7.2
• OS (e.g. MacOS version or Linux type): macOS 14.4
• Any other relevant version information (e.g. PyTorch or TensorFlow version):

Additional context

• Add anything else about the problem here that you want to share.

[TobyRoseman]

What are the dtypes of text_seq, pred_semantic, and ref_audio? It looks like at least one of them is complex64.

The Core ML Framework does not support models with complex inputs.