Add support to Rocm (AMD) onnx provider

Hi, in [models/yolox.py](https://github.com/Unstructured-IO/unstructured-inference/blob/main/unstructured_inference/models/yolox.py#L76) and detectron2onnx.py, we can add support to AMD GPUs by modifying `ordered_providers` list to
```python
ordered_providers = [
    "ROCMExecutionProvider",
    "TensorrtExecutionProvider",
    "CUDAExecutionProvider"
]
```