XlaCallModule or another way of loading StableHLO code

## 🌱 Describe your Feature Request
Are there any plans on supporting a venue for loading StableHLO code?
Currently, I can generate a StableHLO MLIR module and load it into a TensorFlow graph using the XlaCallModule operation, but it isn't supported by the converter.

## How can this feature be used?
This allows for exchanging MLIR modules between platforms and converting them to run on-device.

## Describe alternatives you've considered
I also tried TFLite, but that also doesn't support the operation.

## Additional context
Any other suggestions for converting an MLIR module into CoreML or a way of running them on-device would be awesome!