[Documentation] Question: Is 16bit quantization supported by the python tool?

[MustafaIsmail-smarteye]

Describe the documentation issue

Question: Is 16bit quantization supported by the python tool?

There is no mention of 16bit quantization in the documentation but from the python tool it looks like I can set QuanType as QInt16 and QUInt16.

Is it officially supported? Are there any known issues or limitations?

Page / URL

https://onnxruntime.ai/docs/performance/model-optimizations/quantization.html

[xadupre]

Did you try and run into an error?

[MustafaIsmail-smarteye]

Did you try and run into an error?

No honestly just wondering why it is never mentioned in the documentation. Is it still experimental or is it officially supported?

[xadupre]

It should. There are unit tests checking it is working.

[MustafaIsmail-smarteye]

Thanks! Would be great to add a section in the documentation.