[Feature] RTMO quantize #3240
Description
What is the feature?
The static int8 quantization accuracy I'm doing for rtmo has also dropped significantly, with almost all the output confidence levels being 0. But fp16 is acceptable.
Any other context?
No response