Has anyone tried this quantization method on Regression mode? If so how much performance was degraded with the quantized model?
Has anyone tried this quantization method on Regression mode? If so how much performance was degraded with the quantized model?