confusion in Tutorials > Quantization Recipe

[XA23i]

In 3. Post Training Static Quantization
" This method converts both the weights and the activations to 8-bit integers beforehand so there won’t be on-the-fly conversion on the activations during the inference, as the dynamic quantization does, hence improving the performance significantly."

but I think the performance of Post Training Static Quantization is weaker than Post Training dynamic quantization since dynamic quantization can calibrate for each input.

cc @jerryzh168 @z-a-f @vkuzo

[jmarintur]

Hi @svekars and @XA23i, what about this?

" This method converts both the weights and the activations to 8-bit integers beforehand so there won’t be on-the-fly conversion on the activations during the inference, as the dynamic quantization does. While post-training static quantization can significantly enhance inference speed and reduce model size, this method may degrade the original model's performance."

[svekars]

@jerryzh168 any thoughts?

[XA23i]

yes, I think it makes sense. Thank you.