int8 onnx model

#5
by sunseeker001 - opened

Is there a int8 quantized model? Not just compress the weights, but also compress the activation layer.

Thanks.

Sign up or log in to comment