Sort out quantized int support

Open dsyme opened this issue 5 years ago • 1 comments

Torch has good support for quantized ints, we whould make this availble.

May 24 '20 02:05 dsyme

e.g.

torch.quantize_per_channel(input, scales, zero_points, axis, dtype) → Tensor

and the other basic support for quantized tensors in torch::*quant*

Nov 13 '20 15:11 dsyme