torch.Tensor.q_per_channel_axis#

Tensor.q_per_channel_axis() → int#

Given a Tensor quantized by linear (affine) per-channel quantization, returns the index of dimension on which per-channel quantization is applied.