Conv2d

class torch.ao.nn.qat.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros', qconfig=None, device=None, dtype=None) [source]

A Conv2d module attached with FakeQuantize modules for weight, used for quantization aware training.

We adopt the same interface as torch.nn.Conv2d, please see https://pytorch.org/docs/stable/nn.html?highlight=conv2d#torch.nn.Conv2d for documentation.

Similar to torch.nn.Conv2d, with FakeQuantize modules initialized to default.

Variables

weight_fake_quant – fake quant module for weight