torch.nn.utils.fuse_conv_bn_weights — PyTorch 2.7 documentation (original) (raw)

torch.nn.utils.fuse_conv_bn_weights(conv_w, conv_b, bn_rm, bn_rv, bn_eps, bn_w, bn_b, transpose=False)[source][source]¶

Fuse convolutional module parameters and BatchNorm module parameters into new convolutional module parameters.

Parameters

• conv_w (torch.Tensor) – Convolutional weight.
• conv_b (Optional _[_torch.Tensor]) – Convolutional bias.
• bn_rm (torch.Tensor) – BatchNorm running mean.
• bn_rv (torch.Tensor) – BatchNorm running variance.
• bn_eps (float) – BatchNorm epsilon.
• bn_w (Optional _[_torch.Tensor]) – BatchNorm weight.
• bn_b (Optional _[_torch.Tensor]) – BatchNorm bias.
• transpose (bool, optional) – If True, transpose the conv weight. Defaults to False.

Returns

Fused convolutional weight and bias.

Return type

Tuple[torch.nn.Parameter, torch.nn.Parameter]