torch.utils.flop_counter.conv_flop_count#
- torch.utils.flop_counter.conv_flop_count(x_shape, w_shape, out_shape, transposed=False)[source]#
Count flops for convolution.
Note only multiplication is counted. Computation for bias are ignored. Flops for a transposed convolution are calculated as flops = (x_shape[2:] * prod(w_shape) * batch_size). :param x_shape: The input shape before convolution. :type x_shape: list(int) :param w_shape: The filter shape. :type w_shape: list(int) :param out_shape: The output shape after convolution. :type out_shape: list(int) :param transposed: is the convolution transposed :type transposed: bool
- Returns:
the number of flops
- Return type:
