PyTorch eager

import torch
import torch.nn as nn

class Model(nn.Module):
    """
    Model that performs a 3D transposed convolution, scaling, average pooling, bias addition, and scaling.
    """
    def __init__(self, in_channels, out_channels, kernel_size, stride, padding, scale1, scale2, bias_shape):
        super(Model, self).__init__()
        self.conv_transpose = nn.ConvTranspose3d(in_channels, out_channels, kernel_size, stride=stride, padding=padding)
        self.scale1 = nn.Parameter(torch.tensor(scale1))
        self.avg_pool = nn.AvgPool3d(kernel_size=2)
        self.bias = nn.Parameter(torch.randn(bias_shape))
        self.scale2 = nn.Parameter(torch.tensor(scale2))

    def forward(self, x):
        x = self.conv_transpose(x)
        x = x * self.scale1
        x = self.avg_pool(x)
        x = x + self.bias
        x = x * self.scale2
        return x

batch_size = 128
in_channels = 3
out_channels = 16
depth, height, width = 16, 32, 32
kernel_size = 3
stride = 2
padding = 1
scale1 = 0.5
scale2 = 1.0
bias_shape = (out_channels, 1, 1, 1)

def get_inputs():
    return [torch.rand(batch_size, in_channels, depth, height, width)]

def get_init_inputs():
    return [in_channels, out_channels, kernel_size, stride, padding, scale1, scale2, bias_shape]