import torch
import torch.nn as nn

class Model(nn.Module):
    def __init__(self):
        super(Model, self).__init__()
        self.linear = nn.Linear(1,1)

    def forward(self, x):
        y = self.linear(x)
        return y

x = torch.zeros(1)

model = Model()

trace = torch.jit.trace(x)(model)

trace.save('trace.pt')