bug in models > ali.py

[jaks19]

Hello, in models > ali.py, line 81 of function set_z, I believe is a bug?

def set_z(self, var=None, volatile=False):
        if var is None:
            self.normal_z = var
        else:
            if self.gpu_ids:
                self.normal_z = Variable(torch.randn((self.opt.batch_size, self.encoder.k)).cuda(), volatile=volatile)
            else:
                self.normal_z = Variable(torch.randn((self.opt.batch_size, self.encoder.k)), volatile=volatile)

Would we not like to have something like if var is not None for the first condition? Otherwise, when GibbsNet has sampled a z from the unclamped chain and passed it to ALI as implemented below, we are losing this z and simply restarting it in ALI, essentially losing the unclamped chain?

def forward(self, volatile=False):
        self.sampling()

        # clamped chain : ALI model
        self.ali_model.set_z(var=self.z)
        self.ali_model.set_input(self.x.data, is_z_given=True)

        self.ali_model.forward()

Any help is appreciated. Thanks.

[thuangb]

yeah, I think it should be if var is not None: