TypeError: add_param_group() missing 1 required positional argument: 'param_group'

[langineer]

trying to run training for the Grace model:

dataset = get_node_dataset('cora')
dataloader = DataLoader(dataset[0])
encoder = Encoder(feat_dim=dataset[0].x.shape[1], hidden_dim=embed_dim, n_layers=2, gnn='gcn', node_level=True, graph_level=False)
grace = GRACE(dim=embed_dim, dropE_rate_1=0.2, dropE_rate_2=0.4, maskN_rate_1=0.3, maskN_rate_2=0.4, tau=0.4, device=0)
grace.train(encoders=encoder, data_loader= dataloader,epochs=10, per_epoch_out=False, optimizer = torch.optim.Adam)

getting an error from contrastive.py:
optimizer.add_param_group({"params": self.proj_head_n.parameters()})

TypeError: add_param_group() missing 1 required positional argument: 'param_group'

any pointers?