f3d0dfcb7caf50d123df02773495e5b52a92bd76,test/nn/conv/test_gat_conv.py,,test_gat_conv,#,5

Before Change

    assert conv((x, x), edge_index)[0].size() == (num_nodes, 2 * out_channels)
    jitcls = conv.jittable(x=x, edge_index=edge_index, full_eval=True)
    jitconv = jitcls(in_channels, out_channels, heads=2, dropout=0.5)
    jitconv.load_state_dict(conv.state_dict())
    jittedconv = torch.jit.script(jitconv)
    conv.eval()
    jitconv.eval()
    jittedconv.eval()
    assert (torch.abs(conv(x, edge_index)[0] -
            jitconv(x, edge_index)[0]) < 1e-6).all().item()
    assert (torch.abs(conv(x, edge_index)[0] -

After Change

    assert conv._alpha is None

    model = Net2()
    model.conv = conv.jittable(x=x, edge_index=edge_index)
    model = torch.jit.script(model)
    assert model(x, edge_index)[0].tolist() == result[0].tolist()
    assert model(x, edge_index)[1][0].tolist() == result[1][0].tolist()
    assert model(x, edge_index)[1][1].tolist() == result[1][1].tolist()

    conv = GATConv(in_channels, out_channels, heads=2, concat=False)
    out = conv(x, edge_index)
    assert out.size() == (num_nodes, out_channels)
    assert conv((x, x), edge_index).tolist() == out.tolist()

    model = Net1()

Instances