()
| 2419 | |
| 2420 | |
| 2421 | def test_DeepWalk(): |
| 2422 | dev = F.ctx() |
| 2423 | g = dgl.graph(([0, 1, 2, 1, 2, 0], [1, 2, 0, 0, 1, 2])) |
| 2424 | model = nn.DeepWalk( |
| 2425 | g, emb_dim=8, walk_length=2, window_size=1, fast_neg=True, sparse=True |
| 2426 | ) |
| 2427 | model = model.to(dev) |
| 2428 | dataloader = DataLoader( |
| 2429 | torch.arange(g.num_nodes()), batch_size=16, collate_fn=model.sample |
| 2430 | ) |
| 2431 | optim = SparseAdam(model.parameters(), lr=0.01) |
| 2432 | walk = next(iter(dataloader)).to(dev) |
| 2433 | loss = model(walk) |
| 2434 | loss.backward() |
| 2435 | optim.step() |
| 2436 | |
| 2437 | model = nn.DeepWalk( |
| 2438 | g, emb_dim=8, walk_length=2, window_size=1, fast_neg=False, sparse=False |
| 2439 | ) |
| 2440 | model = model.to(dev) |
| 2441 | dataloader = DataLoader( |
| 2442 | torch.arange(g.num_nodes()), batch_size=16, collate_fn=model.sample |
| 2443 | ) |
| 2444 | optim = Adam(model.parameters(), lr=0.01) |
| 2445 | walk = next(iter(dataloader)).to(dev) |
| 2446 | loss = model(walk) |
| 2447 | loss.backward() |
| 2448 | optim.step() |
| 2449 | |
| 2450 | |
| 2451 | @pytest.mark.parametrize("max_degree", [2, 6]) |
nothing calls this directly
no test coverage detected