It seems it cannot run on the GPU（4060ti）

[laugh-xixi]

tab_trainer = Trainer(
model=tab_model,
objective="binary",
optimizers=torch.optim.Adam(tab_model.parameters(), lr=0.001),
metrics=[Accuracy, Precision, F1Score, Recall],
seed=random_state_tmp, # 使用相同的随机种子
device="cuda"
)
I selected CUDA for the device here, but it still runs slowly.GPU usage shows 2%