Bare Model

The Bare model follows the incremental learning schemes, but no CL technique is applied to the model.

Node-level Problems

class NCClassILBareMinibatchTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as NCMinibatchTrainer.

class NCClassILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as NCTrainer.

class NCDomainILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as NCTrainer.

class NCTaskILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

inference(model, _curr_batch, training_states)

The event function to execute inference step. For task-IL, we need to additionally consider task information for the inference step.

Parameters:

• model (torch.nn.Module) – the current trained model.

• curr_batch (object) – the data (or minibatch) for the current iteration.

• curr_training_states (dict) – the dictionary containing the current training states.

Returns:

A dictionary containing the inference results, such as prediction result and loss.

class NCTimeILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as NCTrainer.

Link-level Problems

class LCClassILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

class LCTaskILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

inference(model, _curr_batch, training_states)

The event function to execute inference step. For task-IL, we need to additionally consider task information for the inference step.

Parameters:

• model (torch.nn.Module) – the current trained model.

• curr_batch (object) – the data (or minibatch) for the current iteration.

• curr_training_states (dict) – the dictionary containing the current training states.

Returns:

A dictionary containing the inference results, such as prediction result and loss.

prepareLoader(_curr_dataset, curr_training_states)

The event function to generate dataloaders from the given dataset for the current task. For task-IL, we need to additionally consider task information.

Parameters:

• curr_dataset (object) – The dataset for the current task. Its type is dgl.graph for node-level and link-level problem, and dgl.data.DGLDataset for graph-level problem.

• curr_training_states (dict) – the dictionary containing the current training states.

Returns:

A tuple containing three dataloaders. The trainer considers the first dataloader, second dataloader, and third dataloader as dataloaders for training, validation, and test, respectively.

class LCTimeILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as LCTrainer.

class LPDomainILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as LPTrainer.

class LPTimeILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as LPTrainer.

Graph-level Problems

class GCClassILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

class GCDomainILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as GCClassILBareTrainer.

class GCTaskILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

inference(model, _curr_batch, training_states)

The event function to execute inference step. For task-IL, we need to additionally consider task information for the inference step.

Parameters:

• model (torch.nn.Module) – the current trained model.

• curr_batch (object) – the data (or minibatch) for the current iteration.

• curr_training_states (dict) – the dictionary containing the current training states.

Returns:

A dictionary containing the inference results, such as prediction result and loss.

class GCTimeILBareTrainer(model, scenario, optimizer_fn, loss_fn, device, **kwargs)

This trainer has the same behavior as GCClassILBareTrainer.