dicee.trainer.torch_trainer_ddp

Classes

TorchDDPTrainer	Abstract base class for KGE model trainers.
NodeTrainer

Functions

make_iterable_verbose(→ Iterable)

Module Contents

dicee.trainer.torch_trainer_ddp.make_iterable_verbose(iterable_object, verbose, desc='Default', position=None, leave=True) → Iterable

class dicee.trainer.torch_trainer_ddp.TorchDDPTrainer(args, callbacks)

Bases: dicee.abstracts.AbstractTrainer

Abstract base class for KGE model trainers.

Provides the callback dispatch mechanism shared by all concrete trainer implementations (TorchTrainer, TorchDDPTrainer, etc.). Sub-classes call the on_* hooks at the appropriate points in the training loop so that any registered AbstractCallback can react.

Parameters:

• args (argparse.Namespace or similar) – Processed configuration object. Must expose at least random_seed (int).

• callbacks (list of AbstractCallback) – Ordered list of callback instances to invoke at each lifecycle hook.

fit(*args, **kwargs)

class dicee.trainer.torch_trainer_ddp.NodeTrainer(trainer, model: torch.nn.Module, train_dataset_loader: torch.utils.data.DataLoader, callbacks, num_epochs: int)

trainer

local_rank

global_rank

optimizer

train_dataset_loader

loss_func

callbacks

model

num_epochs

loss_history = []

ctx

scaler

extract_input_outputs(z: list)

train()