dicee.trainer.torch_trainer_ddp
Classes
A Trainer based on torch.nn.parallel.DistributedDataParallel |
|
Functions
|
Module Contents
- dicee.trainer.torch_trainer_ddp.make_iterable_verbose(iterable_object, verbose, desc='Default', position=None, leave=True) Iterable [source]
- class dicee.trainer.torch_trainer_ddp.TorchDDPTrainer(args, callbacks)[source]
Bases:
dicee.abstracts.AbstractTrainer
A Trainer based on torch.nn.parallel.DistributedDataParallel
Arguments
- entity_idxs
mapping.
- relation_idxs
mapping.
- form
?
- store
?
- label_smoothing_rate
Using hard targets (0,1) drives weights to infinity. An outlier produces enormous gradients.
- Return type:
torch.utils.data.Dataset