modelzoo.transformers.pytorch.gpt2.model.Gpt2Model#

class modelzoo.transformers.pytorch.gpt2.model.Gpt2Model[source]#

Bases: torch.nn.Module

GPT-2 models

Methods

build_model

forward

The forward pass on the input data.
__call__(*args: Any, **kwargs: Any) → Any#

Call self as a function.

__init__(params)[source]#
static __new__(cls, *args: Any, **kwargs: Any) → Any#
forward(data, reduce_batch=True, output_logits=False)[source]#

The forward pass on the input data. This method returns the loss tensor if output_logits is False. If output_logits is True, the model call will also return the output logits tensor in addition to the loss as a (loss, lm_logits) tuple.

This may be useful for performing post processing on the model’s output logits.

previous

modelzoo.transformers.pytorch.gpt2.model

next

modelzoo.transformers.pytorch.gpt2.run