ScalingNmtEnFr

class opennmt.models.ScalingNmtEnFr(*args, **kwargs)[source]

Defines a big Transformer model using the En-Fr hyperparameters from https://arxiv.org/abs/1806.00187.

The architecture is equivalent to transformer_vaswani_wmt_en_fr_big in Fairseq.

Inherits from: opennmt.models.ScalingNmtEnDe

auto_config(num_replicas=1)[source]

Returns automatic configuration values specific to this model.

Parameters: num_replicas – The number of synchronous model replicas used for the training.
Returns: A partial training configuration.