View source on GitHub
|
Networks are combinations of tf.keras layers (and possibly other networks).
They are tf.keras models that would not be trained alone. It encapsulates
common network structures like a transformer encoder into an easily
handled object with a standardized configuration.
Classes
class AlbertEncoder: ALBERT (https://arxiv.org/abs/1810.04805) text encoder network.
class BertEncoder: Bi-directional Transformer-based encoder network.
class BertEncoderV2: Bi-directional Transformer-based encoder network.
class Classification: Classification network head for BERT modeling.
class EncoderScaffold: Bi-directional Transformer-based encoder network scaffold.
class FNet: FNet encoder network.
class FunnelTransformerEncoder: Funnel Transformer-based encoder network.
class MobileBERTEncoder: A Keras functional API implementation for MobileBERT encoder.
class PackedSequenceEmbedding: An embedding network supporting packed sequences and position ids.
class SpanLabeling: Span labeling network head for BERT modeling.
class SparseMixer: Sparse Mixer encoder network.
class XLNetBase: Base XLNet model.
class XLNetSpanLabeling: Span labeling network head for XLNet on SQuAD2.0.