How to Customize BERT Encoders with TensorFlow Model Garden

Content Overview Learning objectives Install and import Install the TensorFlow Model Garden pip package Import TensorFlow and other libraries Canonical BERT encoder Customize BERT encoder Use EncoderScaffold Use nlp.layers.TransformerScaffold Build a new encoder Learning objectives Install and import Install the TensorFlow Model Garden pip package Import TensorFlow and other libraries Canonical BERT encoder Customize BERT encoder Use EncoderScaffold Use nlp.layers.TransformerScaffold Build a new encoder Learning objectives The TensorFlow Models NLP library is a collection of tools for building and training modern high-performance natural language models. TensorFlow Models NLP library The tfm.nlp.networks.EncoderScaffold is the core of this library, and lots of new network architectures are proposed to improve the encoder. In this Colab notebook, we will learn how to customize the encoder to employ new network architectures. tfm.nlp.networks.EncoderScaffold Install and import Install the TensorFlow Model Garden pip package tf-models-official is the stable Model Garden package. Note that it may not include the latest changes in the tensorflow_models github repo. To include latest changes, you may install tf-models-nightly, which is the nightly Model Garden package created daily automatically. pip will install all models and dependencies automatically. tf-models-official is the stable Model Garden package. Note that it may not include the latest changes in the tensorflow_models github repo. To include latest changes, you may install tf-models-nightly, which is the nightly Model Garden package created daily automatically. tf-models-official tensorflow_models tf-models-nightly pip will install all models and dependencies automatically. pip pip install -q opencv-python pip install -q opencv-python pip install -q tf-models-official pip install -q tf-models-official Import Tensorflow and other libraries import numpy as np import tensorflow as tf import tensorflow_models as tfm nlp = tfm.nlp import numpy as np import tensorflow as tf import tensorflow_models as tfm nlp = tfm.nlp 2023-12-14 12:09:32.926415: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2023-12-14 12:09:32.926462: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2023-12-14 12:09:32.927992: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered 2023-12-14 12:09:32.926415: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2023-12-14 12:09:32.926462: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2023-12-14 12:09:32.927992: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered Canonical BERT encoder Before learning how to customize the encoder, let's firstly create a canonical BERT enoder and use it to instantiate a bert_classifier.BertClassifier for classification task. bert_classifier.BertClassifier cfg = { "vocab_size": 100, "hidden_size": 32, "num_layers": 3, "num_attention_heads": 4, "intermediate_size": 64, "activation": tfm.utils.activations.gelu, "dropout_rate": 0.1, "attention_dropout_rate": 0.1, "max_sequence_length": 16, "type_vocab_size": 2, "initializer": tf.keras.initializers.TruncatedNormal(stddev=0.02), } bert_encoder = nlp.networks.BertEncoder(**cfg) def build_classifier(bert_encoder): return nlp.models.BertClassifier(bert_encoder, num_classes=2) canonical_classifier_model = build_classifier(bert_encoder) cfg = { "vocab_size": 100, "hidden_size": 32, "num_layers": 3, "num_attention_heads": 4, "intermediate_size": 64, "activation": tfm.utils.activations.gelu, "dropout_rate": 0.1, "attention_dropout_rate": 0.1, "max_sequence_length": 16, "type_vocab_size": 2, "initializer": tf.keras.initializers.TruncatedNormal(stddev=0.02), } bert_encoder = nlp.networks.BertEncoder(**cfg) def build_classifier(bert_encoder): return nlp.models.BertClassifier(bert_encoder, num_classes=2) canonical_classifier_model = build_classifier(bert_encoder) canonical_classifier_model can be trained using the training data. For details about how to train the model, please see the Fine tuning bert notebook. We skip the code that trains the model here. canonical_classifier_model Fine tuning bert After training, we can apply the model to do prediction. def predict(model): batch_size = 3 np.random.seed(0) word_ids = np.random.randint( cfg["vocab_size"], size=(batch_size, cfg["max_sequence_length"])) mask = np.random.randint(2, size=(batch_size, cfg["max_sequence_length"])) type_ids = np.random.randint( cfg["type_vocab_size"], size=(batch_size, cfg["max_sequence_length"])) print(model([word_ids, mask, type_ids], training=False)) predict(canonical_classifier_model) def predict(model): batch_size = 3 np.random.seed(0) word_ids = np.random.randint( cfg["vocab_size"], size=(batch_size, cfg["max_sequence_length"])) mask = np.random.randint(2, size=(batch_size, cfg["max_sequence_length"])) type_ids = np.random.randint( cfg["type_vocab_size"], size=(batch_size, cfg["max_sequence_length"])) print(model([word_ids, mask, type_ids], training=False)) predict(canonical_classifier_model) tf.Tensor( [[ 0.03545166 0.30729884] [ 0.00677404 0.17251147] [-0.07276718 0.17345032]], shape=(3, 2), dtype=float32) tf.Tensor( [[ 0.03545166 0.30729884] [ 0.00677404 0.17251147] [-0.07276718 0.17345032]], shape=(3, 2), dtype=float32) Customize BERT encoder One BERT encoder consists of an embedding network and multiple transformer blocks, and each transformer block contains an attention layer and a feedforward layer. We provide easy ways to customize each of those components via (1) EncoderScaffold and (2) TransformerScaffold. EncoderScaffold TransformerScaffold Use EncoderScaffold networks.EncoderScaffold allows users to provide a custom embedding subnetwork (which will replace the standard embedding logic) and/or a custom hidden layer class (which will replace the Transformer instantiation in the encoder). networks.EncoderScaffold Transformer Without Customization Without any customization, networks.EncoderScaffold behaves the same the canonical networks.BertEncoder. networks.EncoderScaffold networks.BertEncoder As shown in the following example, networks.EncoderScaffold can load networks.BertEncoder's weights and output the same values: networks.EncoderScaffold networks.BertEncoder default_hidden_cfg = dict( num_attention_heads=cfg["num_attention_heads"], intermediate_size=cfg["intermediate_size"], intermediate_activation=cfg["activation"], dropout_rate=cfg["dropout_rate"], attention_dropout_rate=cfg["attention_dropout_rate"], kernel_initializer=cfg["initializer"], ) default_embedding_cfg = dict( vocab_size=cfg["vocab_size"], type_vocab_size=cfg["type_vocab_size"], hidden_size=cfg["hidden_size"], initializer=cfg["initializer"], dropout_rate=cfg["dropout_rate"], max_seq_length=cfg["max_sequence_length"] ) default_kwargs = dict( hidden_cfg=default_hidden_cfg, embedding_cfg=default_embedding_cfg, num_hidden_instances=cfg["num_layers"], pooled_output_dim=cfg["hidden_size"], return_all_layer_outputs=True, pooler_layer_initializer=cfg["initializer"], ) encoder_scaffold = nlp.networks.EncoderScaffold(**default_kwargs) classifier_model_from_encoder_scaffold = build_classifier(encoder_scaffold) classifier_model_from_encoder_scaffold.set_weights( canonical_classifier_model.get_weights()) predict(classifier_model_from_encoder_scaffold) default_hidden_cfg = dict( num_attention_heads=cfg["num_attention_heads"], intermediate_size=cfg["intermediate_size"], intermediate_activation=cfg["activation"], dropout_rate=cfg["dropout_rate"], attention_dropout_rate=cfg["attention_dropout_rate"], kernel_initializer=cfg["initializer"], ) default_embedding_cfg = dict( vocab_size=cfg["vocab_size"], type_vocab_size=cfg["type_vocab_size"], hidden_size=cfg["hidden_size"], initializer=cfg["initializer"], dropout_rate=cfg["dropout_rate"], max_seq_length=cfg["max_sequence_length"] ) default_kwargs = dict( hidden_cfg=default_hidden_cfg, embedding_cfg=default_embedding_cfg, num_hidden_instances=cfg["num_layers"], pooled_output_dim=cfg["hidden_size"], return_all_layer_outputs=True, pooler_layer_initializer=cfg["initializer"], ) encoder_scaffold = nlp.networks.EncoderScaffold(**default_kwargs) classifier_model_from_encoder_scaffold = build_classifier(encoder_scaffold) classifier_model_from_encoder_scaffold.set_weights( canonical_classifier_model.get_weights()) predict(classifier_model_from_encoder_scaffold) WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. tf.Tensor( [[ 0.03545166 0.30729884] [ 0.00677404 0.17251147] [-0.07276718 0.17345032]], shape=(3, 2), dtype=float32) WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. tf.Tensor( [[ 0.03545166 0.30729884] [ 0.00677404 0.17251147] [-0.07276718 0.17345032]], shape=(3, 2), dtype=float32) Customize Embedding Next, we show how to use a customized embedding network. We first build an embedding network that would replace the default network. This one will have 2 inputs (mask and word_ids) instead of 3, and won't use positional embeddings. mask word_ids word_ids = tf.keras.layers.Input( shape=(cfg['max_sequence_length'],), dtype=tf.int32, name="input_word_ids") mask = tf.keras.layers.Input( shape=(cfg['max_sequence_length'],), dtype=tf.int32, name="input_mask") embedding_layer = nlp.layers.OnDeviceEmbedding( vocab_size=cfg['vocab_size'], embedding_width=cfg['hidden_size'], initializer=cfg["initializer"], name="word_embeddings") word_embeddings = embedding_layer(word_ids) attention_mask = nlp.layers.SelfAttentionMask()([word_embeddings, mask]) new_embedding_network = tf.keras.Model([word_ids, mask], [word_embeddings, attention_mask]) word_ids = tf.keras.layers.Input( shape=(cfg['max_sequence_length'],), dtype=tf.int32, name="input_word_ids") mask = tf.keras.layers.Input( shape=(cfg['max_sequence_length'],), dtype=tf.int32, name="input_mask") embedding_layer = nlp.layers.OnDeviceEmbedding( vocab_size=cfg['vocab_size'], embedding_width=cfg['hidden_size'], initializer=cfg["initializer"], name="word_embeddings") word_embeddings = embedding_layer(word_ids) attention_mask = nlp.layers.SelfAttentionMask()([word_embeddings, mask]) new_embedding_network = tf.keras.Model([word_ids, mask], [word_embeddings, attention_mask]) /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( Inspecting new_embedding_network, we can see it takes two inputs: input_word_ids and input_mask. new_embedding_network input_word_ids input_mask tf.keras.utils.plot_model(new_embedding_network, show_shapes=True, dpi=48) tf.keras.utils.plot_model(new_embedding_network, show_shapes=True, dpi=48) We can then build a new encoder using the above new_embedding_network. new_embedding_network kwargs = dict(default_kwargs) # Use new embedding network. kwargs['embedding_cls'] = new_embedding_network kwargs['embedding_data'] = embedding_layer.embeddings encoder_with_customized_embedding = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_customized_embedding) # ... Train the model ... print(classifier_model.inputs) # Assert that there are only two inputs. assert len(classifier_model.inputs) == 2 kwargs = dict(default_kwargs) # Use new embedding network. kwargs['embedding_cls'] = new_embedding_network kwargs['embedding_data'] = embedding_layer.embeddings encoder_with_customized_embedding = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_customized_embedding) # ... Train the model ... print(classifier_model.inputs) # Assert that there are only two inputs. assert len(classifier_model.inputs) == 2 WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( [ , ] WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. WARNING:absl:The `Transformer` layer is deprecated. Please directly use `TransformerEncoderBlock`. /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( [ , ] Customized Transformer Users can also override the hidden_cls argument in networks.EncoderScaffold's constructor employ a customized Transformer layer. hidden_cls networks.EncoderScaffold See the source of nlp.layers.ReZeroTransformer for how to implement a customized Transformer layer. the source of nlp.layers.ReZeroTransformer The following is an example of using nlp.layers.ReZeroTransformer: nlp.layers.ReZeroTransformer kwargs = dict(default_kwargs) # Use ReZeroTransformer. kwargs['hidden_cls'] = nlp.layers.ReZeroTransformer encoder_with_rezero_transformer = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_rezero_transformer) # ... Train the model ... predict(classifier_model) # Assert that the variable `rezero_alpha` from ReZeroTransformer exists. assert 'rezero_alpha' in ''.join([x.name for x in classifier_model.trainable_weights]) kwargs = dict(default_kwargs) # Use ReZeroTransformer. kwargs['hidden_cls'] = nlp.layers.ReZeroTransformer encoder_with_rezero_transformer = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_rezero_transformer) # ... Train the model ... predict(classifier_model) # Assert that the variable `rezero_alpha` from ReZeroTransformer exists. assert 'rezero_alpha' in ''.join([x.name for x in classifier_model.trainable_weights]) tf.Tensor( [[-0.08663296 0.09281035] [-0.07291833 0.36477187] [-0.08730186 0.1503254 ]], shape=(3, 2), dtype=float32) tf.Tensor( [[-0.08663296 0.09281035] [-0.07291833 0.36477187] [-0.08730186 0.1503254 ]], shape=(3, 2), dtype=float32) Use nlp.layers.TransformerScaffold nlp.layers.TransformerScaffold The above method of customizing the model requires rewriting the whole nlp.layers.Transformer layer, while sometimes you may only want to customize either attention layer or feedforward block. In this case, nlp.layers.TransformerScaffold can be used. nlp.layers.Transformer nlp.layers.TransformerScaffold Customize Attention Layer User can also override the attention_cls argument in layers.TransformerScaffold's constructor to employ a customized Attention layer. attention_cls layers.TransformerScaffold See the source of nlp.layers.TalkingHeadsAttention for how to implement a customized Attention layer. the source of nlp.layers.TalkingHeadsAttention Attention Following is an example of using nlp.layers.TalkingHeadsAttention: nlp.layers.TalkingHeadsAttention # Use TalkingHeadsAttention hidden_cfg = dict(default_hidden_cfg) hidden_cfg['attention_cls'] = nlp.layers.TalkingHeadsAttention kwargs = dict(default_kwargs) kwargs['hidden_cls'] = nlp.layers.TransformerScaffold kwargs['hidden_cfg'] = hidden_cfg encoder = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder) # ... Train the model ... predict(classifier_model) # Assert that the variable `pre_softmax_weight` from TalkingHeadsAttention exists. assert 'pre_softmax_weight' in ''.join([x.name for x in classifier_model.trainable_weights]) # Use TalkingHeadsAttention hidden_cfg = dict(default_hidden_cfg) hidden_cfg['attention_cls'] = nlp.layers.TalkingHeadsAttention kwargs = dict(default_kwargs) kwargs['hidden_cls'] = nlp.layers.TransformerScaffold kwargs['hidden_cfg'] = hidden_cfg encoder = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder) # ... Train the model ... predict(classifier_model) # Assert that the variable `pre_softmax_weight` from TalkingHeadsAttention exists. assert 'pre_softmax_weight' in ''.join([x.name for x in classifier_model.trainable_weights]) tf.Tensor( [[-0.20591784 0.09203205] [-0.0056177 -0.10278902] [-0.21681327 -0.12282 ]], shape=(3, 2), dtype=float32) tf.Tensor( [[-0.20591784 0.09203205] [-0.0056177 -0.10278902] [-0.21681327 -0.12282 ]], shape=(3, 2), dtype=float32) tf.keras.utils.plot_model(encoder_with_rezero_transformer, show_shapes=True, dpi=48) tf.keras.utils.plot_model(encoder_with_rezero_transformer, show_shapes=True, dpi=48) Customize Feedforward Layer Similiarly, one could also customize the feedforward layer. See the source of nlp.layers.GatedFeedforward for how to implement a customized feedforward layer. the source of nlp.layers.GatedFeedforward Following is an example of using nlp.layers.GatedFeedforward: nlp.layers.GatedFeedforward # Use GatedFeedforward hidden_cfg = dict(default_hidden_cfg) hidden_cfg['feedforward_cls'] = nlp.layers.GatedFeedforward kwargs = dict(default_kwargs) kwargs['hidden_cls'] = nlp.layers.TransformerScaffold kwargs['hidden_cfg'] = hidden_cfg encoder_with_gated_feedforward = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_gated_feedforward) # ... Train the model ... predict(classifier_model) # Assert that the variable `gate` from GatedFeedforward exists. assert 'gate' in ''.join([x.name for x in classifier_model.trainable_weights]) # Use GatedFeedforward hidden_cfg = dict(default_hidden_cfg) hidden_cfg['feedforward_cls'] = nlp.layers.GatedFeedforward kwargs = dict(default_kwargs) kwargs['hidden_cls'] = nlp.layers.TransformerScaffold kwargs['hidden_cfg'] = hidden_cfg encoder_with_gated_feedforward = nlp.networks.EncoderScaffold(**kwargs) classifier_model = build_classifier(encoder_with_gated_feedforward) # ... Train the model ... predict(classifier_model) # Assert that the variable `gate` from GatedFeedforward exists. assert 'gate' in ''.join([x.name for x in classifier_model.trainable_weights]) /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( tf.Tensor( [[-0.10270456 -0.10999684] [-0.03512481 0.15430304] [-0.23601504 -0.18162844]], shape=(3, 2), dtype=float32) /tmpfs/src/tf_docs_env/lib/python3.9/site-packages/keras/src/initializers/initializers.py:120: UserWarning: The initializer TruncatedNormal is unseeded and being called multiple times, which will return identical values each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initializer instance more than once. warnings.warn( tf.Tensor( [[-0.10270456 -0.10999684] [-0.03512481 0.15430304] [-0.23601504 -0.18162844]], shape=(3, 2), dtype=float32) Build a new Encoder Finally, you could also build a new encoder using building blocks in the modeling library. See the source for nlp.networks.AlbertEncoder as an example of how to do this. the source for nlp.networks.AlbertEncoder Here is an example using nlp.networks.AlbertEncoder: nlp.networks.AlbertEncoder albert_encoder = nlp.networks.AlbertEncoder(**cfg) classifier_model = build_classifier(albert_encoder) # ... Train the model ... predict(classifier_model) albert_encoder = nlp.networks.AlbertEncoder(**cfg) classifier_model = build_classifier(albert_encoder) # ... Train the model ... predict(classifier_model) tf.Tensor( [[-0.00369881 -0.2540995 ] [ 0.1235221 -0.2959229 ] [-0.08698564 -0.17653546]], shape=(3, 2), dtype=float32) tf.Tensor( [[-0.00369881 -0.2540995 ] [ 0.1235221 -0.2959229 ] [-0.08698564 -0.17653546]], shape=(3, 2), dtype=float32) Inspecting the albert_encoder, we see it stacks the same Transformer layer multiple times (note the loop-back on the "Transformer" block below.. albert_encoder Transformer tf.keras.utils.plot_model(albert_encoder, show_shapes=True, dpi=48) tf.keras.utils.plot_model(albert_encoder, show_shapes=True, dpi=48) Originally published on the TensorFlow website, this article appears here under a new headline and is licensed under CC BY 4.0. Code samples shared under the Apache 2.0 License. Originally published on the TensorFlow website, this article appears here under a new headline and is licensed under CC BY 4.0. Code samples shared under the Apache 2.0 License. Originally published on the TensorFlow website, this article appears here under a new headline and is licensed under CC BY 4.0. Code samples shared under the Apache 2.0 License. TensorFlow TensorFlow