# MobileBERT¶

## Overview¶

The MobileBERT model was proposed in MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices by Zhiqing Sun, Hongkun Yu, Xiaodan Song, Renjie Liu, Yiming Yang, and Denny Zhou. It’s a bidirectional transformer based on the BERT model, which is compressed and accelerated using several approaches.

The abstract from the paper is the following:

Natural Language Processing (NLP) has recently achieved great success by using huge pre-trained models with hundreds of millions of parameters. However, these models suffer from heavy model sizes and high latency such that they cannot be deployed to resource-limited mobile devices. In this paper, we propose MobileBERT for compressing and accelerating the popular BERT model. Like the original BERT, MobileBERT is task-agnostic, that is, it can be generically applied to various downstream NLP tasks via simple fine-tuning. Basically, MobileBERT is a thin version of BERT_LARGE, while equipped with bottleneck structures and a carefully designed balance between self-attentions and feed-forward networks. To train MobileBERT, we first train a specially designed teacher model, an inverted-bottleneck incorporated BERT_LARGE model. Then, we conduct knowledge transfer from this teacher to MobileBERT. Empirical studies show that MobileBERT is 4.3x smaller and 5.5x faster than BERT_BASE while achieving competitive results on well-known benchmarks. On the natural language inference tasks of GLUE, MobileBERT achieves a GLUEscore o 77.7 (0.6 lower than BERT_BASE), and 62 ms latency on a Pixel 4 phone. On the SQuAD v1.1/v2.0 question answering task, MobileBERT achieves a dev F1 score of 90.0/79.2 (1.5/2.1 higher than BERT_BASE).

Tips:

• MobileBERT is a model with absolute position embeddings so it’s usually advised to pad the inputs on the right rather than the left.

• MobileBERT is similar to BERT and therefore relies on the masked language modeling (MLM) objective. It is therefore efficient at predicting masked tokens and at NLU in general, but is not optimal for text generation. Models trained with a causal language modeling (CLM) objective are better in that regard.

This model was contributed by vshampor. The original code can be found here.

## MobileBertConfig¶

class transformers.MobileBertConfig(vocab_size=30522, hidden_size=512, num_hidden_layers=24, num_attention_heads=4, intermediate_size=512, hidden_act='relu', hidden_dropout_prob=0.0, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=2, initializer_range=0.02, layer_norm_eps=1e-12, pad_token_id=0, embedding_size=128, trigram_input=True, use_bottleneck=True, intra_bottleneck_size=128, use_bottleneck_attention=False, key_query_shared_bottleneck=True, num_feedforward_networks=4, normalization_type='no_norm', classifier_activation=True, classifier_dropout=None, **kwargs)[source]

This is the configuration class to store the configuration of a MobileBertModel or a TFMobileBertModel. It is used to instantiate a MobileBERT model according to the specified arguments, defining the model architecture.

Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.

Parameters
• vocab_size (int, optional, defaults to 30522) – Vocabulary size of the MobileBERT model. Defines the number of different tokens that can be represented by the inputs_ids passed when calling MobileBertModel or TFMobileBertModel.

• hidden_size (int, optional, defaults to 512) – Dimensionality of the encoder layers and the pooler layer.

• num_hidden_layers (int, optional, defaults to 24) – Number of hidden layers in the Transformer encoder.

• num_attention_heads (int, optional, defaults to 4) – Number of attention heads for each attention layer in the Transformer encoder.

• intermediate_size (int, optional, defaults to 512) – Dimensionality of the “intermediate” (often named feed-forward) layer in the Transformer encoder.

• hidden_act (str or function, optional, defaults to "relu") – The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "silu" and "gelu_new" are supported.

• hidden_dropout_prob (float, optional, defaults to 0.0) – The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.

• attention_probs_dropout_prob (float, optional, defaults to 0.1) – The dropout ratio for the attention probabilities.

• max_position_embeddings (int, optional, defaults to 512) – The maximum sequence length that this model might ever be used with. Typically set this to something large just in case (e.g., 512 or 1024 or 2048).

• type_vocab_size (int, optional, defaults to 2) – The vocabulary size of the token_type_ids passed when calling MobileBertModel or TFMobileBertModel.

• initializer_range (float, optional, defaults to 0.02) – The standard deviation of the truncated_normal_initializer for initializing all weight matrices.

• layer_norm_eps (float, optional, defaults to 1e-12) – The epsilon used by the layer normalization layers.

• pad_token_id (int, optional, defaults to 0) – The ID of the token in the word embedding to use as padding.

• embedding_size (int, optional, defaults to 128) – The dimension of the word embedding vectors.

• trigram_input (bool, optional, defaults to True) – Use a convolution of trigram as input.

• use_bottleneck (bool, optional, defaults to True) – Whether to use bottleneck in BERT.

• intra_bottleneck_size (int, optional, defaults to 128) – Size of bottleneck layer output.

• use_bottleneck_attention (bool, optional, defaults to False) – Whether to use attention inputs from the bottleneck transformation.

• key_query_shared_bottleneck (bool, optional, defaults to True) – Whether to use the same linear transformation for query&key in the bottleneck.

• num_feedforward_networks (int, optional, defaults to 4) – Number of FFNs in a block.

• normalization_type (str, optional, defaults to "no_norm") – The normalization type in MobileBERT.

• classifier_dropout (float, optional) – The dropout ratio for the classification head.

Examples:

>>> from transformers import MobileBertModel, MobileBertConfig

>>> # Initializing a MobileBERT configuration
>>> configuration = MobileBertConfig()

>>> # Initializing a model from the configuration above
>>> model = MobileBertModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config


Attributes: pretrained_config_archive_map (Dict[str, str]): A dictionary containing all the available pre-trained checkpoints.

## MobileBertTokenizer¶

class transformers.MobileBertTokenizer(vocab_file, do_lower_case=True, do_basic_tokenize=True, never_split=None, unk_token='[UNK]', sep_token='[SEP]', pad_token='[PAD]', cls_token='[CLS]', mask_token='[MASK]', tokenize_chinese_chars=True, strip_accents=None, **kwargs)[source]

Construct a MobileBERT tokenizer.

BertTokenizer and runs end-to-end tokenization: punctuation splitting and wordpiece.

Refer to superclass BertTokenizer for usage examples and documentation concerning parameters.

## MobileBertTokenizerFast¶

class transformers.MobileBertTokenizerFast(vocab_file=None, tokenizer_file=None, do_lower_case=True, unk_token='[UNK]', sep_token='[SEP]', pad_token='[PAD]', cls_token='[CLS]', mask_token='[MASK]', tokenize_chinese_chars=True, strip_accents=None, **kwargs)[source]

Construct a “fast” MobileBERT tokenizer (backed by HuggingFace’s tokenizers library).

MobileBertTokenizerFast is identical to BertTokenizerFast and runs end-to-end tokenization: punctuation splitting and wordpiece.

Refer to superclass BertTokenizerFast for usage examples and documentation concerning parameters.

slow_tokenizer_class

alias of transformers.models.mobilebert.tokenization_mobilebert.MobileBertTokenizer

## MobileBert specific outputs¶

class transformers.models.mobilebert.modeling_mobilebert.MobileBertForPreTrainingOutput(loss: Optional[torch.FloatTensor] = None, prediction_logits: torch.FloatTensor = None, seq_relationship_logits: torch.FloatTensor = None, hidden_states: Optional[Tuple[torch.FloatTensor]] = None, attentions: Optional[Tuple[torch.FloatTensor]] = None)[source]

Output type of MobileBertForPreTraining.

Parameters
• loss (optional, returned when labels is provided, torch.FloatTensor of shape (1,)) – Total loss as the sum of the masked language modeling loss and the next sequence prediction (classification) loss.

• prediction_logits (torch.FloatTensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• seq_relationship_logits (torch.FloatTensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) –

Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) –

Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

class transformers.models.mobilebert.modeling_tf_mobilebert.TFMobileBertForPreTrainingOutput(loss: Optional[tensorflow.python.framework.ops.Tensor] = None, prediction_logits: tensorflow.python.framework.ops.Tensor = None, seq_relationship_logits: tensorflow.python.framework.ops.Tensor = None, hidden_states: Optional[Tuple[tensorflow.python.framework.ops.Tensor]] = None, attentions: Optional[Tuple[tensorflow.python.framework.ops.Tensor]] = None)[source]

Output type of TFMobileBertForPreTraining.

Parameters
• prediction_logits (tf.Tensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• seq_relationship_logits (tf.Tensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) –

Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) –

Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

## MobileBertModel¶

class transformers.MobileBertModel(config, add_pooling_layer=True)[source]

The bare MobileBert Model transformer outputting raw hidden-states without any specific head on top.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

https://arxiv.org/pdf/2004.02984.pdf

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_hidden_states=None, output_attentions=None, return_dict=None)[source]

The MobileBertModel forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

Returns

A BaseModelOutputWithPooling or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size)) – Sequence of hidden-states at the output of the last layer of the model.

• pooler_output (torch.FloatTensor of shape (batch_size, hidden_size)) – Last layer hidden-state of the first token of the sequence (classification token) after further processing through the layers used for the auxiliary pretraining task. E.g. for BERT-family of models, this returns the classification token after processing through a linear layer and a tanh activation function. The linear layer weights are trained from the next sentence prediction (classification) objective during pretraining.

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

BaseModelOutputWithPooling or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertModel
>>> import torch

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
>>> outputs = model(**inputs)

>>> last_hidden_states = outputs.last_hidden_state


## MobileBertForPreTraining¶

class transformers.MobileBertForPreTraining(config)[source]

MobileBert Model with two heads on top as done during the pretraining: a masked language modeling head and a next sentence prediction (classification) head.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, next_sentence_label=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForPreTraining forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size, sequence_length), optional) – Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size]

• next_sentence_label (torch.LongTensor of shape (batch_size,), optional) –

Labels for computing the next sequence prediction (classification) loss. Input should be a sequence pair (see input_ids docstring) Indices should be in [0, 1]:

• 0 indicates sequence B is a continuation of sequence A,

• 1 indicates sequence B is a random sequence.

Returns

A MobileBertForPreTrainingOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (optional, returned when labels is provided, torch.FloatTensor of shape (1,)) – Total loss as the sum of the masked language modeling loss and the next sequence prediction (classification) loss.

• prediction_logits (torch.FloatTensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• seq_relationship_logits (torch.FloatTensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Examples:

>>> from transformers import MobileBertTokenizer, MobileBertForPreTraining
>>> import torch

>>> input_ids = torch.tensor(tokenizer.encode("Hello, my dog is cute", add_special_tokens=True)).unsqueeze(0)  # Batch size 1
>>> outputs = model(input_ids)

>>> prediction_logits = outputs.prediction_logits
>>> seq_relationship_logits = outputs.seq_relationship_logits


Return type

MobileBertForPreTrainingOutput or tuple(torch.FloatTensor)

class transformers.MobileBertForMaskedLM(config)[source]

MobileBert Model with a language modeling head on top.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForMaskedLM forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size, sequence_length), optional) – Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size]

Returns

A MaskedLMOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) – Masked language modeling (MLM) loss.

• logits (torch.FloatTensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

MaskedLMOutput or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertForMaskedLM
>>> import torch

>>> inputs = tokenizer("The capital of France is [MASK].", return_tensors="pt")
>>> labels = tokenizer("The capital of France is Paris.", return_tensors="pt")["input_ids"]

>>> outputs = model(**inputs, labels=labels)
>>> loss = outputs.loss
>>> logits = outputs.logits


## MobileBertForNextSentencePrediction¶

class transformers.MobileBertForNextSentencePrediction(config)[source]

MobileBert Model with a next sentence prediction (classification) head on top.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None, **kwargs)[source]

The MobileBertForNextSentencePrediction forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size,), optional) –

Labels for computing the next sequence prediction (classification) loss. Input should be a sequence pair (see input_ids docstring) Indices should be in [0, 1].

• 0 indicates sequence B is a continuation of sequence A,

• 1 indicates sequence B is a random sequence.

Returns

A NextSentencePredictorOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when next_sentence_label is provided) – Next sequence prediction (classification) loss.

• logits (torch.FloatTensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Examples:

>>> from transformers import MobileBertTokenizer, MobileBertForNextSentencePrediction
>>> import torch

>>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
>>> next_sentence = "The sky is blue due to the shorter wavelength of blue light."
>>> encoding = tokenizer(prompt, next_sentence, return_tensors='pt')

>>> outputs = model(**encoding, labels=torch.LongTensor([1]))
>>> loss = outputs.loss
>>> logits = outputs.logits


Return type

NextSentencePredictorOutput or tuple(torch.FloatTensor)

## MobileBertForSequenceClassification¶

class transformers.MobileBertForSequenceClassification(config)[source]

MobileBert Model transformer with a sequence classification/regression head on top (a linear layer on top of the pooled output) e.g. for GLUE tasks.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForSequenceClassification forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size,), optional) – Labels for computing the sequence classification/regression loss. Indices should be in [0, ..., config.num_labels - 1]. If config.num_labels == 1 a regression loss is computed (Mean-Square loss), If config.num_labels > 1 a classification loss is computed (Cross-Entropy).

Returns

A SequenceClassifierOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) – Classification (or regression if config.num_labels==1) loss.

• logits (torch.FloatTensor of shape (batch_size, config.num_labels)) – Classification (or regression if config.num_labels==1) scores (before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

SequenceClassifierOutput or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertForSequenceClassification
>>> import torch

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
>>> labels = torch.tensor([1]).unsqueeze(0)  # Batch size 1
>>> outputs = model(**inputs, labels=labels)
>>> loss = outputs.loss
>>> logits = outputs.logits


## MobileBertForMultipleChoice¶

class transformers.MobileBertForMultipleChoice(config)[source]

MobileBert Model with a multiple choice classification head on top (a linear layer on top of the pooled output and a softmax) e.g. for RocStories/SWAG tasks.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForMultipleChoice forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, num_choices, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, num_choices, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, num_choices, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, num_choices, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, num_choices, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size,), optional) – Labels for computing the multiple choice classification loss. Indices should be in [0, ..., num_choices-1] where num_choices is the size of the second dimension of the input tensors. (See input_ids above)

Returns

A MultipleChoiceModelOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) – Classification loss.

• logits (torch.FloatTensor of shape (batch_size, num_choices)) – num_choices is the second dimension of the input tensors. (see input_ids above).

Classification scores (before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

MultipleChoiceModelOutput or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertForMultipleChoice
>>> import torch

>>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
>>> choice0 = "It is eaten with a fork and a knife."
>>> choice1 = "It is eaten while held in the hand."
>>> labels = torch.tensor(0).unsqueeze(0)  # choice0 is correct (according to Wikipedia ;)), batch size 1

>>> encoding = tokenizer([prompt, prompt], [choice0, choice1], return_tensors='pt', padding=True)
>>> outputs = model(**{k: v.unsqueeze(0) for k,v in encoding.items()}, labels=labels)  # batch size is 1

>>> # the linear classifier still needs to be trained
>>> loss = outputs.loss
>>> logits = outputs.logits


## MobileBertForTokenClassification¶

class transformers.MobileBertForTokenClassification(config)[source]

MobileBert Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForTokenClassification forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• labels (torch.LongTensor of shape (batch_size, sequence_length), optional) – Labels for computing the token classification loss. Indices should be in [0, ..., config.num_labels - 1].

Returns

A TokenClassifierOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) – Classification loss.

• logits (torch.FloatTensor of shape (batch_size, sequence_length, config.num_labels)) – Classification scores (before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TokenClassifierOutput or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertForTokenClassification
>>> import torch

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
>>> labels = torch.tensor([1] * inputs["input_ids"].size(1)).unsqueeze(0)  # Batch size 1

>>> outputs = model(**inputs, labels=labels)
>>> loss = outputs.loss
>>> logits = outputs.logits


class transformers.MobileBertForQuestionAnswering(config)[source]

MobileBert Model with a span classification head on top for extractive question-answering tasks like SQuAD (a linear layers on top of the hidden-states output to compute span start logits and span end logits).

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, start_positions=None, end_positions=None, output_attentions=None, output_hidden_states=None, return_dict=None)[source]

The MobileBertForQuestionAnswering forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using BertTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.__call__() for details.

What are input IDs?

• attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (torch.LongTensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple.

• start_positions (torch.LongTensor of shape (batch_size,), optional) – Labels for position (index) of the start of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss.

• end_positions (torch.LongTensor of shape (batch_size,), optional) – Labels for position (index) of the end of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss.

Returns

A QuestionAnsweringModelOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) – Total span extraction loss is the sum of a Cross-Entropy for the start and end positions.

• start_logits (torch.FloatTensor of shape (batch_size, sequence_length)) – Span-start scores (before SoftMax).

• end_logits (torch.FloatTensor of shape (batch_size, sequence_length)) – Span-end scores (before SoftMax).

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of torch.FloatTensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

QuestionAnsweringModelOutput or tuple(torch.FloatTensor)

Example:

>>> from transformers import MobileBertTokenizer, MobileBertForQuestionAnswering
>>> import torch

>>> question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
>>> inputs = tokenizer(question, text, return_tensors='pt')
>>> start_positions = torch.tensor([1])
>>> end_positions = torch.tensor([3])

>>> outputs = model(**inputs, start_positions=start_positions, end_positions=end_positions)
>>> loss = outputs.loss
>>> start_scores = outputs.start_logits
>>> end_scores = outputs.end_logits


## TFMobileBertModel¶

class transformers.TFMobileBertModel(*args, **kwargs)[source]

The bare MobileBert Model transformer outputting raw hidden-states without any specific head on top.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, training=False, **kwargs)[source]

The TFMobileBertModel forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

Returns

A TFBaseModelOutputWithPooling or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• last_hidden_state (tf.Tensor of shape (batch_size, sequence_length, hidden_size)) – Sequence of hidden-states at the output of the last layer of the model.

• pooler_output (tf.Tensor of shape (batch_size, hidden_size)) – Last layer hidden-state of the first token of the sequence (classification token) further processed by a Linear layer and a Tanh activation function. The Linear layer weights are trained from the next sentence prediction (classification) objective during pretraining.

This output is usually not a good summary of the semantic content of the input, you’re often better with averaging or pooling the sequence of hidden-states for the whole input sequence.

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFBaseModelOutputWithPooling or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertModel
>>> import tensorflow as tf

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
>>> outputs = model(inputs)

>>> last_hidden_states = outputs.last_hidden_state


## TFMobileBertForPreTraining¶

class transformers.TFMobileBertForPreTraining(*args, **kwargs)[source]

MobileBert Model with two heads on top as done during the pretraining: a masked language modeling head and a next sentence prediction (classification) head.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, training=False, **kwargs)[source]

The TFMobileBertForPreTraining forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

Returns

A TFMobileBertForPreTrainingOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• prediction_logits (tf.Tensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• seq_relationship_logits (tf.Tensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Examples:

>>> import tensorflow as tf
>>> from transformers import MobileBertTokenizer, TFMobileBertForPreTraining

>>> input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :]  # Batch size 1
>>> outputs = model(input_ids)
>>> prediction_scores, seq_relationship_scores = outputs[:2]


Return type

TFMobileBertForPreTrainingOutput or tuple(tf.Tensor)

class transformers.TFMobileBertForMaskedLM(*args, **kwargs)[source]

MobileBert Model with a language modeling head on top.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, labels=None, training=False, **kwargs)[source]

The TFMobileBertForMaskedLM forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

• labels (tf.Tensor of shape (batch_size, sequence_length), optional) – Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels

Returns

A TFMaskedLMOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (n,), optional, where n is the number of non-masked labels, returned when labels is provided) – Masked language modeling (MLM) loss.

• logits (tf.Tensor of shape (batch_size, sequence_length, config.vocab_size)) – Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFMaskedLMOutput or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertForMaskedLM
>>> import tensorflow as tf

>>> inputs = tokenizer("The capital of France is [MASK].", return_tensors="tf")
>>> inputs["labels"] = tokenizer("The capital of France is Paris.", return_tensors="tf")["input_ids"]

>>> outputs = model(inputs)
>>> loss = outputs.loss
>>> logits = outputs.logits


## TFMobileBertForNextSentencePrediction¶

class transformers.TFMobileBertForNextSentencePrediction(*args, **kwargs)[source]

MobileBert Model with a next sentence prediction (classification) head on top.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, next_sentence_label=None, training=False, **kwargs)[source]

The TFMobileBertForNextSentencePrediction forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

Returns

A TFNextSentencePredictorOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (n,), optional, where n is the number of non-masked labels, returned when next_sentence_label is provided) – Next sentence prediction loss.

• logits (tf.Tensor of shape (batch_size, 2)) – Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Examples:

>>> import tensorflow as tf
>>> from transformers import MobileBertTokenizer, TFMobileBertForNextSentencePrediction

>>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
>>> next_sentence = "The sky is blue due to the shorter wavelength of blue light."
>>> encoding = tokenizer(prompt, next_sentence, return_tensors='tf')

>>> logits = model(encoding['input_ids'], token_type_ids=encoding['token_type_ids'])[0]


Return type

TFNextSentencePredictorOutput or tuple(tf.Tensor)

## TFMobileBertForSequenceClassification¶

class transformers.TFMobileBertForSequenceClassification(*args, **kwargs)[source]

MobileBert Model transformer with a sequence classification/regression head on top (a linear layer on top of the pooled output) e.g. for GLUE tasks.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, labels=None, training=False, **kwargs)[source]

The TFMobileBertForSequenceClassification forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

• labels (tf.Tensor of shape (batch_size,), optional) – Labels for computing the sequence classification/regression loss. Indices should be in [0, ..., config.num_labels - 1]. If config.num_labels == 1 a regression loss is computed (Mean-Square loss), If config.num_labels > 1 a classification loss is computed (Cross-Entropy).

Returns

A TFSequenceClassifierOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (batch_size, ), optional, returned when labels is provided) – Classification (or regression if config.num_labels==1) loss.

• logits (tf.Tensor of shape (batch_size, config.num_labels)) – Classification (or regression if config.num_labels==1) scores (before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFSequenceClassifierOutput or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertForSequenceClassification
>>> import tensorflow as tf

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
>>> inputs["labels"] = tf.reshape(tf.constant(1), (-1, 1)) # Batch size 1

>>> outputs = model(inputs)
>>> loss = outputs.loss
>>> logits = outputs.logits


## TFMobileBertForMultipleChoice¶

class transformers.TFMobileBertForMultipleChoice(*args, **kwargs)[source]

MobileBert Model with a multiple choice classification head on top (a linear layer on top of the pooled output and a softmax) e.g. for RocStories/SWAG tasks.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, labels=None, training=False, **kwargs)[source]

The TFMobileBertForMultipleChoice forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, num_choices, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, num_choices, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, num_choices, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, num_choices, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, num_choices, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

• labels (tf.Tensor of shape (batch_size,), optional) – Labels for computing the multiple choice classification loss. Indices should be in [0, ..., num_choices] where num_choices is the size of the second dimension of the input tensors. (See input_ids above)

Returns

A TFMultipleChoiceModelOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (batch_size, ), optional, returned when labels is provided) – Classification loss.

• logits (tf.Tensor of shape (batch_size, num_choices)) – num_choices is the second dimension of the input tensors. (see input_ids above).

Classification scores (before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFMultipleChoiceModelOutput or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertForMultipleChoice
>>> import tensorflow as tf

>>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
>>> choice0 = "It is eaten with a fork and a knife."
>>> choice1 = "It is eaten while held in the hand."

>>> encoding = tokenizer([prompt, prompt], [choice0, choice1], return_tensors='tf', padding=True)
>>> inputs = {k: tf.expand_dims(v, 0) for k, v in encoding.items()}
>>> outputs = model(inputs)  # batch size is 1

>>> # the linear classifier still needs to be trained
>>> logits = outputs.logits


## TFMobileBertForTokenClassification¶

class transformers.TFMobileBertForTokenClassification(*args, **kwargs)[source]

MobileBert Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks.

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, labels=None, training=False, **kwargs)[source]

The TFMobileBertForTokenClassification forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

• labels (tf.Tensor of shape (batch_size, sequence_length), optional) – Labels for computing the token classification loss. Indices should be in [0, ..., config.num_labels - 1].

Returns

A TFTokenClassifierOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (n,), optional, where n is the number of unmasked labels, returned when labels is provided) – Classification loss.

• logits (tf.Tensor of shape (batch_size, sequence_length, config.num_labels)) – Classification scores (before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFTokenClassifierOutput or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertForTokenClassification
>>> import tensorflow as tf

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
>>> input_ids = inputs["input_ids"]
>>> inputs["labels"] = tf.reshape(tf.constant([1] * tf.size(input_ids).numpy()), (-1, tf.size(input_ids))) # Batch size 1

>>> outputs = model(inputs)
>>> loss = outputs.loss
>>> logits = outputs.logits


class transformers.TFMobileBertForQuestionAnswering(*args, **kwargs)[source]

MobileBert Model with a span classification head on top for extractive question-answering tasks like SQuAD (a linear layers on top of the hidden-states output to compute span start logits and span end logits).

This model inherits from TFPreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a tf.keras.Model subclass. Use it as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and behavior.

Note

TF 2.0 models accepts two formats as inputs:

• having all inputs as keyword arguments (like PyTorch models), or

• having all inputs as a list, tuple or dict in the first positional arguments.

This second option is useful when using tf.keras.Model.fit() method which currently requires having all the tensors in the first argument of the model call function: model(inputs).

If you choose this second option, there are three possibilities you can use to gather all the input Tensors in the first positional argument :

• a single Tensor with input_ids only and nothing else: model(inputs_ids)

• a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: model([input_ids, attention_mask]) or model([input_ids, attention_mask, token_type_ids])

• a dictionary with one or several input Tensors associated to the input names given in the docstring: model({"input_ids": input_ids, "token_type_ids": token_type_ids})

Parameters

config (MobileBertConfig) – Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

call(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None, start_positions=None, end_positions=None, training=False, **kwargs)[source]

The TFMobileBertForQuestionAnswering forward method, overrides the __call__() special method.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Parameters
• input_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length)) –

Indices of input sequence tokens in the vocabulary.

Indices can be obtained using MobileBertTokenizer. See transformers.PreTrainedTokenizer.__call__() and transformers.PreTrainedTokenizer.encode() for details.

What are input IDs?

• attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:

• 1 for tokens that are not masked,

• 0 for tokens that are masked.

• token_type_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Segment token indices to indicate first and second portions of the inputs. Indices are selected in [0, 1]:

• 0 corresponds to a sentence A token,

• 1 corresponds to a sentence B token.

What are token type IDs?

• position_ids (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional) –

Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.max_position_embeddings - 1].

What are position IDs?

• head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional) –

Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) – Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

• output_attentions (bool, optional) – Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• output_hidden_states (bool, optional) – Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.

• return_dict (bool, optional) – Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.

• training (bool, optional, defaults to False) – Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).

• start_positions (tf.Tensor of shape (batch_size,), optional) – Labels for position (index) of the start of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss.

• end_positions (tf.Tensor of shape (batch_size,), optional) – Labels for position (index) of the end of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss.

Returns

A TFQuestionAnsweringModelOutput or a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (MobileBertConfig) and inputs.

• loss (tf.Tensor of shape (batch_size, ), optional, returned when start_positions and end_positions are provided) – Total span extraction loss is the sum of a Cross-Entropy for the start and end positions.

• start_logits (tf.Tensor of shape (batch_size, sequence_length)) – Span-start scores (before SoftMax).

• end_logits (tf.Tensor of shape (batch_size, sequence_length)) – Span-end scores (before SoftMax).

• hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) – Tuple of tf.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

• attentions (tuple(tf.Tensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) – Tuple of tf.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

Return type

TFQuestionAnsweringModelOutput or tuple(tf.Tensor)

Example:

>>> from transformers import MobileBertTokenizer, TFMobileBertForQuestionAnswering
>>> import tensorflow as tf