ELECTRA

The ELECTRA model was proposed in the paper. ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators. ELECTRA is a new pre-training approach which trains two transformer models: the generator and the discriminator. The generator’s role is to replace tokens in a sequence, and is therefore trained as a masked language model. The discriminator, which is the model we’re interested in, tries to identify which tokens were replaced by the generator in the sequence.

The abstract from the paper is the following:

Masked language modeling (MLM) pre-training methods such as BERT corrupt the input by replacing some tokens with [MASK] and then train a model to reconstruct the original tokens. While they produce good results when transferred to downstream NLP tasks, they generally require large amounts of compute to be effective. As an alternative, we propose a more sample-efficient pre-training task called replaced token detection. Instead of masking the input, our approach corrupts it by replacing some tokens with plausible alternatives sampled from a small generator network. Then, instead of training a model that predicts the original identities of the corrupted tokens, we train a discriminative model that predicts whether each token in the corrupted input was replaced by a generator sample or not. Thorough experiments demonstrate this new pre-training task is more efficient than MLM because the task is defined over all input tokens rather than just the small subset that was masked out. As a result, the contextual representations learned by our approach substantially outperform the ones learned by BERT given the same model size, data, and compute. The gains are particularly strong for small models; for example, we train a model on one GPU for 4 days that outperforms GPT (trained using 30x more compute) on the GLUE natural language understanding benchmark. Our approach also works well at scale, where it performs comparably to RoBERTa and XLNet while using less than 1/4 of their compute and outperforms them when using the same amount of compute.

Tips:

  • ELECTRA is the pre-training approach, therefore there is nearly no changes done to the underlying model: BERT. The only change is the separation of the embedding size and the hidden size -> The embedding size is generally smaller, while the hidden size is larger. An additional projection layer (linear) is used to project the embeddings from their embedding size to the hidden size. In the case where the embedding size is the same as the hidden size, no projection layer is used.

  • The ELECTRA checkpoints saved using Google Research’s implementation contain both the generator and discriminator. The conversion script requires the user to name which model to export into the correct architecture. Once converted to the HuggingFace format, these checkpoints may be loaded into all available ELECTRA models, however. This means that the discriminator may be loaded in the ElectraForMaskedLM model, and the generator may be loaded in the ElectraForPreTraining model (the classification head will be randomly initialized as it doesn’t exist in the generator).

The original code can be found here.

ElectraConfig

class transformers.ElectraConfig(vocab_size=30522, embedding_size=128, hidden_size=256, num_hidden_layers=12, num_attention_heads=4, intermediate_size=1024, hidden_act='gelu', hidden_dropout_prob=0.1, 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, **kwargs)[source]

This is the configuration class to store the configuration of a ElectraModel. It is used to instantiate an ELECTRA model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the ELECTRA google/electra-small-discriminator 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 ELECTRA model. Defines the different tokens that can be represented by the inputs_ids passed to the forward method of ElectraModel.

  • embedding_size (int, optional, defaults to 128) – Dimensionality of the encoder layers and the pooler layer.

  • hidden_size (int, optional, defaults to 256) – Dimensionality of the encoder layers and the pooler layer.

  • num_hidden_layers (int, optional, defaults to 12) – 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 1024) – Dimensionality of the “intermediate” (i.e., feed-forward) layer in the Transformer encoder.

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

  • hidden_dropout_prob (float, optional, defaults to 0.1) – The dropout probabilitiy 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 into ElectraModel.

  • 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.

Example:

from transformers import ElectraModel, ElectraConfig

# Initializing a ELECTRA electra-base-uncased style configuration
configuration = ElectraConfig()

# Initializing a model from the electra-base-uncased style configuration
model = ElectraModel(configuration)

# Accessing the model configuration
configuration = model.config
pretrained_config_archive_map

A dictionary containing all the available pre-trained checkpoints.

Type

Dict[str, str]

ElectraTokenizer

class transformers.ElectraTokenizer(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, **kwargs)[source]

Constructs an Electra tokenizer. ElectraTokenizer is identical to BertTokenizer and runs end-to-end tokenization: punctuation splitting + wordpiece.

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

ElectraTokenizerFast

class transformers.ElectraTokenizerFast(vocab_file, do_lower_case=True, unk_token='[UNK]', sep_token='[SEP]', pad_token='[PAD]', cls_token='[CLS]', mask_token='[MASK]', clean_text=True, tokenize_chinese_chars=True, strip_accents=True, wordpieces_prefix='##', **kwargs)[source]

Constructs a “Fast” Electra Fast tokenizer (backed by HuggingFace’s tokenizers library).

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

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

ElectraModel

class transformers.ElectraModel(config)[source]

The bare Electra Model transformer outputting raw hidden-states without any specific head on top. Identical to the BERT model except that it uses an additional linear layer between the embedding layer and the encoder if the hidden size and embedding size are different.Both the generator and discriminator checkpoints may be loaded into this model. This model is a PyTorch torch.nn.Module sub-class. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (ElectraConfig) – 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.

config_class

alias of transformers.configuration_electra.ElectraConfig

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

The ElectraModel 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    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, defaults to None) –

    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, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – 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.

  • encoder_hidden_states (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder.

  • encoder_attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) – Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

Returns

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.

hidden_states (tuple(torch.FloatTensor), optional, returned 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 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

tuple(torch.FloatTensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

from transformers import ElectraModel, ElectraTokenizer
import torch

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = ElectraModel.from_pretrained('google/electra-small-discriminator')

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

last_hidden_states = outputs[0]  # The last hidden-state is the first element of the output tuple
get_input_embeddings()[source]

Returns the model’s input embeddings.

Returns

A torch module mapping vocabulary to hidden states.

Return type

nn.Module

set_input_embeddings(value)[source]

Set model’s input embeddings

Parameters

value (nn.Module) – A module mapping vocabulary to hidden states.

ElectraForPreTraining

class transformers.ElectraForPreTraining(config)[source]

Electra model with a binary classification head on top as used during pre-training for identifying generated tokens.

It is recommended to load the discriminator checkpoint into that model. This model is a PyTorch torch.nn.Module sub-class. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (ElectraConfig) – 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)[source]

The ElectraForPreTraining 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    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, defaults to None) –

    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, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – 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.

  • encoder_hidden_states (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder.

  • encoder_attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) – Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

  • labels (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) – Labels for computing the ELECTRA loss. Input should be a sequence of tokens (see input_ids docstring) Indices should be in [0, 1]. 0 indicates the token is an original token, 1 indicates the token was replaced.

Returns

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

Total loss of the ELECTRA objective.

scores (torch.FloatTensor of shape (batch_size, sequence_length))

Prediction scores of the head (scores for each token before SoftMax).

hidden_states (tuple(torch.FloatTensor), optional, returned 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 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

tuple(torch.FloatTensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

from transformers import ElectraTokenizer, ElectraForPreTraining
import torch

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = ElectraForPreTraining.from_pretrained('google/electra-small-discriminator')

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_scores, seq_relationship_scores = outputs[:2]

ElectraForMaskedLM

class transformers.ElectraForMaskedLM(config)[source]

Electra model with a language modeling head on top.

Even though both the discriminator and generator may be loaded into this model, the generator is the only model of the two to have been trained for the masked language modeling task. This model is a PyTorch torch.nn.Module sub-class. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (ElectraConfig) – 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, masked_lm_labels=None)[source]

The ElectraForMaskedLM 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    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, defaults to None) –

    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, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – 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.

  • encoder_hidden_states (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder.

  • encoder_attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) – Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

  • masked_lm_labels (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) – 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

masked_lm_loss (optional, returned when masked_lm_labels is provided) torch.FloatTensor of shape (1,):

Masked language modeling loss.

prediction_scores (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 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 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 ElectraTokenizer, ElectraForMaskedLM
import torch

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-generator')
model = ElectraForMaskedLM.from_pretrained('google/electra-small-generator')

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

loss, prediction_scores = outputs[:2]

Return type

tuple(torch.FloatTensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

get_output_embeddings()[source]

Returns the model’s output embeddings.

Returns

A torch module mapping hidden states to vocabulary.

Return type

nn.Module

ElectraForTokenClassification

class transformers.ElectraForTokenClassification(config)[source]

Electra model with a token classification head on top.

Both the discriminator and generator may be loaded into this model. This model is a PyTorch torch.nn.Module sub-class. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

Parameters

config (ElectraConfig) – 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)[source]

The ElectraForTokenClassification 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • token_type_ids (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) –

    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, defaults to None) –

    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, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – 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.

  • encoder_hidden_states (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional, defaults to None) – Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder.

  • encoder_attention_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional, defaults to None) – Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

  • labels (torch.LongTensor of shape (batch_size, sequence_length), optional, defaults to None) – Labels for computing the token classification loss. Indices should be in [0, ..., config.num_labels - 1].

Returns

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

Classification loss.

scores (torch.FloatTensor of shape (batch_size, sequence_length, config.num_labels))

Classification scores (before SoftMax).

hidden_states (tuple(torch.FloatTensor), optional, returned 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 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

tuple(torch.FloatTensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

from transformers import ElectraTokenizer, ElectraForTokenClassification
import torch

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = ElectraForTokenClassification.from_pretrained('google/electra-small-discriminator')

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

loss, scores = outputs[:2]

TFElectraModel

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

The bare Electra Model transformer outputting raw hidden-states without any specific head on top. Identical to the BERT model except that it uses an additional linear layer between the embedding layer and the encoder if the hidden size and embedding size are different.Both the generator and discriminator checkpoints may be loaded into this model. This model is a tf.keras.Model sub-class. 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 (ElectraConfig) – 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(inputs, **kwargs)[source]

The TFElectraModel 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (Numpy array or tf.Tensor of shape (batch_size, sequence_length, embedding_dim), optional, defaults to None) – 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.

  • training (boolean, optional, defaults to False) – Whether to activate dropout modules (if set to True) during training or to de-activate them (if set to False) for evaluation.

Returns

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.

hidden_states (tuple(tf.Tensor), optional, returned 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 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

tuple(tf.Tensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

import tensorflow as tf
from transformers import ElectraTokenizer, TFElectraModel

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = TFElectraModel.from_pretrained('google/electra-small-discriminator')
input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :]  # Batch size 1
outputs = model(input_ids)
last_hidden_states = outputs[0]  # The last hidden-state is the first element of the output tuple
get_input_embeddings()[source]

Returns the model’s input embeddings.

Returns

A torch module mapping vocabulary to hidden states.

Return type

tf.keras.layers.Layer

TFElectraForPreTraining

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

Electra model with a binary classification head on top as used during pre-training for identifying generated tokens.

Even though both the discriminator and generator may be loaded into this model, the discriminator is the only model of the two to have the correct classification head to be used for this model.

This model is a tf.keras.Model sub-class. 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 (ElectraConfig): 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, training=False)[source]

The TFElectraForPreTraining 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (Numpy array or tf.Tensor of shape (batch_size, sequence_length, embedding_dim), optional, defaults to None) – 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.

  • training (boolean, optional, defaults to False) – Whether to activate dropout modules (if set to True) during training or to de-activate them (if set to False) for evaluation.

Returns

scores (Numpy array or tf.Tensor of shape (batch_size, sequence_length, config.num_labels)):

Prediction scores of the head (scores for each token before SoftMax).

hidden_states (tuple(tf.Tensor), optional, returned 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 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

tuple(tf.Tensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

import tensorflow as tf
from transformers import ElectraTokenizer, TFElectraForPreTraining

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = TFElectraForPreTraining.from_pretrained('google/electra-small-discriminator')
input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :]  # Batch size 1
outputs = model(input_ids)
scores = outputs[0]
get_input_embeddings()[source]

Returns the model’s input embeddings.

Returns

A torch module mapping vocabulary to hidden states.

Return type

tf.keras.layers.Layer

TFElectraForMaskedLM

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

Electra model with a language modeling head on top.

Even though both the discriminator and generator may be loaded into this model, the generator is the only model of the two to have been trained for the masked language modeling task.

This model is a tf.keras.Model sub-class. 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 (ElectraConfig): 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, training=False)[source]

The TFElectraForMaskedLM 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (Numpy array or tf.Tensor of shape (batch_size, sequence_length, embedding_dim), optional, defaults to None) – 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.

  • training (boolean, optional, defaults to False) – Whether to activate dropout modules (if set to True) during training or to de-activate them (if set to False) for evaluation.

Returns

prediction_scores (Numpy array or 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 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 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

tuple(tf.Tensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

import tensorflow as tf
from transformers import ElectraTokenizer, TFElectraForMaskedLM

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-generator')
model = TFElectraForMaskedLM.from_pretrained('google/electra-small-generator')
input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :]  # Batch size 1
outputs = model(input_ids)
prediction_scores = outputs[0]
get_input_embeddings()[source]

Returns the model’s input embeddings.

Returns

A torch module mapping vocabulary to hidden states.

Return type

tf.keras.layers.Layer

get_output_embeddings()[source]

Returns the model’s output embeddings.

Returns

A torch module mapping hidden states to vocabulary.

Return type

tf.keras.layers.Layer

TFElectraForTokenClassification

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

Electra model with a token classification head on top.

Both the discriminator and generator may be loaded into this model.

This model is a tf.keras.Model sub-class. 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 (ElectraConfig): 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, training=False)[source]

The TFElectraForTokenClassification 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 transformers.ElectraTokenizer. See transformers.PreTrainedTokenizer.encode() and transformers.PreTrainedTokenizer.encode_plus() for details.

    What are input IDs?

  • attention_mask (Numpy array or tf.Tensor of shape (batch_size, sequence_length), optional, defaults to None) –

    Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]: 1 for tokens that are NOT MASKED, 0 for MASKED tokens.

    What are attention masks?

  • head_mask (Numpy array or tf.Tensor of shape (num_heads,) or (num_layers, num_heads), optional, defaults to None) – Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]: 1 indicates the head is not masked, 0 indicates the head is masked.

  • inputs_embeds (Numpy array or tf.Tensor of shape (batch_size, sequence_length, embedding_dim), optional, defaults to None) – 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.

  • training (boolean, optional, defaults to False) – Whether to activate dropout modules (if set to True) during training or to de-activate them (if set to False) for evaluation.

Returns

scores (Numpy array or tf.Tensor of shape (batch_size, sequence_length, config.num_labels)):

Classification scores (before SoftMax).

hidden_states (tuple(tf.Tensor), optional, returned 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 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

tuple(tf.Tensor) comprising various elements depending on the configuration (ElectraConfig) and inputs

Examples:

import tensorflow as tf
from transformers import ElectraTokenizer, TFElectraForTokenClassification

tokenizer = ElectraTokenizer.from_pretrained('google/electra-small-discriminator')
model = TFElectraForTokenClassification.from_pretrained('google/electra-small-discriminator')
input_ids = tf.constant(tokenizer.encode("Hello, my dog is cute"))[None, :]  # Batch size 1
outputs = model(input_ids)
scores = outputs[0]