Transformers documentation

XLM

Join the Hugging Face community

to get started

# XLM

## Overview

The XLM model was proposed in Cross-lingual Language Model Pretraining by Guillaume Lample, Alexis Conneau. It’s a transformer pretrained using one of the following objectives:

• a causal language modeling (CLM) objective (next token prediction),
• a masked language modeling (MLM) objective (BERT-like), or
• a Translation Language Modeling (TLM) object (extension of BERT’s MLM to multiple language inputs)

The abstract from the paper is the following:

Recent studies have demonstrated the efficiency of generative pretraining for English natural language understanding. In this work, we extend this approach to multiple languages and show the effectiveness of cross-lingual pretraining. We propose two methods to learn cross-lingual language models (XLMs): one unsupervised that only relies on monolingual data, and one supervised that leverages parallel data with a new cross-lingual language model objective. We obtain state-of-the-art results on cross-lingual classification, unsupervised and supervised machine translation. On XNLI, our approach pushes the state of the art by an absolute gain of 4.9% accuracy. On unsupervised machine translation, we obtain 34.3 BLEU on WMT’16 German-English, improving the previous state of the art by more than 9 BLEU. On supervised machine translation, we obtain a new state of the art of 38.5 BLEU on WMT’16 Romanian-English, outperforming the previous best approach by more than 4 BLEU. Our code and pretrained models will be made publicly available.

Tips:

• XLM has many different checkpoints, which were trained using different objectives: CLM, MLM or TLM. Make sure to select the correct objective for your task (e.g. MLM checkpoints are not suitable for generation).
• XLM has multilingual checkpoints which leverage a specific lang parameter. Check out the multi-lingual page for more information.

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

## XLMConfig

### class transformers.XLMConfig

< >

( vocab_size = 30145 emb_dim = 2048 n_layers = 12 n_heads = 16 dropout = 0.1 attention_dropout = 0.1 gelu_activation = True sinusoidal_embeddings = False causal = False asm = False n_langs = 1 use_lang_emb = True max_position_embeddings = 512 embed_init_std = 0.02209708691207961 layer_norm_eps = 1e-12 init_std = 0.02 bos_index = 0 eos_index = 1 pad_index = 2 unk_index = 3 mask_index = 5 is_encoder = True summary_type = 'first' summary_use_proj = True summary_activation = None summary_proj_to_labels = True summary_first_dropout = 0.1 start_n_top = 5 end_n_top = 5 mask_token_id = 0 lang_id = 0 pad_token_id = 2 bos_token_id = 0 **kwargs )

Parameters

• vocab_size (int, optional, defaults to 30145) — Vocabulary size of the BERT model. Defines the number of different tokens that can be represented by the inputs_ids passed when calling XLMModel or TFXLMModel.
• emb_dim (int, optional, defaults to 2048) — Dimensionality of the encoder layers and the pooler layer.
• n_layer (int, optional, defaults to 12) — Number of hidden layers in the Transformer encoder.
• n_head (int, optional, defaults to 16) — Number of attention heads for each attention layer in the Transformer encoder.
• dropout (float, optional, defaults to 0.1) — The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
• attention_dropout (float, optional, defaults to 0.1) — The dropout probability for the attention mechanism
• gelu_activation (bool, optional, defaults to True) — Whether or not to use gelu for the activations instead of relu.
• sinusoidal_embeddings (bool, optional, defaults to False) — Whether or not to use sinusoidal positional embeddings instead of absolute positional embeddings.
• causal (bool, optional, defaults to False) — Whether or not the model should behave in a causal manner. Causal models use a triangular attention mask in order to only attend to the left-side context instead if a bidirectional context.
• asm (bool, optional, defaults to False) — Whether or not to use an adaptive log softmax projection layer instead of a linear layer for the prediction layer.
• n_langs (int, optional, defaults to 1) — The number of languages the model handles. Set to 1 for monolingual models.
• use_lang_emb (bool, optional, defaults to True) — Whether to use language embeddings. Some models use additional language embeddings, see the multilingual models page for information on how to use them.
• 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).
• embed_init_std (float, optional, defaults to 2048^-0.5) — The standard deviation of the truncated_normal_initializer for initializing the embedding matrices.
• init_std (int, optional, defaults to 50257) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices except the embedding matrices.
• layer_norm_eps (float, optional, defaults to 1e-12) — The epsilon used by the layer normalization layers.
• bos_index (int, optional, defaults to 0) — The index of the beginning of sentence token in the vocabulary.
• eos_index (int, optional, defaults to 1) — The index of the end of sentence token in the vocabulary.
• pad_index (int, optional, defaults to 2) — The index of the padding token in the vocabulary.
• unk_index (int, optional, defaults to 3) — The index of the unknown token in the vocabulary.
• mask_index (int, optional, defaults to 5) — The index of the masking token in the vocabulary.
• is_encoder(bool, optional, defaults to True) — Whether or not the initialized model should be a transformer encoder or decoder as seen in Vaswani et al.
• summary_type (string, optional, defaults to “first”) — Argument used when doing sequence summary. Used in the sequence classification and multiple choice models.

Has to be one of the following options:

• "last": Take the last token hidden state (like XLNet).
• "first": Take the first token hidden state (like BERT).
• "mean": Take the mean of all tokens hidden states.
• "cls_index": Supply a Tensor of classification token position (like GPT/GPT-2).
• "attn": Not implemented now, use multi-head attention.
• summary_use_proj (bool, optional, defaults to True) — Argument used when doing sequence summary. Used in the sequence classification and multiple choice models.

Whether or not to add a projection after the vector extraction.

• summary_activation (str, optional) — Argument used when doing sequence summary. Used in the sequence classification and multiple choice models.

Pass "tanh" for a tanh activation to the output, any other value will result in no activation.

• summary_proj_to_labels (bool, optional, defaults to True) — Used in the sequence classification and multiple choice models.

Whether the projection outputs should have config.num_labels or config.hidden_size classes.

• summary_first_dropout (float, optional, defaults to 0.1) — Used in the sequence classification and multiple choice models.

The dropout ratio to be used after the projection and activation.

• start_n_top (int, optional, defaults to 5) — Used in the SQuAD evaluation script.
• end_n_top (int, optional, defaults to 5) — Used in the SQuAD evaluation script.
• mask_token_id (int, optional, defaults to 0) — Model agnostic parameter to identify masked tokens when generating text in an MLM context.
• lang_id (int, optional, defaults to 1) — The ID of the language used by the model. This parameter is used when generating text in a given language.

This is the configuration class to store the configuration of a XLMModel or a TFXLMModel. It is used to instantiate a XLM 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 xlm-mlm-en-2048 architecture.

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

Examples:

>>> from transformers import XLMConfig, XLMModel

>>> # Initializing a XLM configuration
>>> configuration = XLMConfig()

>>> # Initializing a model from the configuration
>>> model = XLMModel(configuration)

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

## XLMTokenizer

### class transformers.XLMTokenizer

< >

( vocab_file merges_file unk_token = '<unk>' bos_token = '<s>' sep_token = '</s>' pad_token = '<pad>' cls_token = '</s>' mask_token = '<special1>' additional_special_tokens = ['<special0>', '<special1>', '<special2>', '<special3>', '<special4>', '<special5>', '<special6>', '<special7>', '<special8>', '<special9>'] lang2id = None id2lang = None do_lowercase_and_remove_accent = True **kwargs )

Parameters

• vocab_file (str) — Vocabulary file.
• merges_file (str) — Merges file.
• unk_token (str, optional, defaults to "<unk>") — The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this token instead.
• bos_token (str, optional, defaults to "<s>") — The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.

When building a sequence using special tokens, this is not the token that is used for the beginning of sequence. The token used is the cls_token.

• sep_token (str, optional, defaults to "</s>") — The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for sequence classification or for a text and a question for question answering. It is also used as the last token of a sequence built with special tokens.
• pad_token (str, optional, defaults to "<pad>") — The token used for padding, for example when batching sequences of different lengths.
• cls_token (str, optional, defaults to "</s>") — The classifier token which is used when doing sequence classification (classification of the whole sequence instead of per-token classification). It is the first token of the sequence when built with special tokens.
• mask_token (str, optional, defaults to "<special1>") — The token used for masking values. This is the token used when training this model with masked language modeling. This is the token which the model will try to predict.
• additional_special_tokens (List[str], optional, defaults to ["<special0>","<special1>","<special2>","<special3>","<special4>","<special5>","<special6>","<special7>","<special8>","<special9>"]) — List of additional special tokens.
• lang2id (Dict[str, int], optional) — Dictionary mapping languages string identifiers to their IDs.
• id2lang (Dict[int, str], optional) — Dictionary mapping language IDs to their string identifiers.
• do_lowercase_and_remove_accent (bool, optional, defaults to True) — Whether to lowercase and remove accents when tokenizing.

Construct an XLM tokenizer. Based on Byte-Pair Encoding. The tokenization process is the following:

• Moses preprocessing and tokenization for most supported languages.
• Language specific tokenization for Chinese (Jieba), Japanese (KyTea) and Thai (PyThaiNLP).
• Optionally lowercases and normalizes all inputs text.
• The arguments special_tokens and the function set_special_tokens, can be used to add additional symbols (like ”classify”) to a vocabulary.
• The lang2id attribute maps the languages supported by the model with their IDs if provided (automatically set for pretrained vocabularies).
• The id2lang attributes does reverse mapping if provided (automatically set for pretrained vocabularies).

This tokenizer inherits from PreTrainedTokenizer which contains most of the main methods. Users should refer to this superclass for more information regarding those methods.

#### build_inputs_with_special_tokens

< >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None ) List[int]

Parameters

• token_ids_0 (List[int]) — List of IDs to which the special tokens will be added.
• token_ids_1 (List[int], optional) — Optional second list of IDs for sequence pairs.

Returns

List[int]

List of input IDs with the appropriate special tokens.

Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. An XLM sequence has the following format:

• single sequence: <s> X </s>
• pair of sequences: <s> A </s> B </s>

< >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None already_has_special_tokens: bool = False ) List[int]

Parameters

• token_ids_0 (List[int]) — List of IDs.
• token_ids_1 (List[int], optional) — Optional second list of IDs for sequence pairs.
• already_has_special_tokens (bool, optional, defaults to False) — Whether or not the token list is already formatted with special tokens for the model.

Returns

List[int]

A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.

Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding special tokens using the tokenizer prepare_for_model method.

#### create_token_type_ids_from_sequences

< >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None ) List[int]

Parameters

• token_ids_0 (List[int]) — List of IDs.
• token_ids_1 (List[int], optional) — Optional second list of IDs for sequence pairs.

Returns

List[int]

List of token type IDs according to the given sequence(s).

Create a mask from the two sequences passed to be used in a sequence-pair classification task. An XLM sequence pair mask has the following format:

0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
| first sequence    | second sequence |

If token_ids_1 is None, this method only returns the first portion of the mask (0s).

#### save_vocabulary

< >

( save_directory: str filename_prefix: typing.Optional[str] = None )

## XLM specific outputs

< >

( loss: typing.Optional[torch.FloatTensor] = None start_top_log_probs: typing.Optional[torch.FloatTensor] = None start_top_index: typing.Optional[torch.LongTensor] = None end_top_log_probs: typing.Optional[torch.FloatTensor] = None end_top_index: typing.Optional[torch.LongTensor] = None cls_logits: typing.Optional[torch.FloatTensor] = None hidden_states: typing.Optional[typing.Tuple[torch.FloatTensor]] = None attentions: typing.Optional[typing.Tuple[torch.FloatTensor]] = None )

Parameters

• loss (torch.FloatTensor of shape (1,), optional, returned if both start_positions and end_positions are provided) — Classification loss as the sum of start token, end token (and is_impossible if provided) classification losses.
• start_top_log_probs (torch.FloatTensor of shape (batch_size, config.start_n_top), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the top config.start_n_top start token possibilities (beam-search).
• start_top_index (torch.LongTensor of shape (batch_size, config.start_n_top), optional, returned if start_positions or end_positions is not provided) — Indices for the top config.start_n_top start token possibilities (beam-search).
• end_top_log_probs (torch.FloatTensor of shape (batch_size, config.start_n_top * config.end_n_top), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the top config.start_n_top * config.end_n_top end token possibilities (beam-search).
• end_top_index (torch.LongTensor of shape (batch_size, config.start_n_top * config.end_n_top), optional, returned if start_positions or end_positions is not provided) — Indices for the top config.start_n_top * config.end_n_top end token possibilities (beam-search).
• cls_logits (torch.FloatTensor of shape (batch_size,), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the is_impossible label of the answers.
• 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.

Base class for outputs of question answering models using a SquadHead.

## XLMModel

### class transformers.XLMModel

< >

( config )

Parameters

• config (XLMConfig) — 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.

The bare XLM 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.BaseModelOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_outputs.BaseModelOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.BaseModelOutput 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 (XLMConfig) 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.

• 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.

The XLMModel forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMModel
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = XLMModel.from_pretrained("xlm-mlm-en-2048")

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

>>> last_hidden_states = outputs.last_hidden_state

< >

( config )

Parameters

• config (XLMConfig) — 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.

The XLM Model transformer with a language modeling head on top (linear layer with weights tied to the input embeddings).

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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.MaskedLMOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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 language modeling. Note that the labels are shifted inside the model, i.e. you can set labels = input_ids Indices are selected in [-100, 0, ..., config.vocab_size] All labels set to -100 are ignored (masked), the loss is only computed for labels in [0, ..., config.vocab_size]

Returns

transformers.modeling_outputs.MaskedLMOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.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 (XLMConfig) 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.

The XLMWithLMHeadModel forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMWithLMHeadModel
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")

>>> inputs = tokenizer("The capital of France is <special1>.", 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

## XLMForSequenceClassification

### class transformers.XLMForSequenceClassification

< >

( config )

Parameters

• config (XLMConfig) — 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.

XLM Model 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.SequenceClassifierOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_outputs.SequenceClassifierOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.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 (XLMConfig) 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.

The XLMForSequenceClassification forward method, overrides the __call__ special method.

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.

Example of single-label classification:

>>> from transformers import XLMTokenizer, XLMForSequenceClassification
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = XLMForSequenceClassification.from_pretrained("xlm-mlm-en-2048")

>>> 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

Example of multi-label classification:

>>> from transformers import XLMTokenizer, XLMForSequenceClassification
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = XLMForSequenceClassification.from_pretrained("xlm-mlm-en-2048", problem_type="multi_label_classification")

>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
>>> labels = torch.tensor([[1, 1]], dtype=torch.float)  # need dtype=float for BCEWithLogitsLoss
>>> outputs = model(**inputs, labels=labels)
>>> loss = outputs.loss
>>> logits = outputs.logits

## XLMForMultipleChoice

### class transformers.XLMForMultipleChoice

< >

( config *inputs **kwargs )

Parameters

• config (XLMConfig) — 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.

XLM 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.MultipleChoiceModelOutput or tuple(torch.FloatTensor)

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 XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, num_choices, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_outputs.MultipleChoiceModelOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.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 (XLMConfig) 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.

The XLMForMultipleChoice forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMForMultipleChoice
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = XLMForMultipleChoice.from_pretrained("xlm-mlm-en-2048")

>>> 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

## XLMForTokenClassification

### class transformers.XLMForTokenClassification

< >

( config )

Parameters

• config (XLMConfig) — 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.

XLM 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.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 (XLMConfig) 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.

The XLMForTokenClassification forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMForTokenClassification
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = XLMForTokenClassification.from_pretrained("xlm-mlm-en-2048")

>>> 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

< >

( config )

Parameters

• config (XLMConfig) — 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.

XLM 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None start_positions = None end_positions = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.QuestionAnsweringModelOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_outputs.QuestionAnsweringModelOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.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 (XLMConfig) 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.

The XLMForQuestionAnsweringSimple forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMForQuestionAnsweringSimple
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")

>>> 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

< >

( config )

Parameters

• config (XLMConfig) — 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.

XLM Model with a beam-search 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.

#### forward

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None start_positions = None end_positions = None is_impossible = None cls_index = None p_mask = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.models.xlm.modeling_xlm.XLMForQuestionAnsweringOutput or tuple(torch.FloatTensor)

Parameters

• input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.

Indices can be obtained using XLMTokenizer. See PreTrainedTokenizer.encode() and 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.

• langs (torch.LongTensor of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (torch.LongTensor of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, torch.FloatTensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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.
• is_impossible (torch.LongTensor of shape (batch_size,), optional) — Labels whether a question has an answer or no answer (SQuAD 2.0)
• cls_index (torch.LongTensor of shape (batch_size,), optional) — Labels for position (index) of the classification token to use as input for computing plausibility of the answer.
• p_mask (torch.FloatTensor of shape (batch_size, sequence_length), optional) — Optional mask of tokens which can’t be in answers (e.g. [CLS], [PAD], …). 1.0 means token should be masked. 0.0 mean token is not masked.

Returns

transformers.models.xlm.modeling_xlm.XLMForQuestionAnsweringOutput or tuple(torch.FloatTensor)

A transformers.models.xlm.modeling_xlm.XLMForQuestionAnsweringOutput 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 (XLMConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned if both start_positions and end_positions are provided) — Classification loss as the sum of start token, end token (and is_impossible if provided) classification losses.

• start_top_log_probs (torch.FloatTensor of shape (batch_size, config.start_n_top), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the top config.start_n_top start token possibilities (beam-search).

• start_top_index (torch.LongTensor of shape (batch_size, config.start_n_top), optional, returned if start_positions or end_positions is not provided) — Indices for the top config.start_n_top start token possibilities (beam-search).

• end_top_log_probs (torch.FloatTensor of shape (batch_size, config.start_n_top * config.end_n_top), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the top config.start_n_top * config.end_n_top end token possibilities (beam-search).

• end_top_index (torch.LongTensor of shape (batch_size, config.start_n_top * config.end_n_top), optional, returned if start_positions or end_positions is not provided) — Indices for the top config.start_n_top * config.end_n_top end token possibilities (beam-search).

• cls_logits (torch.FloatTensor of shape (batch_size,), optional, returned if start_positions or end_positions is not provided) — Log probabilities for the is_impossible label of the answers.

• 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.

The XLMForQuestionAnswering forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, XLMForQuestionAnswering
>>> import torch

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")

>>> input_ids = torch.tensor(tokenizer.encode("Hello, my dog is cute", add_special_tokens=True)).unsqueeze(
...     0
>>> )  # Batch size 1
>>> start_positions = torch.tensor([1])
>>> end_positions = torch.tensor([3])

>>> outputs = model(input_ids, start_positions=start_positions, end_positions=end_positions)
>>> loss = outputs.loss

## TFXLMModel

### class transformers.TFXLMModel

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

The bare XLM 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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None training = False **kwargs ) transformers.modeling_tf_outputs.TFBaseModelOutput or tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_tf_outputs.TFBaseModelOutput or tuple(tf.Tensor)

A transformers.modeling_tf_outputs.TFBaseModelOutput 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 (XLMConfig) 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.

• hidden_states (tuple(tf.FloatTensor), 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.

The TFXLMModel forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMModel
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = TFXLMModel.from_pretrained("xlm-mlm-en-2048")

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

>>> last_hidden_states = outputs.last_hidden_state

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

The XLM Model transformer with a language modeling head on top (linear layer with weights tied to the input embeddings).

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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None training = False **kwargs ) transformers.models.xlm.modeling_tf_xlm.TFXLMWithLMHeadModelOutputor tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.models.xlm.modeling_tf_xlm.TFXLMWithLMHeadModelOutputor tuple(tf.Tensor)

A transformers.models.xlm.modeling_tf_xlm.TFXLMWithLMHeadModelOutputor a tuple of tf.Tensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (XLMConfig) and inputs.

• 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.

The TFXLMWithLMHeadModel forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMWithLMHeadModel
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")

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

## TFXLMForSequenceClassification

### class transformers.TFXLMForSequenceClassification

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

XLM Model 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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) transformers.modeling_tf_outputs.TFSequenceClassifierOutput or tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_tf_outputs.TFSequenceClassifierOutput or tuple(tf.Tensor)

A transformers.modeling_tf_outputs.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 (XLMConfig) 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.

The TFXLMForSequenceClassification forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMForSequenceClassification
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = TFXLMForSequenceClassification.from_pretrained("xlm-mlm-en-2048")

>>> 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

## TFXLMForMultipleChoice

### class transformers.TFXLMForMultipleChoice

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

XLM 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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) transformers.modeling_tf_outputs.TFMultipleChoiceModelOutput or tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, num_choices, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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).

Returns

transformers.modeling_tf_outputs.TFMultipleChoiceModelOutput or tuple(tf.Tensor)

A transformers.modeling_tf_outputs.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 (XLMConfig) 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.

The TFXLMForMultipleChoice forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMForMultipleChoice
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = TFXLMForMultipleChoice.from_pretrained("xlm-mlm-en-2048")

>>> 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

## TFXLMForTokenClassification

### class transformers.TFXLMForTokenClassification

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

XLM 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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) transformers.modeling_tf_outputs.TFTokenClassifierOutput or tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_tf_outputs.TFTokenClassifierOutput or tuple(tf.Tensor)

A transformers.modeling_tf_outputs.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 (XLMConfig) 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.

The TFXLMForTokenClassification forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMForTokenClassification
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> model = TFXLMForTokenClassification.from_pretrained("xlm-mlm-en-2048")

>>> 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

< >

( *args **kwargs )

Parameters

• config (XLMConfig) — 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.

XLM Model with a span classification head on top for extractive question-answering tasks like SQuAD (a linear layer 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.

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})

#### call

< >

( input_ids = None attention_mask = None langs = None token_type_ids = None position_ids = None lengths = None cache = 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 ) transformers.modeling_tf_outputs.TFQuestionAnsweringModelOutput or tuple(tf.Tensor)

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 BertTokenizer. See PreTrainedTokenizer.call() and 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.

• langs (tf.Tensor or Numpy array of shape (batch_size, sequence_length), optional) — A parallel sequence of tokens to be used to indicate the language of each token in the input. Indices are languages ids which can be obtained from the language names by using two conversion mappings provided in the configuration of the model (only provided for multilingual models). More precisely, the language name to language id mapping is in model.config.lang2id (which is a dictionary string to int) and the language id to language name mapping is in model.config.id2lang (dictionary int to string).

See usage examples detailed in the multilingual documentation.

• 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?

• lengths (tf.Tensor or Numpy array of shape (batch_size,), optional) — Length of each sentence that can be used to avoid performing attention on padding token indices. You can also use attention_mask for the same result (see above), kept here for compatibility. Indices selected in [0, ..., input_ids.size(-1)].
• cache (Dict[str, tf.Tensor], optional) — Dictionary string to torch.FloatTensor that contains precomputed hidden states (key and values in the attention blocks) as computed by the model (see cache output below). Can be used to speed up sequential decoding.

The dictionary object will be modified in-place during the forward pass to add newly computed hidden-states.

• 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

transformers.modeling_tf_outputs.TFQuestionAnsweringModelOutput or tuple(tf.Tensor)

A transformers.modeling_tf_outputs.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 (XLMConfig) 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.

The TFXLMForQuestionAnsweringSimple forward method, overrides the __call__ special method.

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.

Example:

>>> from transformers import XLMTokenizer, TFXLMForQuestionAnsweringSimple
>>> import tensorflow as tf

>>> tokenizer = XLMTokenizer.from_pretrained("xlm-mlm-en-2048")
>>> answer = " ".join(all_tokens[tf.math.argmax(start_logits, 1)[0] : tf.math.argmax(end_logits, 1)[0] + 1])