Transformers documentation

Funnel Transformer

Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

Funnel Transformer

Overview

The Funnel Transformer model was proposed in the paper Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing. It is a bidirectional transformer model, like BERT, but with a pooling operation after each block of layers, a bit like in traditional convolutional neural networks (CNN) in computer vision.

The abstract from the paper is the following:

With the success of language pretraining, it is highly desirable to develop more efficient architectures of good scalability that can exploit the abundant unlabeled data at a lower cost. To improve the efficiency, we examine the much-overlooked redundancy in maintaining a full-length token-level presentation, especially for tasks that only require a single-vector presentation of the sequence. With this intuition, we propose Funnel-Transformer which gradually compresses the sequence of hidden states to a shorter one and hence reduces the computation cost. More importantly, by re-investing the saved FLOPs from length reduction in constructing a deeper or wider model, we further improve the model capacity. In addition, to perform token-level predictions as required by common pretraining objectives, Funnel-Transformer is able to recover a deep representation for each token from the reduced hidden sequence via a decoder. Empirically, with comparable or fewer FLOPs, Funnel-Transformer outperforms the standard Transformer on a wide variety of sequence-level prediction tasks, including text classification, language understanding, and reading comprehension.

Tips:

  • Since Funnel Transformer uses pooling, the sequence length of the hidden states changes after each block of layers. The base model therefore has a final sequence length that is a quarter of the original one. This model can be used directly for tasks that just require a sentence summary (like sequence classification or multiple choice). For other tasks, the full model is used; this full model has a decoder that upsamples the final hidden states to the same sequence length as the input.
  • The Funnel Transformer checkpoints are all available with a full version and a base version. The first ones should be used for FunnelModel, FunnelForPreTraining, FunnelForMaskedLM, FunnelForTokenClassification and class:~transformers.FunnelForQuestionAnswering. The second ones should be used for FunnelBaseModel, FunnelForSequenceClassification and FunnelForMultipleChoice.

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

FunnelConfig

class transformers.FunnelConfig < >

( vocab_size = 30522 block_sizes = [4, 4, 4] block_repeats = None num_decoder_layers = 2 d_model = 768 n_head = 12 d_head = 64 d_inner = 3072 hidden_act = 'gelu_new' hidden_dropout = 0.1 attention_dropout = 0.1 activation_dropout = 0.0 max_position_embeddings = 512 type_vocab_size = 3 initializer_range = 0.1 initializer_std = None layer_norm_eps = 1e-09 pooling_type = 'mean' attention_type = 'relative_shift' separate_cls = True truncate_seq = True pool_q_only = True **kwargs )

Parameters

  • vocab_size (int, optional, defaults to 30522) — Vocabulary size of the Funnel transformer. Defines the number of different tokens that can be represented by the inputs_ids passed when calling FunnelModel or TFFunnelModel.
  • block_sizes (List[int], optional, defaults to [4, 4, 4]) — The sizes of the blocks used in the model.
  • block_repeats (List[int], optional) — If passed along, each layer of each block is repeated the number of times indicated.
  • num_decoder_layers (int, optional, defaults to 2) — The number of layers in the decoder (when not using the base model).
  • d_model (int, optional, defaults to 768) — Dimensionality of the model’s hidden states.
  • n_head (int, optional, defaults to 12) — Number of attention heads for each attention layer in the Transformer encoder.
  • d_head (int, optional, defaults to 64) — Dimensionality of the model’s heads.
  • d_inner (int, optional, defaults to 3072) — Inner dimension in the feed-forward blocks.
  • hidden_act (str or callable, optional, defaults to "gelu_new") — The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "silu" and "gelu_new" are supported.
  • hidden_dropout (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 probabilities.
  • activation_dropout (float, optional, defaults to 0.0) — The dropout probability used between the two layers of the feed-forward blocks.
  • max_position_embeddings (int, optional, defaults to 512) — The maximum sequence length that this model might ever be used with. Typically set this to something large just in case (e.g., 512 or 1024 or 2048).
  • type_vocab_size (int, optional, defaults to 3) — The vocabulary size of the token_type_ids passed when calling FunnelModel or TFFunnelModel.
  • initializer_range (float, optional, defaults to 0.1) — The standard deviation of the uniform initializer for initializing all weight matrices in attention layers.
  • initializer_std (float, optional) — The standard deviation of the normal initializer for initializing the embedding matrix and the weight of linear layers. Will default to 1 for the embedding matrix and the value given by Xavier initialization for linear layers.
  • layer_norm_eps (float, optional, defaults to 1e-9) — The epsilon used by the layer normalization layers.
  • pooling_type (str, optional, defaults to "mean") — Possible values are "mean" or "max". The way pooling is performed at the beginning of each block.
  • attention_type (str, optional, defaults to "relative_shift") — Possible values are "relative_shift" or "factorized". The former is faster on CPU/GPU while the latter is faster on TPU.
  • separate_cls (bool, optional, defaults to True) — Whether or not to separate the cls token when applying pooling.
  • truncate_seq (bool, optional, defaults to False) — When using separate_cls, whether or not to truncate the last token when pooling, to avoid getting a sequence length that is not a multiple of 2.
  • pool_q_only (bool, optional, defaults to False) — Whether or not to apply the pooling only to the query or to query, key and values for the attention layers.

This is the configuration class to store the configuration of a FunnelModel or a TFBertModel. It is used to instantiate a Funnel Transformer 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 Funnel Transformer funnel-transformer/small architecture.

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

FunnelTokenizer

class transformers.FunnelTokenizer < >

( vocab_file do_lower_case = True do_basic_tokenize = True never_split = None unk_token = '<unk>' sep_token = '<sep>' pad_token = '<pad>' cls_token = '<cls>' mask_token = '<mask>' bos_token = '<s>' eos_token = '</s>' tokenize_chinese_chars = True strip_accents = None **kwargs )

Construct a Funnel Transformer tokenizer.

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

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

build_inputs_with_special_tokens < >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None ) β†’ 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. A BERT sequence has the following format:

  • single sequence: [CLS] X [SEP]
  • pair of sequences: [CLS] A [SEP] B [SEP]
get_special_tokens_mask < >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None already_has_special_tokens: bool = False ) β†’ 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 ) β†’ 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. A Funnel Transformer sequence pair mask has the following format:

2 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).

FunnelTokenizerFast

class transformers.FunnelTokenizerFast < >

( vocab_file = None tokenizer_file = None do_lower_case = True unk_token = '<unk>' sep_token = '<sep>' pad_token = '<pad>' cls_token = '<cls>' mask_token = '<mask>' bos_token = '<s>' eos_token = '</s>' clean_text = True tokenize_chinese_chars = True strip_accents = None wordpieces_prefix = '##' **kwargs )

Construct a β€œfast” Funnel Transformer tokenizer (backed by HuggingFace’s tokenizers library).

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

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

create_token_type_ids_from_sequences < >

( token_ids_0: typing.List[int] token_ids_1: typing.Optional[typing.List[int]] = None ) β†’ 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. A Funnel Transformer sequence pair mask has the following format:

2 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).

Funnel specific outputs

class transformers.models.funnel.modeling_funnel.FunnelForPreTrainingOutput < >

( loss: typing.Optional[torch.FloatTensor] = None logits: FloatTensor = None hidden_states: typing.Optional[typing.Tuple[torch.FloatTensor]] = None attentions: typing.Optional[typing.Tuple[torch.FloatTensor]] = None )

Parameters

  • loss (optional, returned when labels is provided, torch.FloatTensor of shape (1,)) — Total loss of the ELECTRA-style objective.
  • logits (torch.FloatTensor of shape (batch_size, sequence_length)) — Prediction scores of the head (scores for each token before SoftMax).
  • hidden_states (tuple(torch.FloatTensor), optional, returned when 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.

Output type of FunnelForPreTraining.

class transformers.models.funnel.modeling_tf_funnel.TFFunnelForPreTrainingOutput < >

( logits: Tensor = None hidden_states: typing.Optional[typing.Tuple[tensorflow.python.framework.ops.Tensor]] = None attentions: typing.Optional[typing.Tuple[tensorflow.python.framework.ops.Tensor]] = None )

Parameters

  • logits (tf.Tensor of shape (batch_size, sequence_length)) — Prediction scores of the head (scores for each 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.

Output type of FunnelForPreTraining.

FunnelBaseModel

class transformers.FunnelBaseModel < >

( config )

Parameters

  • config (FunnelConfig) — 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 base Funnel Transformer Model transformer outputting raw hidden-states without upsampling head (also called decoder) or any task-specific head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None position_ids = None head_mask = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 (FunnelConfig) and inputs.

  • last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size)) β€” 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) β€” 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) β€” 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 FunnelBaseModel 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 FunnelTokenizer, FunnelBaseModel
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = FunnelBaseModel.from_pretrained("funnel-transformer/small-base")

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

>>> last_hidden_states = outputs.last_hidden_state

FunnelModel

class transformers.FunnelModel < >

( config )

Parameters

  • config (FunnelConfig) — 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 Funnel Transformer Model transformer outputting raw hidden-states without any specific head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 (FunnelConfig) and inputs.

  • last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size)) β€” 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) β€” 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) β€” 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 FunnelModel 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 FunnelTokenizer, FunnelModel
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = FunnelModel.from_pretrained("funnel-transformer/small")

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

>>> last_hidden_states = outputs.last_hidden_state

FunnelModelForPreTraining

forward < >

( input_ids = None attention_mask = None token_type_ids = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ transformers.models.funnel.modeling_funnel.FunnelForPreTrainingOutput 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 ELECTRA-style loss. Input should be a sequence of tokens (see input_ids docstring) Indices should be in [0, 1]:

    • 0 indicates the token is an original token,
    • 1 indicates the token was replaced.

A transformers.models.funnel.modeling_funnel.FunnelForPreTrainingOutput 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 (FunnelConfig) and inputs.

  • loss (optional, returned when labels is provided, torch.FloatTensor of shape (1,)) β€” Total loss of the ELECTRA-style objective.

  • logits (torch.FloatTensor of shape (batch_size, sequence_length)) β€” Prediction scores of the head (scores for each token before SoftMax).

  • hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) β€” 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) β€” 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 FunnelForPreTraining 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.

Examples:

>>> from transformers import FunnelTokenizer, FunnelForPreTraining
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = FunnelForPreTraining.from_pretrained("funnel-transformer/small")

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

FunnelForMaskedLM

class transformers.FunnelForMaskedLM < >

( config )

Parameters

  • config (FunnelConfig) — 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.

Funnel Transformer Model with a language modeling head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • inputs_embeds (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.
  • output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.
  • output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
  • labels (torch.LongTensor of shape (batch_size, sequence_length), optional) — Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size]

Returns

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 (FunnelConfig) and inputs.

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

  • logits (torch.FloatTensor of shape (batch_size, sequence_length, config.vocab_size)) β€” 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) β€” 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) β€” 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 FunnelForMaskedLM 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 FunnelTokenizer, FunnelForMaskedLM
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = FunnelForMaskedLM.from_pretrained("funnel-transformer/small")

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

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

FunnelForSequenceClassification

class transformers.FunnelForSequenceClassification < >

( config )

Parameters

  • config (FunnelConfig) — 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.

Funnel Transformer Model with a sequence classification/regression head on top (two linear layer on top of the first timestep of the last hidden state) e.g. for GLUE tasks.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 (FunnelConfig) and inputs.

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

  • logits (torch.FloatTensor of shape (batch_size, config.num_labels)) β€” 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) β€” 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) β€” 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 FunnelForSequenceClassification 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 FunnelTokenizer, FunnelForSequenceClassification
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = FunnelForSequenceClassification.from_pretrained("funnel-transformer/small-base")

>>> 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 FunnelTokenizer, FunnelForSequenceClassification
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = FunnelForSequenceClassification.from_pretrained("funnel-transformer/small-base", 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

FunnelForMultipleChoice

class transformers.FunnelForMultipleChoice < >

( config )

Parameters

  • config (FunnelConfig) — 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.

Funnel Transformer Model with a multiple choice classification head on top (two linear layer on top of the first timestep of the last hidden state, and a softmax) e.g. for RocStories/SWAG tasks.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 (FunnelConfig) and inputs.

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

  • logits (torch.FloatTensor of shape (batch_size, num_choices)) β€” 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) β€” 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) β€” 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 FunnelForMultipleChoice 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 FunnelTokenizer, FunnelForMultipleChoice
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = FunnelForMultipleChoice.from_pretrained("funnel-transformer/small-base")

>>> 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

FunnelForTokenClassification

class transformers.FunnelForTokenClassification < >

( config )

Parameters

  • config (FunnelConfig) — 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.

Funnel Transformer 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.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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 (FunnelConfig) and inputs.

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

  • logits (torch.FloatTensor of shape (batch_size, sequence_length, config.num_labels)) β€” 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) β€” 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) β€” 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 FunnelForTokenClassification 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 FunnelTokenizer, FunnelForTokenClassification
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = FunnelForTokenClassification.from_pretrained("funnel-transformer/small")

>>> 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

FunnelForQuestionAnswering

class transformers.FunnelForQuestionAnswering < >

( config )

Parameters

  • config (FunnelConfig) — 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.

Funnel Transformer 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).

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None start_positions = None end_positions = None output_attentions = None output_hidden_states = None return_dict = None ) β†’ 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 BertTokenizer. 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.

    What are attention masks?

  • 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?

  • 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.

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 (FunnelConfig) and inputs.

  • loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) β€” 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)) β€” Span-start scores (before SoftMax).

  • end_logits (torch.FloatTensor of shape (batch_size, sequence_length)) β€” 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) β€” 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) β€” 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 FunnelForQuestionAnswering 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 FunnelTokenizer, FunnelForQuestionAnswering
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = FunnelForQuestionAnswering.from_pretrained("funnel-transformer/small")

>>> 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

TFFunnelBaseModel

class transformers.TFFunnelBaseModel < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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 base Funnel Transformer Model transformer outputting raw hidden-states without upsampling head (also called decoder) or any task-specific head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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).

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 (FunnelConfig) and inputs.

  • last_hidden_state (tf.Tensor of shape (batch_size, sequence_length, hidden_size)) β€” 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) β€” 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) β€” 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 TFFunnelBaseModel 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 FunnelTokenizer, TFFunnelBaseModel
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = TFFunnelBaseModel.from_pretrained("funnel-transformer/small-base")

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

>>> last_hidden_states = outputs.last_hidden_state

TFFunnelModel

class transformers.TFFunnelModel < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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 Funnel Transformer Model transformer outputting raw hidden-states without any specific head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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).

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 (FunnelConfig) and inputs.

  • last_hidden_state (tf.Tensor of shape (batch_size, sequence_length, hidden_size)) β€” 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) β€” 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) β€” 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 TFFunnelModel 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 FunnelTokenizer, TFFunnelModel
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = TFFunnelModel.from_pretrained("funnel-transformer/small")

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

>>> last_hidden_states = outputs.last_hidden_state

TFFunnelModelForPreTraining

class transformers.TFFunnelForPreTraining < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

Funnel model with a binary classification head on top as used during pretraining for identifying generated tokens.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None training = False **kwargs ) β†’ transformers.models.funnel.modeling_tf_funnel.TFFunnelForPreTrainingOutput 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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).

A transformers.models.funnel.modeling_tf_funnel.TFFunnelForPreTrainingOutput 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 (FunnelConfig) and inputs.

  • logits (tf.Tensor of shape (batch_size, sequence_length)) β€” Prediction scores of the head (scores for each token before SoftMax).

  • hidden_states (tuple(tf.Tensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) β€” 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) β€” 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 TFFunnelForPreTraining 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.

Examples:

>>> from transformers import FunnelTokenizer, TFFunnelForPreTraining
>>> import torch

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = TFFunnelForPreTraining.from_pretrained("funnel-transformer/small")

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

TFFunnelForMaskedLM

class transformers.TFFunnelForMaskedLM < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

Funnel Model with a language modeling head on top.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) β†’ transformers.modeling_tf_outputs.TFMaskedLMOutput 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • inputs_embeds (tf.Tensor of shape (batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.
  • output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.
  • output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.
  • training (bool, optional, defaults to False) — Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).
  • labels (tf.Tensor of shape (batch_size, sequence_length), optional) — Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size]

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

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

  • logits (tf.Tensor of shape (batch_size, sequence_length, config.vocab_size)) β€” 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) β€” 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) β€” 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 TFFunnelForMaskedLM 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 FunnelTokenizer, TFFunnelForMaskedLM
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = TFFunnelForMaskedLM.from_pretrained("funnel-transformer/small")

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

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

TFFunnelForSequenceClassification

class transformers.TFFunnelForSequenceClassification < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

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

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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).

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 (FunnelConfig) and inputs.

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

  • logits (tf.Tensor of shape (batch_size, config.num_labels)) β€” 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) β€” 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) β€” 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 TFFunnelForSequenceClassification 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 FunnelTokenizer, TFFunnelForSequenceClassification
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = TFFunnelForSequenceClassification.from_pretrained("funnel-transformer/small-base")

>>> 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

TFFunnelForMultipleChoice

class transformers.TFFunnelForMultipleChoice < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

Funnel 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.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • inputs_embeds (tf.Tensor of shape (batch_size, num_choices, sequence_length, hidden_size), optional) — Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix.
  • output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.
  • output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the config will be used instead.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple. This argument can be used in eager mode, in graph mode the value will always be set to True.
  • training (bool, optional, defaults to False) — Whether or not to use the model in training mode (some modules like dropout modules have different behaviors between training and evaluation).
  • labels (tf.Tensor of shape (batch_size,), optional) — Labels for computing the multiple choice classification loss. Indices should be in [0, ..., num_choices] where num_choices is the size of the second dimension of the input tensors. (See input_ids above)

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 (FunnelConfig) and inputs.

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

  • logits (tf.Tensor of shape (batch_size, num_choices)) β€” 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) β€” 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) β€” 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 TFFunnelForMultipleChoice 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 FunnelTokenizer, TFFunnelForMultipleChoice
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small-base")
>>> model = TFFunnelForMultipleChoice.from_pretrained("funnel-transformer/small-base")

>>> 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

TFFunnelForTokenClassification

class transformers.TFFunnelForTokenClassification < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

Funnel 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.

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None labels = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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].

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 (FunnelConfig) and inputs.

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

  • logits (tf.Tensor of shape (batch_size, sequence_length, config.num_labels)) β€” 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) β€” 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) β€” 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 TFFunnelForTokenClassification 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 FunnelTokenizer, TFFunnelForTokenClassification
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = TFFunnelForTokenClassification.from_pretrained("funnel-transformer/small")

>>> 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

TFFunnelForQuestionAnswering

class transformers.TFFunnelForQuestionAnswering < >

( *args **kwargs )

Parameters

  • config (XxxConfig) — 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.

Funnel 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).

The Funnel Transformer model was proposed in Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.

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 token_type_ids = None inputs_embeds = None output_attentions = None output_hidden_states = None return_dict = None start_positions = None end_positions = None training = False **kwargs ) β†’ 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 FunnelTokenizer. 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.

    What are attention masks?

  • 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?

  • 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.

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 (FunnelConfig) and inputs.

  • loss (tf.Tensor of shape (batch_size, ), optional, returned when start_positions and end_positions are provided) β€” 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)) β€” Span-start scores (before SoftMax).

  • end_logits (tf.Tensor of shape (batch_size, sequence_length)) β€” 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) β€” 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) β€” 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 TFFunnelForQuestionAnswering 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 FunnelTokenizer, TFFunnelForQuestionAnswering
>>> import tensorflow as tf

>>> tokenizer = FunnelTokenizer.from_pretrained("funnel-transformer/small")
>>> model = TFFunnelForQuestionAnswering.from_pretrained("funnel-transformer/small")

>>> question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
>>> input_dict = tokenizer(question, text, return_tensors="tf")
>>> outputs = model(input_dict)
>>> start_logits = outputs.start_logits
>>> end_logits = outputs.end_logits

>>> all_tokens = tokenizer.convert_ids_to_tokens(input_dict["input_ids"].numpy()[0])
>>> answer = " ".join(all_tokens[tf.math.argmax(start_logits, 1)[0] : tf.math.argmax(end_logits, 1)[0] + 1])