SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

TF GPT2

Browse Source
This commit is contained in:
Lysandre
2020-01-17 10:28:12 -05:00
committed by Lysandre Debut
parent bd0d3fd76e
commit 1487b840d3
3 changed files with 150 additions and 139 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docs/source/model_doc/gpt2.rst
View File

@@ -30,6 +30,10 @@ Tips:
See `reusing the past in generative models <../quickstart.html#using-the-past>`_ for more information on the usage
of this argument.
`Write With Transformer <https://transformer.huggingface.co/doc/gpt2-large>`__ is a webapp created and hosted by
Hugging Face showcasing the generative capabilities of several models. GPT-2 is one of them and is available in five
different sizes: small, medium, large, xl and a distilled version of the small checkpoint: distilgpt-2.
``GPT2Config``
~~~~~~~~~~~~~~~~~~~~~

6
src/transformers/modeling_tf_bert.py
View File

@@ -589,12 +589,6 @@ BERT_START_DOCSTRING = r"""
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.
.. _`BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding`:
https://arxiv.org/abs/1810.04805
.. _`tf.keras.Model`:
https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/Model
.. note::
TF 2.0 models accepts two formats as inputs:

279
src/transformers/modeling_tf_gpt2.py
View File

@@ -22,7 +22,7 @@ import numpy as np
import tensorflow as tf
from .configuration_gpt2 import GPT2Config
from .file_utils import add_start_docstrings
from .file_utils import add_start_docstrings, add_start_docstrings_to_callable
from .modeling_tf_utils import (
TFConv1D,
TFPreTrainedModel,
@@ -368,36 +368,25 @@ class TFGPT2PreTrainedModel(TFPreTrainedModel):
base_model_prefix = "transformer"
GPT2_START_DOCSTRING = r""" OpenAI GPT-2 model was proposed in
`Language Models are Unsupervised Multitask Learners`_
by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
It's a causal (unidirectional) transformer pre-trained using language modeling on a very large
corpus of ~40 GB of text data.
GPT2_START_DOCSTRING = r"""
This model is a tf.keras.Model `tf.keras.Model`_ sub-class. Use it as a regular TF 2.0 Keras Model and
refer to the TF 2.0 documentation for all matter related to general usage and behavior.
.. _`Language Models are Unsupervised Multitask Learners`:
https://openai.com/blog/better-language-models/
.. _`tf.keras.Model`:
https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/Model
Note on the model inputs:
.. note::
TF 2.0 models accepts two formats as inputs:
- having all inputs as keyword arguments (like PyTorch models), or
- having all inputs as a list, tuple or dict in the first positional arguments.
This second option is usefull 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)`.
This second option is useful when using :obj:`tf.keras.Model.fit()` method which currently requires having
all the tensors in the first argument of the model call function: :obj:`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 :
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 single Tensor with input_ids only and nothing else: :obj:`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 associaed to the input names given in the docstring:
`model({'input_ids': input_ids, 'token_type_ids': token_type_ids})`
:obj:`model([input_ids, attention_mask])` or :obj:`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:
:obj:`model({'input_ids': input_ids, 'token_type_ids': token_type_ids})`
Parameters:
config (:class:`~transformers.GPT2Config`): Model configuration class with all the parameters of the model.
@@ -405,35 +394,43 @@ GPT2_START_DOCSTRING = r""" OpenAI GPT-2 model was proposed in
Check out the :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model weights.
"""
GPT2_INPUTS_DOCSTRING = r""" Inputs:
**input_ids**: ``Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, sequence_length)``:
Indices of input sequence tokens in the vocabulary.
GPT-2 is a model with absolute position embeddings so it's usually advised to pad the inputs on
the right rather than the left.
Indices can be obtained using :class:`transformers.BPT2Tokenizer`.
GPT2_INPUTS_DOCSTRING = r"""
Args:
input_ids (:obj:`Numpy array` or :obj:`tf.Tensor` of shape :obj:`(batch_size, sequence_length)`):
Indices of input sequence tokens in the vocabulary.
Indices can be obtained using :class:`transformers.GPT2Tokenizer`.
See :func:`transformers.PreTrainedTokenizer.encode` and
:func:`transformers.PreTrainedTokenizer.convert_tokens_to_ids` for details.
**past**:
list of ``Numpy array`` or ``tf.Tensor`` (one for each layer):
that contains pre-computed hidden-states (key and values in the attention blocks) as computed by the model
(see `past` output below). Can be used to speed up sequential decoding.
**attention_mask**: (`optional`) ``Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, sequence_length)``:
:func:`transformers.PreTrainedTokenizer.encode_plus` for details.
`What are input IDs? <../glossary.html#input-ids>`__
past (:obj:`List[tf.Tensor]` of length :obj:`config.n_layers`):
Contains pre-computed hidden-states (key and values in the attention blocks) as computed by the model
(see `past` output below). Can be used to speed up sequential decoding. The token ids which have their past given to this model
should not be passed as input ids as they have already been computed.
attention_mask (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length)`, `optional`, defaults to :obj:`None`):
Mask to avoid performing attention on padding token indices.
Mask values selected in ``[0, 1]``:
``1`` for tokens that are NOT MASKED, ``0`` for MASKED tokens.
**token_type_ids**: (`optional`) ```Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, sequence_length)``:
A parallel sequence of tokens (can be used to indicate various portions of the inputs).
The embeddings from these tokens will be summed with the respective token embeddings.
Indices are selected in the vocabulary (unlike BERT which has a specific vocabulary for segment indices).
**position_ids**: (`optional`) ```Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, sequence_length)``:
`What are attention masks? <../glossary.html#attention-mask>`__
token_type_ids (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length)`, `optional`, defaults to :obj:`None`):
Segment token indices to indicate first and second portions of the inputs.
Indices are selected in ``[0, 1]``: ``0`` corresponds to a `sentence A` token, ``1``
corresponds to a `sentence B` token
`What are token type IDs? <../glossary.html#token-type-ids>`_
position_ids (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length)`, `optional`, defaults to :obj:`None`):
Indices of positions of each input sequence tokens in the position embeddings.
Selected in the range ``[0, config.max_position_embeddings - 1]``.
**head_mask**: (`optional`) ``Numpy array`` or ``tf.Tensor`` of shape ``(num_heads,)`` or ``(num_layers, num_heads)``:
`What are position IDs? <../glossary.html#position-ids>`_
head_mask (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(num_heads,)` or :obj:`(num_layers, num_heads)`, `optional`, defaults to :obj:`None`):
Mask to nullify selected heads of the self-attention modules.
Mask values selected in ``[0, 1]``:
``1`` indicates the head is **not masked**, ``0`` indicates the head is **masked**.
**inputs_embeds**: (`optional`) ``Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, sequence_length, embedding_dim)``:
Optionally, instead of passing ``input_ids`` you can choose to directly pass an embedded representation.
:obj:`1` indicates the head is **not masked**, :obj:`0` indicates the head is **masked**.
input_embeds (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`, defaults to :obj:`None`):
Optionally, instead of passing :obj:`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.
"""
@@ -442,24 +439,35 @@ GPT2_INPUTS_DOCSTRING = r""" Inputs:
@add_start_docstrings(
"The bare GPT2 Model transformer outputing raw hidden-states without any specific head on top.",
GPT2_START_DOCSTRING,
GPT2_INPUTS_DOCSTRING,
)
class TFGPT2Model(TFGPT2PreTrainedModel):
r"""
Outputs: `Tuple` comprising various elements depending on the configuration (config) and inputs:
**last_hidden_state**: ``tf.Tensor`` of shape ``(batch_size, sequence_length, hidden_size)``
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
self.transformer = TFGPT2MainLayer(config, name="transformer")
@add_start_docstrings_to_callable(GPT2_INPUTS_DOCSTRING)
def call(self, inputs, **kwargs):
r"""
Return:
:obj:`tuple(tf.Tensor)` comprising various elements depending on the configuration (config) and inputs:
last_hidden_state (:obj:`tf.Tensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`):
Sequence of hidden-states at the last layer of the model.
**past**:
list of ``tf.Tensor`` (one for each layer) of shape ``(2, batch_size, num_heads, sequence_length, embed_size_per_head)``:
that contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding.
**hidden_states**: (`optional`, returned when ``config.output_hidden_states=True``)
list of ``tf.Tensor`` (one for the output of each layer + the output of the embeddings)
of shape ``(batch_size, sequence_length, hidden_size)``:
past (:obj:`List[tf.Tensor]` of length :obj:`config.n_layers` with each tensor of shape :obj:`(2, batch_size, num_heads, sequence_length, embed_size_per_head)`):
Contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding. The token ids which have their past given to this model
should not be passed as input ids as they have already been computed.
hidden_states (:obj:`tuple(tf.Tensor)` `optional`, returned when ``config.output_hidden_states=True``):
Tuple of :obj:`tf.Tensor` (one for the output of the embeddings + one for the output of each layer)
of shape :obj:`(batch_size, sequence_length, hidden_size)`.
Hidden-states of the model at the output of each layer plus the initial embedding outputs.
**attentions**: (`optional`, returned when ``config.output_attentions=True``)
list 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.
attentions (:obj:`tuple(tf.Tensor)`, `optional`, returned when ``config.output_attentions=True``):
Tuple of :obj:`tf.Tensor` (one for each layer) of shape
:obj:`(batch_size, num_heads, sequence_length, sequence_length)`.
Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
heads.
Examples::
@@ -473,38 +481,46 @@ class TFGPT2Model(TFGPT2PreTrainedModel):
last_hidden_states = outputs[0] # The last hidden-state is the first element of the output tuple
"""
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
self.transformer = TFGPT2MainLayer(config, name="transformer")
def call(self, inputs, **kwargs):
outputs = self.transformer(inputs, **kwargs)
return outputs
@add_start_docstrings(
"""The GPT2 Model transformer with a language modeling head on top
(linear layer with weights tied to the input embeddings). """,
(linear layer with weights tied to the input embeddings). """,
GPT2_START_DOCSTRING,
GPT2_INPUTS_DOCSTRING,
)
class TFGPT2LMHeadModel(TFGPT2PreTrainedModel):
r"""
Outputs: `Tuple` comprising various elements depending on the configuration (config) and inputs:
**prediction_scores**: `tf.Tensor`` of shape ``(batch_size, sequence_length, config.vocab_size)``
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
self.transformer = TFGPT2MainLayer(config, name="transformer")
def get_output_embeddings(self):
return self.transformer.wte
@add_start_docstrings_to_callable(GPT2_INPUTS_DOCSTRING)
def call(self, inputs, **kwargs):
r"""
Return:
:obj:`tuple(tf.Tensor)` comprising various elements depending on the configuration (:obj:`~transformers.GPT2Config`) and inputs:
prediction_scores (:obj:`tf.Tensor` of shape :obj:`(batch_size, sequence_length, config.vocab_size)`):
Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
**past**:
list of `tf.Tensor`` (one for each layer) of shape ``(2, batch_size, num_heads, sequence_length, embed_size_per_head)``:
that contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding.
**hidden_states**: (`optional`, returned when ``config.output_hidden_states=True``)
list of `tf.Tensor`` (one for the output of each layer + the output of the embeddings)
of shape ``(batch_size, sequence_length, hidden_size)``:
past (:obj:`List[tf.Tensor]` of length :obj:`config.n_layers` with each tensor of shape :obj:`(2, batch_size, num_heads, sequence_length, embed_size_per_head)`):
Contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding. The token ids which have their past given to this model
should not be passed as input ids as they have already been computed.
hidden_states (:obj:`tuple(tf.Tensor)`, `optional`, returned when ``config.output_hidden_states=True``):
Tuple of :obj:`tf.Tensor` (one for the output of the embeddings + one for the output of each layer)
of shape :obj:`(batch_size, sequence_length, hidden_size)`.
Hidden-states of the model at the output of each layer plus the initial embedding outputs.
**attentions**: (`optional`, returned when ``config.output_attentions=True``)
list 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.
attentions (:obj:`tuple(tf.Tensor)`, `optional`, returned when ``config.output_attentions=True``):
Tuple of :obj:`tf.Tensor` (one for each layer) of shape
:obj:`(batch_size, num_heads, sequence_length, sequence_length)`.
Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
heads.
Examples::
@@ -518,16 +534,7 @@ class TFGPT2LMHeadModel(TFGPT2PreTrainedModel):
outputs = model(input_ids)
logits = outputs[0]
"""
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
self.transformer = TFGPT2MainLayer(config, name="transformer")
def get_output_embeddings(self):
return self.transformer.wte
def call(self, inputs, **kwargs):
"""
transformer_outputs = self.transformer(inputs, **kwargs)
hidden_states = transformer_outputs[0]
@@ -540,35 +547,65 @@ class TFGPT2LMHeadModel(TFGPT2PreTrainedModel):
@add_start_docstrings(
"""The GPT2 Model transformer with a language modeling and a multiple-choice classification
head on top e.g. for RocStories/SWAG tasks. The two heads are two linear layers.
The language modeling head has its weights tied to the input embeddings,
the classification head takes as input the input of a specified classification token index in the input sequence).
head on top e.g. for RocStories/SWAG tasks. The two heads are two linear layers.
The language modeling head has its weights tied to the input embeddings,
the classification head takes as input the input of a specified classification token index in the input sequence).
""",
GPT2_START_DOCSTRING,
GPT2_INPUTS_DOCSTRING,
)
class TFGPT2DoubleHeadsModel(TFGPT2PreTrainedModel):
r"""
**mc_token_ids**: (`optional`, default to index of the last token of the input) ``Numpy array`` or ``tf.Tensor`` of shape ``(batch_size, num_choices)``:
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
config.num_labels = 1
self.transformer = TFGPT2MainLayer(config, name="transformer")
self.multiple_choice_head = TFSequenceSummary(
config, initializer_range=config.initializer_range, name="multiple_choice_head"
)
def get_output_embeddings(self):
return self.transformer.wte
@add_start_docstrings_to_callable(GPT2_INPUTS_DOCSTRING)
def call(
self,
inputs,
past=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
mc_token_ids=None,
training=False,
):
r"""
mc_token_ids (:obj:`tf.Tensor` or :obj:`Numpy array` of shape :obj:`(batch_size, num_choices)`, `optional`, default to index of the last token of the input)
Index of the classification token in each input sequence.
Selected in the range ``[0, input_ids.size(-1) - 1[``.
Outputs: `Tuple` comprising various elements depending on the configuration (config) and inputs:
**lm_prediction_scores**: `tf.Tensor`` of shape ``(batch_size, num_choices, sequence_length, config.vocab_size)``
Return:
:obj:`tuple(tf.Tensor)` comprising various elements depending on the configuration (:obj:`~transformers.GPT2Config`) and inputs:
lm_prediction_scores (:obj:`tf.Tensor` of shape :obj:`(batch_size, num_choices, sequence_length, config.vocab_size)`):
Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
**mc_prediction_scores**: `tf.Tensor`` of shape ``(batch_size, num_choices)``
Prediction scores of the multiplechoice classification head (scores for each choice before SoftMax).
**past**:
list of `tf.Tensor`` (one for each layer) of shape ``(2, batch_size, num_heads, sequence_length, embed_size_per_head)``:
that contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding.
**hidden_states**: (`optional`, returned when ``config.output_hidden_states=True``)
list of `tf.Tensor`` (one for the output of each layer + the output of the embeddings)
of shape ``(batch_size, sequence_length, hidden_size)``:
mc_prediction_scores (:obj:`tf.Tensor` of shape :obj:`(batch_size, num_choices)`):
Prediction scores of the multiple choice classification head (scores for each choice before SoftMax).
past (:obj:`List[tf.Tensor]` of length :obj:`config.n_layers` with each tensor of shape :obj:`(2, batch_size, num_heads, sequence_length, embed_size_per_head)`):
Contains pre-computed hidden-states (key and values in the attention blocks).
Can be used (see `past` input) to speed up sequential decoding. The token ids which have their past given to this model
should not be passed as input ids as they have already been computed.
hidden_states (:obj:`tuple(tf.Tensor)`, `optional`, returned when ``config.output_hidden_states=True``):
Tuple of :obj:`tf.Tensor` (one for the output of the embeddings + one for the output of each layer)
of shape :obj:`(batch_size, sequence_length, hidden_size)`.
Hidden-states of the model at the output of each layer plus the initial embedding outputs.
**attentions**: (`optional`, returned when ``config.output_attentions=True``)
list 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.
attentions (:obj:`tuple(tf.Tensor)`, `optional`, returned when ``config.output_attentions=True``):
Tuple of :obj:`tf.Tensor` (one for each layer) of shape
:obj:`(batch_size, num_heads, sequence_length, sequence_length)`.
Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
heads.
Examples::
@@ -595,31 +632,7 @@ class TFGPT2DoubleHeadsModel(TFGPT2PreTrainedModel):
outputs = model(input_ids, mc_token_ids=mc_token_ids)
lm_prediction_scores, mc_prediction_scores = outputs[:2]
"""
def __init__(self, config, *inputs, **kwargs):
super().__init__(config, *inputs, **kwargs)
config.num_labels = 1
self.transformer = TFGPT2MainLayer(config, name="transformer")
self.multiple_choice_head = TFSequenceSummary(
config, initializer_range=config.initializer_range, name="multiple_choice_head"
)
def get_output_embeddings(self):
return self.transformer.wte
def call(
self,
inputs,
past=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
mc_token_ids=None,
training=False,
):
"""
if isinstance(inputs, (tuple, list)):
input_ids = inputs[0]
past = inputs[1] if len(inputs) > 1 else past