Source code for transformers.models.bert_generation.modeling_bert_generation

# coding=utf-8
# Copyright 2020 The Google AI Language Team Authors and The HuggingFace Inc. team.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""PyTorch BERT model specific for generation. """


import torch
import torch.utils.checkpoint
from torch import nn
from torch.nn import CrossEntropyLoss

from ...file_utils import (
    add_code_sample_docstrings,
    add_start_docstrings,
    add_start_docstrings_to_model_forward,
    replace_return_docstrings,
)
from ...modeling_outputs import BaseModelOutputWithPastAndCrossAttentions, CausalLMOutputWithCrossAttentions
from ...modeling_utils import PreTrainedModel
from ...utils import logging
from ..bert.modeling_bert import BertEncoder
from .configuration_bert_generation import BertGenerationConfig


logger = logging.get_logger(__name__)

_CHECKPOINT_FOR_DOC = "google/bert_for_seq_generation_L-24_bbc_encoder"
_CONFIG_FOR_DOC = "BertGenerationConfig"
_TOKENIZER_FOR_DOC = "BertGenerationTokenizer"


def load_tf_weights_in_bert_generation(
    model, tf_hub_path, model_class, is_encoder_named_decoder=False, is_encoder=False
):
    try:
        import numpy as np
        import tensorflow.compat.v1 as tf

        import tensorflow_hub as hub
        import tensorflow_text  # noqa: F401

        tf.disable_eager_execution()
    except ImportError:
        logger.error(
            "Loading a TensorFlow model in PyTorch, requires TensorFlow to be installed. Please see "
            "https://www.tensorflow.org/install/ for installation instructions."
        )
        raise
    tf_model = hub.Module(tf_hub_path)
    init = tf.global_variables_initializer()
    with tf.Session() as sess:
        init.run()
        all_variables = tf_model.variable_map
        keep_track_variables = all_variables.copy()
        for key in list(all_variables.keys()):
            if "global" in key:
                logger.info(f"Skipping {key}...")
                continue
            if not is_encoder:
                model_pointer = getattr(model, model_class)
            else:
                model_pointer = model
            is_embedding = False
            logger.info(f"Trying to match {key}...")
            # remove start_string = "module/bert/"
            sub_layers = key.split("/")[2:]
            if is_encoder_named_decoder and sub_layers[0] == "encoder":
                logger.info(f"Skipping encoder layer {key} for decoder")
                continue
            if is_encoder and sub_layers[0] == "decoder":
                logger.info(f"Skipping decoder layer {key} for encoder")
                continue
            for i, sub_layer in enumerate(sub_layers):
                if sub_layer == "embeddings":
                    is_embedding = True
                elif sub_layer == "LayerNorm":
                    is_embedding = False
                if "layer" in sub_layer:
                    model_pointer = model_pointer.layer[int(sub_layer.split("_")[-1])]
                elif sub_layer in ["kernel", "gamma"]:
                    model_pointer = model_pointer.weight
                elif sub_layer == "beta":
                    model_pointer = model_pointer.bias
                elif sub_layer == "encdec":
                    model_pointer = model_pointer.crossattention.self
                elif sub_layer == "encdec_output":
                    model_pointer = model_pointer.crossattention.output
                elif is_encoder_named_decoder and sub_layer == "decoder":
                    model_pointer = model_pointer.encoder
                else:
                    if sub_layer == "attention" and "encdec" in sub_layers[i + 1]:
                        continue
                    try:
                        model_pointer = getattr(model_pointer, sub_layer)
                    except AttributeError:
                        logger.info(f"Skipping to initialize {key} at {sub_layer}...")
                        raise AttributeError

            array = np.asarray(sess.run(all_variables[key]))
            if not is_embedding:
                logger.info(f"Transposing numpy weight of shape {array.shape} for {key}")
                array = np.transpose(array)
            else:
                model_pointer = model_pointer.weight

            try:
                assert (
                    model_pointer.shape == array.shape
                ), f"Pointer shape {model_pointer.shape} and array shape {array.shape} mismatched"
            except AssertionError as e:
                e.args += (model_pointer.shape, array.shape)
                raise
            logger.info(f"Initialize PyTorch weight {key}")

            model_pointer.data = torch.from_numpy(array.astype(np.float32))
            keep_track_variables.pop(key, None)

        logger.info(f"Weights not copied to PyTorch model: {', '.join(keep_track_variables.keys())}")
        return model


class BertGenerationEmbeddings(nn.Module):
    """Construct the embeddings from word and position embeddings."""

    def __init__(self, config):
        super().__init__()
        self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=config.pad_token_id)
        self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
        # self.LayerNorm is not snake-cased to stick with TensorFlow model variable name and be able to load
        # any TensorFlow checkpoint file
        self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
        self.dropout = nn.Dropout(config.hidden_dropout_prob)

        # position_ids (1, len position emb) is contiguous in memory and exported when serialized
        self.register_buffer("position_ids", torch.arange(config.max_position_embeddings).expand((1, -1)))

    def forward(self, input_ids=None, position_ids=None, inputs_embeds=None, past_key_values_length=0):
        if input_ids is not None:
            input_shape = input_ids.size()
        else:
            input_shape = inputs_embeds.size()[:-1]

        seq_length = input_shape[1]

        if position_ids is None:
            position_ids = self.position_ids[:, past_key_values_length : seq_length + past_key_values_length]

        if inputs_embeds is None:
            inputs_embeds = self.word_embeddings(input_ids)
        position_embeddings = self.position_embeddings(position_ids)

        embeddings = inputs_embeds + position_embeddings
        embeddings = self.LayerNorm(embeddings)
        embeddings = self.dropout(embeddings)
        return embeddings


class BertGenerationPreTrainedModel(PreTrainedModel):
    """
    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
    models.
    """

    config_class = BertGenerationConfig
    base_model_prefix = "bert"
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def _init_weights(self, module):
        """Initialize the weights"""
        if isinstance(module, nn.Linear):
            # Slightly different from the TF version which uses truncated_normal for initialization
            # cf https://github.com/pytorch/pytorch/pull/5617
            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
            if module.bias is not None:
                module.bias.data.zero_()
        elif isinstance(module, nn.Embedding):
            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
            if module.padding_idx is not None:
                module.weight.data[module.padding_idx].zero_()
        elif isinstance(module, nn.LayerNorm):
            module.bias.data.zero_()
            module.weight.data.fill_(1.0)


BERT_GENERATION_START_DOCSTRING = r"""

    This model inherits from :class:`~transformers.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 <https://pytorch.org/docs/stable/nn.html#torch.nn.Module>`__
    subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to
    general usage and behavior.

    Parameters:
        config (:class:`~transformers.BertGenerationConfig`): 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 :meth:`~transformers.PreTrainedModel.from_pretrained` method to load the model
            weights.
"""

BERT_GENERATION_INPUTS_DOCSTRING = r"""
    Args:
        input_ids (:obj:`torch.LongTensor` of shape :obj:`({0})`):
            Indices of input sequence tokens in the vocabulary.

            Indices can be obtained using :class:`~transformers.BertGenerationTokenizer`. See
            :meth:`transformers.PreTrainedTokenizer.__call__` and :meth:`transformers.PreTrainedTokenizer.encode` for
            details.

            `What are input IDs? <../glossary.html#input-ids>`__
        attention_mask (:obj:`torch.FloatTensor` of shape :obj:`({0})`, `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? <../glossary.html#attention-mask>`__
        position_ids (:obj:`torch.LongTensor` of shape :obj:`({0})`, `optional`):
            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range ``[0,
            config.max_position_embeddings - 1]``.

            `What are position IDs? <../glossary.html#position-ids>`_
        head_mask (:obj:`torch.FloatTensor` of shape :obj:`(num_heads,)` or :obj:`(num_layers, num_heads)`, `optional`):
            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 (:obj:`torch.FloatTensor` of shape :obj:`({0}, hidden_size)`, `optional`):
            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 :obj:`input_ids` indices into associated
            vectors than the model's internal embedding lookup matrix.
        output_attentions (:obj:`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 (:obj:`bool`, `optional`):
            Whether or not to return the hidden states of all layers. See ``hidden_states`` under returned tensors for
            more detail.
        return_dict (:obj:`bool`, `optional`):
            Whether or not to return a :class:`~transformers.file_utils.ModelOutput` instead of a plain tuple.
"""


[docs]@add_start_docstrings( "The bare BertGeneration model transformer outputting raw hidden-states without any specific head on top.", BERT_GENERATION_START_DOCSTRING, ) class BertGenerationEncoder(BertGenerationPreTrainedModel): """ The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of cross-attention is added between the self-attention layers, following the architecture described in `Attention is all you need <https://arxiv.org/abs/1706.03762>`__ by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin. This model should be used when leveraging Bert or Roberta checkpoints for the :class:`~transformers.EncoderDecoderModel` class as described in `Leveraging Pre-trained Checkpoints for Sequence Generation Tasks <https://arxiv.org/abs/1907.12461>`__ by Sascha Rothe, Shashi Narayan, and Aliaksei Severyn. To behave as an decoder the model needs to be initialized with the :obj:`is_decoder` argument of the configuration set to :obj:`True`. To be used in a Seq2Seq model, the model needs to initialized with both :obj:`is_decoder` argument and :obj:`add_cross_attention` set to :obj:`True`; an :obj:`encoder_hidden_states` is then expected as an input to the forward pass. """ def __init__(self, config): super().__init__(config) self.config = config self.embeddings = BertGenerationEmbeddings(config) self.encoder = BertEncoder(config) # Initialize weights and apply final processing self.post_init() def get_input_embeddings(self): return self.embeddings.word_embeddings def set_input_embeddings(self, value): self.embeddings.word_embeddings = value def _prune_heads(self, heads_to_prune): """ Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base class PreTrainedModel """ for layer, heads in heads_to_prune.items(): self.encoder.layer[layer].attention.prune_heads(heads)
[docs] @add_start_docstrings_to_model_forward(BERT_GENERATION_INPUTS_DOCSTRING.format("batch_size, sequence_length")) @add_code_sample_docstrings( processor_class=_TOKENIZER_FOR_DOC, checkpoint=_CHECKPOINT_FOR_DOC, output_type=BaseModelOutputWithPastAndCrossAttentions, config_class=_CONFIG_FOR_DOC, ) def forward( self, input_ids=None, attention_mask=None, position_ids=None, head_mask=None, inputs_embeds=None, encoder_hidden_states=None, encoder_attention_mask=None, past_key_values=None, use_cache=None, output_attentions=None, output_hidden_states=None, return_dict=None, ): r""" encoder_hidden_states (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`): Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder. encoder_attention_mask (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`): Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in ``[0, 1]``: ``1`` for tokens that are NOT MASKED, ``0`` for MASKED tokens. past_key_values (:obj:`tuple(tuple(torch.FloatTensor))` of length :obj:`config.n_layers` with each tuple having 4 tensors of shape :obj:`(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`): Contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding. If :obj:`past_key_values` are used, the user can optionally input only the last :obj:`decoder_input_ids` (those that don't have their past key value states given to this model) of shape :obj:`(batch_size, 1)` instead of all :obj:`decoder_input_ids` of shape :obj:`(batch_size, sequence_length)`. use_cache (:obj:`bool`, `optional`): If set to :obj:`True`, :obj:`past_key_values` key value states are returned and can be used to speed up decoding (see :obj:`past_key_values`). """ output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions output_hidden_states = ( output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states ) return_dict = return_dict if return_dict is not None else self.config.use_return_dict if self.config.is_decoder: use_cache = use_cache if use_cache is not None else self.config.use_cache else: use_cache = False if input_ids is not None and inputs_embeds is not None: raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time") elif input_ids is not None: input_shape = input_ids.size() elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] else: raise ValueError("You have to specify either input_ids or inputs_embeds") batch_size, seq_length = input_shape device = input_ids.device if input_ids is not None else inputs_embeds.device # past_key_values_length past_key_values_length = past_key_values[0][0].shape[2] if past_key_values is not None else 0 if attention_mask is None: attention_mask = torch.ones(((batch_size, seq_length + past_key_values_length)), device=device) # We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length] # ourselves in which case we just need to make it broadcastable to all heads. extended_attention_mask = None if not use_cache: extended_attention_mask: torch.Tensor = self.get_extended_attention_mask( attention_mask, input_shape, device ) # If a 2D or 3D attention mask is provided for the cross-attention # we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length] if self.config.is_decoder and encoder_hidden_states is not None: encoder_batch_size, encoder_sequence_length, _ = encoder_hidden_states.size() encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length) if encoder_attention_mask is None: encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device) encoder_extended_attention_mask = self.invert_attention_mask(encoder_attention_mask) else: encoder_extended_attention_mask = None # Prepare head mask if needed # 1.0 in head_mask indicate we keep the head # attention_probs has shape bsz x n_heads x N x N # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads] # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length] head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers) embedding_output = self.embeddings( input_ids=input_ids, position_ids=position_ids, inputs_embeds=inputs_embeds, past_key_values_length=past_key_values_length, ) encoder_outputs = self.encoder( embedding_output, attention_mask=extended_attention_mask, head_mask=head_mask, encoder_hidden_states=encoder_hidden_states, encoder_attention_mask=encoder_extended_attention_mask, past_key_values=past_key_values, use_cache=use_cache, output_attentions=output_attentions, output_hidden_states=output_hidden_states, return_dict=return_dict, ) sequence_output = encoder_outputs[0] if not return_dict: return (sequence_output,) + encoder_outputs[1:] return BaseModelOutputWithPastAndCrossAttentions( last_hidden_state=sequence_output, past_key_values=encoder_outputs.past_key_values, hidden_states=encoder_outputs.hidden_states, attentions=encoder_outputs.attentions, cross_attentions=encoder_outputs.cross_attentions, )
class BertGenerationOnlyLMHead(nn.Module): def __init__(self, config): super().__init__() self.decoder = nn.Linear(config.hidden_size, config.vocab_size) self.bias = nn.Parameter(torch.zeros(config.vocab_size)) self.decoder.bias = self.bias def forward(self, hidden_states): logits = self.decoder(hidden_states) return logits def _tie_weights(self): # To tie those two weights if they get disconnected (on TPU or when the bias is resized) self.bias = self.decoder.bias
[docs]@add_start_docstrings( """BertGeneration Model with a `language modeling` head on top for CLM fine-tuning. """, BERT_GENERATION_START_DOCSTRING, ) class BertGenerationDecoder(BertGenerationPreTrainedModel): def __init__(self, config): super().__init__(config) if not config.is_decoder: logger.warning("If you want to use `BertGenerationDecoder` as a standalone, add `is_decoder=True.`") self.bert = BertGenerationEncoder(config) self.lm_head = BertGenerationOnlyLMHead(config) # Initialize weights and apply final processing self.post_init() def get_output_embeddings(self): return self.lm_head.decoder def set_output_embeddings(self, new_embeddings): self.lm_head.decoder = new_embeddings
[docs] @add_start_docstrings_to_model_forward(BERT_GENERATION_INPUTS_DOCSTRING.format("batch_size, sequence_length")) @replace_return_docstrings(output_type=CausalLMOutputWithCrossAttentions, config_class=_CONFIG_FOR_DOC) def forward( self, input_ids=None, attention_mask=None, position_ids=None, head_mask=None, inputs_embeds=None, encoder_hidden_states=None, encoder_attention_mask=None, labels=None, past_key_values=None, use_cache=None, output_attentions=None, output_hidden_states=None, return_dict=None, ): r""" encoder_hidden_states (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length, hidden_size)`, `optional`): Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder. encoder_attention_mask (:obj:`torch.FloatTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`): Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in ``[0, 1]``: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size, sequence_length)`, `optional`): Labels for computing the left-to-right language modeling loss (next word prediction). 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]`` past_key_values (:obj:`tuple(tuple(torch.FloatTensor))` of length :obj:`config.n_layers` with each tuple having 4 tensors of shape :obj:`(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`): Contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding. If :obj:`past_key_values` are used, the user can optionally input only the last :obj:`decoder_input_ids` (those that don't have their past key value states given to this model) of shape :obj:`(batch_size, 1)` instead of all :obj:`decoder_input_ids` of shape :obj:`(batch_size, sequence_length)`. use_cache (:obj:`bool`, `optional`): If set to :obj:`True`, :obj:`past_key_values` key value states are returned and can be used to speed up decoding (see :obj:`past_key_values`). Returns: Example:: >>> from transformers import BertGenerationTokenizer, BertGenerationDecoder, BertGenerationConfig >>> import torch >>> tokenizer = BertGenerationTokenizer.from_pretrained('google/bert_for_seq_generation_L-24_bbc_encoder') >>> config = BertGenerationConfig.from_pretrained("google/bert_for_seq_generation_L-24_bbc_encoder") >>> config.is_decoder = True >>> model = BertGenerationDecoder.from_pretrained('google/bert_for_seq_generation_L-24_bbc_encoder', config=config) >>> inputs = tokenizer("Hello, my dog is cute", return_token_type_ids=False, return_tensors="pt") >>> outputs = model(**inputs) >>> prediction_logits = outputs.logits """ return_dict = return_dict if return_dict is not None else self.config.use_return_dict if labels is not None: use_cache = False outputs = self.bert( input_ids, attention_mask=attention_mask, position_ids=position_ids, head_mask=head_mask, inputs_embeds=inputs_embeds, encoder_hidden_states=encoder_hidden_states, encoder_attention_mask=encoder_attention_mask, past_key_values=past_key_values, use_cache=use_cache, output_attentions=output_attentions, output_hidden_states=output_hidden_states, return_dict=return_dict, ) sequence_output = outputs[0] prediction_scores = self.lm_head(sequence_output) lm_loss = None if labels is not None: # we are doing next-token prediction; shift prediction scores and input ids by one shifted_prediction_scores = prediction_scores[:, :-1, :].contiguous() labels = labels[:, 1:].contiguous() loss_fct = CrossEntropyLoss() lm_loss = loss_fct(shifted_prediction_scores.view(-1, self.config.vocab_size), labels.view(-1)) if not return_dict: output = (prediction_scores,) + outputs[1:] return ((lm_loss,) + output) if lm_loss is not None else output return CausalLMOutputWithCrossAttentions( loss=lm_loss, logits=prediction_scores, past_key_values=outputs.past_key_values, hidden_states=outputs.hidden_states, attentions=outputs.attentions, cross_attentions=outputs.cross_attentions, )
def prepare_inputs_for_generation(self, input_ids, past=None, attention_mask=None, **model_kwargs): input_shape = input_ids.shape # if model is used as a decoder in encoder-decoder model, the decoder attention mask is created on the fly if attention_mask is None: attention_mask = input_ids.new_ones(input_shape) # cut decoder_input_ids if past is used if past is not None: input_ids = input_ids[:, -1:] return {"input_ids": input_ids, "attention_mask": attention_mask, "past_key_values": past} def _reorder_cache(self, past, beam_idx): reordered_past = () for layer_past in past: reordered_past += (tuple(past_state.index_select(0, beam_idx) for past_state in layer_past),) return reordered_past