1053 lines
42 KiB
Python
1053 lines
42 KiB
Python
# coding=utf-8
|
|
# Copyright 2018 The HuggingFace Inc. team, Microsoft Corporation.
|
|
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
|
|
#
|
|
# 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 MPNet model."""
|
|
|
|
|
|
import math
|
|
from typing import Optional, Tuple, Union
|
|
|
|
import torch
|
|
from torch import nn
|
|
from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
|
|
|
|
from ...activations import ACT2FN, gelu
|
|
from ...modeling_outputs import (
|
|
BaseModelOutput,
|
|
BaseModelOutputWithPooling,
|
|
MaskedLMOutput,
|
|
MultipleChoiceModelOutput,
|
|
QuestionAnsweringModelOutput,
|
|
SequenceClassifierOutput,
|
|
TokenClassifierOutput,
|
|
)
|
|
from ...modeling_utils import PreTrainedModel
|
|
from ...pytorch_utils import find_pruneable_heads_and_indices, prune_linear_layer
|
|
from ...utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward, logging
|
|
from .configuration_mpnet import MPNetConfig
|
|
|
|
|
|
logger = logging.get_logger(__name__)
|
|
|
|
_CHECKPOINT_FOR_DOC = "microsoft/mpnet-base"
|
|
_CONFIG_FOR_DOC = "MPNetConfig"
|
|
|
|
|
|
from ..deprecated._archive_maps import MPNET_PRETRAINED_MODEL_ARCHIVE_LIST # noqa: F401, E402
|
|
|
|
|
|
class MPNetPreTrainedModel(PreTrainedModel):
|
|
config_class = MPNetConfig
|
|
base_model_prefix = "mpnet"
|
|
|
|
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)
|
|
|
|
|
|
class MPNetEmbeddings(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.padding_idx = 1
|
|
self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=self.padding_idx)
|
|
self.position_embeddings = nn.Embedding(
|
|
config.max_position_embeddings, config.hidden_size, padding_idx=self.padding_idx
|
|
)
|
|
|
|
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
self.register_buffer(
|
|
"position_ids", torch.arange(config.max_position_embeddings).expand((1, -1)), persistent=False
|
|
)
|
|
|
|
def forward(self, input_ids=None, position_ids=None, inputs_embeds=None, **kwargs):
|
|
if position_ids is None:
|
|
if input_ids is not None:
|
|
position_ids = create_position_ids_from_input_ids(input_ids, self.padding_idx)
|
|
else:
|
|
position_ids = self.create_position_ids_from_inputs_embeds(inputs_embeds)
|
|
|
|
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[:, :seq_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
|
|
|
|
def create_position_ids_from_inputs_embeds(self, inputs_embeds):
|
|
"""
|
|
We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.
|
|
|
|
Args:
|
|
inputs_embeds: torch.Tensor
|
|
|
|
Returns: torch.Tensor
|
|
"""
|
|
input_shape = inputs_embeds.size()[:-1]
|
|
sequence_length = input_shape[1]
|
|
|
|
position_ids = torch.arange(
|
|
self.padding_idx + 1, sequence_length + self.padding_idx + 1, dtype=torch.long, device=inputs_embeds.device
|
|
)
|
|
return position_ids.unsqueeze(0).expand(input_shape)
|
|
|
|
|
|
class MPNetSelfAttention(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
if config.hidden_size % config.num_attention_heads != 0 and not hasattr(config, "embedding_size"):
|
|
raise ValueError(
|
|
f"The hidden size ({config.hidden_size}) is not a multiple of the number of attention "
|
|
f"heads ({config.num_attention_heads})"
|
|
)
|
|
|
|
self.num_attention_heads = config.num_attention_heads
|
|
self.attention_head_size = int(config.hidden_size / config.num_attention_heads)
|
|
self.all_head_size = self.num_attention_heads * self.attention_head_size
|
|
|
|
self.q = nn.Linear(config.hidden_size, self.all_head_size)
|
|
self.k = nn.Linear(config.hidden_size, self.all_head_size)
|
|
self.v = nn.Linear(config.hidden_size, self.all_head_size)
|
|
self.o = nn.Linear(config.hidden_size, config.hidden_size)
|
|
|
|
self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
|
|
|
|
def transpose_for_scores(self, x):
|
|
new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.attention_head_size)
|
|
x = x.view(*new_x_shape)
|
|
return x.permute(0, 2, 1, 3)
|
|
|
|
def forward(
|
|
self,
|
|
hidden_states,
|
|
attention_mask=None,
|
|
head_mask=None,
|
|
position_bias=None,
|
|
output_attentions=False,
|
|
**kwargs,
|
|
):
|
|
q = self.q(hidden_states)
|
|
k = self.k(hidden_states)
|
|
v = self.v(hidden_states)
|
|
|
|
q = self.transpose_for_scores(q)
|
|
k = self.transpose_for_scores(k)
|
|
v = self.transpose_for_scores(v)
|
|
|
|
# Take the dot product between "query" and "key" to get the raw attention scores.
|
|
attention_scores = torch.matmul(q, k.transpose(-1, -2))
|
|
attention_scores = attention_scores / math.sqrt(self.attention_head_size)
|
|
|
|
# Apply relative position embedding (precomputed in MPNetEncoder) if provided.
|
|
if position_bias is not None:
|
|
attention_scores += position_bias
|
|
|
|
if attention_mask is not None:
|
|
attention_scores = attention_scores + attention_mask
|
|
|
|
# Normalize the attention scores to probabilities.
|
|
attention_probs = nn.functional.softmax(attention_scores, dim=-1)
|
|
|
|
attention_probs = self.dropout(attention_probs)
|
|
|
|
if head_mask is not None:
|
|
attention_probs = attention_probs * head_mask
|
|
|
|
c = torch.matmul(attention_probs, v)
|
|
|
|
c = c.permute(0, 2, 1, 3).contiguous()
|
|
new_c_shape = c.size()[:-2] + (self.all_head_size,)
|
|
c = c.view(*new_c_shape)
|
|
|
|
o = self.o(c)
|
|
|
|
outputs = (o, attention_probs) if output_attentions else (o,)
|
|
return outputs
|
|
|
|
|
|
class MPNetAttention(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.attn = MPNetSelfAttention(config)
|
|
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
|
|
self.pruned_heads = set()
|
|
|
|
def prune_heads(self, heads):
|
|
if len(heads) == 0:
|
|
return
|
|
heads, index = find_pruneable_heads_and_indices(
|
|
heads, self.attn.num_attention_heads, self.attn.attention_head_size, self.pruned_heads
|
|
)
|
|
|
|
self.attn.q = prune_linear_layer(self.attn.q, index)
|
|
self.attn.k = prune_linear_layer(self.attn.k, index)
|
|
self.attn.v = prune_linear_layer(self.attn.v, index)
|
|
self.attn.o = prune_linear_layer(self.attn.o, index, dim=1)
|
|
|
|
self.attn.num_attention_heads = self.attn.num_attention_heads - len(heads)
|
|
self.attn.all_head_size = self.attn.attention_head_size * self.attn.num_attention_heads
|
|
self.pruned_heads = self.pruned_heads.union(heads)
|
|
|
|
def forward(
|
|
self,
|
|
hidden_states,
|
|
attention_mask=None,
|
|
head_mask=None,
|
|
position_bias=None,
|
|
output_attentions=False,
|
|
**kwargs,
|
|
):
|
|
self_outputs = self.attn(
|
|
hidden_states,
|
|
attention_mask,
|
|
head_mask,
|
|
position_bias,
|
|
output_attentions=output_attentions,
|
|
)
|
|
attention_output = self.LayerNorm(self.dropout(self_outputs[0]) + hidden_states)
|
|
outputs = (attention_output,) + self_outputs[1:] # add attentions if we output them
|
|
return outputs
|
|
|
|
|
|
# Copied from transformers.models.bert.modeling_bert.BertIntermediate
|
|
class MPNetIntermediate(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
|
|
if isinstance(config.hidden_act, str):
|
|
self.intermediate_act_fn = ACT2FN[config.hidden_act]
|
|
else:
|
|
self.intermediate_act_fn = config.hidden_act
|
|
|
|
def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
|
|
hidden_states = self.dense(hidden_states)
|
|
hidden_states = self.intermediate_act_fn(hidden_states)
|
|
return hidden_states
|
|
|
|
|
|
# Copied from transformers.models.bert.modeling_bert.BertOutput
|
|
class MPNetOutput(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.dense = nn.Linear(config.intermediate_size, config.hidden_size)
|
|
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
|
|
def forward(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor) -> torch.Tensor:
|
|
hidden_states = self.dense(hidden_states)
|
|
hidden_states = self.dropout(hidden_states)
|
|
hidden_states = self.LayerNorm(hidden_states + input_tensor)
|
|
return hidden_states
|
|
|
|
|
|
class MPNetLayer(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.attention = MPNetAttention(config)
|
|
self.intermediate = MPNetIntermediate(config)
|
|
self.output = MPNetOutput(config)
|
|
|
|
def forward(
|
|
self,
|
|
hidden_states,
|
|
attention_mask=None,
|
|
head_mask=None,
|
|
position_bias=None,
|
|
output_attentions=False,
|
|
**kwargs,
|
|
):
|
|
self_attention_outputs = self.attention(
|
|
hidden_states,
|
|
attention_mask,
|
|
head_mask,
|
|
position_bias=position_bias,
|
|
output_attentions=output_attentions,
|
|
)
|
|
attention_output = self_attention_outputs[0]
|
|
outputs = self_attention_outputs[1:] # add self attentions if we output attention weights
|
|
|
|
intermediate_output = self.intermediate(attention_output)
|
|
layer_output = self.output(intermediate_output, attention_output)
|
|
outputs = (layer_output,) + outputs
|
|
return outputs
|
|
|
|
|
|
class MPNetEncoder(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.config = config
|
|
self.n_heads = config.num_attention_heads
|
|
self.layer = nn.ModuleList([MPNetLayer(config) for _ in range(config.num_hidden_layers)])
|
|
self.relative_attention_bias = nn.Embedding(config.relative_attention_num_buckets, self.n_heads)
|
|
|
|
def forward(
|
|
self,
|
|
hidden_states: torch.Tensor,
|
|
attention_mask: Optional[torch.Tensor] = None,
|
|
head_mask: Optional[torch.Tensor] = None,
|
|
output_attentions: bool = False,
|
|
output_hidden_states: bool = False,
|
|
return_dict: bool = False,
|
|
**kwargs,
|
|
):
|
|
position_bias = self.compute_position_bias(hidden_states)
|
|
all_hidden_states = () if output_hidden_states else None
|
|
all_attentions = () if output_attentions else None
|
|
for i, layer_module in enumerate(self.layer):
|
|
if output_hidden_states:
|
|
all_hidden_states = all_hidden_states + (hidden_states,)
|
|
|
|
layer_outputs = layer_module(
|
|
hidden_states,
|
|
attention_mask,
|
|
head_mask[i],
|
|
position_bias,
|
|
output_attentions=output_attentions,
|
|
**kwargs,
|
|
)
|
|
hidden_states = layer_outputs[0]
|
|
|
|
if output_attentions:
|
|
all_attentions = all_attentions + (layer_outputs[1],)
|
|
|
|
# Add last layer
|
|
if output_hidden_states:
|
|
all_hidden_states = all_hidden_states + (hidden_states,)
|
|
|
|
if not return_dict:
|
|
return tuple(v for v in [hidden_states, all_hidden_states, all_attentions] if v is not None)
|
|
return BaseModelOutput(
|
|
last_hidden_state=hidden_states,
|
|
hidden_states=all_hidden_states,
|
|
attentions=all_attentions,
|
|
)
|
|
|
|
def compute_position_bias(self, x, position_ids=None, num_buckets=32):
|
|
bsz, qlen, klen = x.size(0), x.size(1), x.size(1)
|
|
if position_ids is not None:
|
|
context_position = position_ids[:, :, None]
|
|
memory_position = position_ids[:, None, :]
|
|
else:
|
|
context_position = torch.arange(qlen, dtype=torch.long)[:, None]
|
|
memory_position = torch.arange(klen, dtype=torch.long)[None, :]
|
|
|
|
relative_position = memory_position - context_position
|
|
|
|
rp_bucket = self.relative_position_bucket(relative_position, num_buckets=num_buckets)
|
|
rp_bucket = rp_bucket.to(x.device)
|
|
values = self.relative_attention_bias(rp_bucket)
|
|
values = values.permute([2, 0, 1]).unsqueeze(0)
|
|
values = values.expand((bsz, -1, qlen, klen)).contiguous()
|
|
return values
|
|
|
|
@staticmethod
|
|
def relative_position_bucket(relative_position, num_buckets=32, max_distance=128):
|
|
ret = 0
|
|
n = -relative_position
|
|
|
|
num_buckets //= 2
|
|
ret += (n < 0).to(torch.long) * num_buckets
|
|
n = torch.abs(n)
|
|
|
|
max_exact = num_buckets // 2
|
|
is_small = n < max_exact
|
|
|
|
val_if_large = max_exact + (
|
|
torch.log(n.float() / max_exact) / math.log(max_distance / max_exact) * (num_buckets - max_exact)
|
|
).to(torch.long)
|
|
|
|
val_if_large = torch.min(val_if_large, torch.full_like(val_if_large, num_buckets - 1))
|
|
ret += torch.where(is_small, n, val_if_large)
|
|
return ret
|
|
|
|
|
|
# Copied from transformers.models.bert.modeling_bert.BertPooler
|
|
class MPNetPooler(nn.Module):
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.dense = nn.Linear(config.hidden_size, config.hidden_size)
|
|
self.activation = nn.Tanh()
|
|
|
|
def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
|
|
# We "pool" the model by simply taking the hidden state corresponding
|
|
# to the first token.
|
|
first_token_tensor = hidden_states[:, 0]
|
|
pooled_output = self.dense(first_token_tensor)
|
|
pooled_output = self.activation(pooled_output)
|
|
return pooled_output
|
|
|
|
|
|
MPNET_START_DOCSTRING = r"""
|
|
|
|
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](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 ([`MPNetConfig`]): 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 [`~PreTrainedModel.from_pretrained`] method to load the model weights.
|
|
"""
|
|
|
|
MPNET_INPUTS_DOCSTRING = r"""
|
|
Args:
|
|
input_ids (`torch.LongTensor` of shape `({0})`):
|
|
Indices of input sequence tokens in the vocabulary.
|
|
|
|
Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
|
|
[`PreTrainedTokenizer.__call__`] for details.
|
|
|
|
[What are input IDs?](../glossary#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?](../glossary#attention-mask)
|
|
position_ids (`torch.LongTensor` of shape `({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#position-ids)
|
|
head_mask (`torch.FloatTensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*):
|
|
Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:
|
|
|
|
- 1 indicates the head is **not masked**,
|
|
- 0 indicates the head is **masked**.
|
|
|
|
inputs_embeds (`torch.FloatTensor` of shape `({0}, 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 [`~utils.ModelOutput`] instead of a plain tuple.
|
|
"""
|
|
|
|
|
|
@add_start_docstrings(
|
|
"The bare MPNet Model transformer outputting raw hidden-states without any specific head on top.",
|
|
MPNET_START_DOCSTRING,
|
|
)
|
|
class MPNetModel(MPNetPreTrainedModel):
|
|
def __init__(self, config, add_pooling_layer=True):
|
|
super().__init__(config)
|
|
self.config = config
|
|
|
|
self.embeddings = MPNetEmbeddings(config)
|
|
self.encoder = MPNetEncoder(config)
|
|
self.pooler = MPNetPooler(config) if add_pooling_layer else None
|
|
|
|
# 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)
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=BaseModelOutputWithPooling,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
**kwargs,
|
|
) -> Union[Tuple[torch.Tensor], BaseModelOutputWithPooling]:
|
|
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 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:
|
|
self.warn_if_padding_and_no_attention_mask(input_ids, attention_mask)
|
|
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")
|
|
|
|
device = input_ids.device if input_ids is not None else inputs_embeds.device
|
|
|
|
if attention_mask is None:
|
|
attention_mask = torch.ones(input_shape, device=device)
|
|
extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(attention_mask, input_shape)
|
|
|
|
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)
|
|
encoder_outputs = self.encoder(
|
|
embedding_output,
|
|
attention_mask=extended_attention_mask,
|
|
head_mask=head_mask,
|
|
output_attentions=output_attentions,
|
|
output_hidden_states=output_hidden_states,
|
|
return_dict=return_dict,
|
|
)
|
|
sequence_output = encoder_outputs[0]
|
|
pooled_output = self.pooler(sequence_output) if self.pooler is not None else None
|
|
|
|
if not return_dict:
|
|
return (sequence_output, pooled_output) + encoder_outputs[1:]
|
|
|
|
return BaseModelOutputWithPooling(
|
|
last_hidden_state=sequence_output,
|
|
pooler_output=pooled_output,
|
|
hidden_states=encoder_outputs.hidden_states,
|
|
attentions=encoder_outputs.attentions,
|
|
)
|
|
|
|
|
|
class MPNetForMaskedLM(MPNetPreTrainedModel):
|
|
_tied_weights_keys = ["lm_head.decoder"]
|
|
|
|
def __init__(self, config):
|
|
super().__init__(config)
|
|
|
|
self.mpnet = MPNetModel(config, add_pooling_layer=False)
|
|
self.lm_head = MPNetLMHead(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
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=MaskedLMOutput,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
labels: Optional[torch.LongTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
) -> Union[Tuple[torch.Tensor], MaskedLMOutput]:
|
|
r"""
|
|
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]`
|
|
"""
|
|
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
|
|
|
outputs = self.mpnet(
|
|
input_ids,
|
|
attention_mask=attention_mask,
|
|
position_ids=position_ids,
|
|
head_mask=head_mask,
|
|
inputs_embeds=inputs_embeds,
|
|
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)
|
|
|
|
masked_lm_loss = None
|
|
if labels is not None:
|
|
loss_fct = CrossEntropyLoss()
|
|
masked_lm_loss = loss_fct(prediction_scores.view(-1, self.config.vocab_size), labels.view(-1))
|
|
|
|
if not return_dict:
|
|
output = (prediction_scores,) + outputs[2:]
|
|
return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output
|
|
|
|
return MaskedLMOutput(
|
|
loss=masked_lm_loss,
|
|
logits=prediction_scores,
|
|
hidden_states=outputs.hidden_states,
|
|
attentions=outputs.attentions,
|
|
)
|
|
|
|
|
|
class MPNetLMHead(nn.Module):
|
|
"""MPNet Head for masked and permuted language modeling."""
|
|
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.dense = nn.Linear(config.hidden_size, config.hidden_size)
|
|
self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
|
|
|
|
self.decoder = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
|
|
self.bias = nn.Parameter(torch.zeros(config.vocab_size))
|
|
|
|
# Need a link between the two variables so that the bias is correctly resized with `resize_token_embeddings`
|
|
self.decoder.bias = self.bias
|
|
|
|
def forward(self, features, **kwargs):
|
|
x = self.dense(features)
|
|
x = gelu(x)
|
|
x = self.layer_norm(x)
|
|
|
|
# project back to size of vocabulary with bias
|
|
x = self.decoder(x)
|
|
|
|
return x
|
|
|
|
|
|
@add_start_docstrings(
|
|
"""
|
|
MPNet Model transformer with a sequence classification/regression head on top (a linear layer on top of the pooled
|
|
output) e.g. for GLUE tasks.
|
|
""",
|
|
MPNET_START_DOCSTRING,
|
|
)
|
|
class MPNetForSequenceClassification(MPNetPreTrainedModel):
|
|
def __init__(self, config):
|
|
super().__init__(config)
|
|
|
|
self.num_labels = config.num_labels
|
|
self.mpnet = MPNetModel(config, add_pooling_layer=False)
|
|
self.classifier = MPNetClassificationHead(config)
|
|
|
|
# Initialize weights and apply final processing
|
|
self.post_init()
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=SequenceClassifierOutput,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
labels: Optional[torch.LongTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
) -> Union[Tuple[torch.Tensor], SequenceClassifierOutput]:
|
|
r"""
|
|
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).
|
|
"""
|
|
|
|
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
|
|
|
outputs = self.mpnet(
|
|
input_ids,
|
|
attention_mask=attention_mask,
|
|
position_ids=position_ids,
|
|
head_mask=head_mask,
|
|
inputs_embeds=inputs_embeds,
|
|
output_attentions=output_attentions,
|
|
output_hidden_states=output_hidden_states,
|
|
return_dict=return_dict,
|
|
)
|
|
sequence_output = outputs[0]
|
|
logits = self.classifier(sequence_output)
|
|
|
|
loss = None
|
|
if labels is not None:
|
|
if self.config.problem_type is None:
|
|
if self.num_labels == 1:
|
|
self.config.problem_type = "regression"
|
|
elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
|
|
self.config.problem_type = "single_label_classification"
|
|
else:
|
|
self.config.problem_type = "multi_label_classification"
|
|
|
|
if self.config.problem_type == "regression":
|
|
loss_fct = MSELoss()
|
|
if self.num_labels == 1:
|
|
loss = loss_fct(logits.squeeze(), labels.squeeze())
|
|
else:
|
|
loss = loss_fct(logits, labels)
|
|
elif self.config.problem_type == "single_label_classification":
|
|
loss_fct = CrossEntropyLoss()
|
|
loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
|
|
elif self.config.problem_type == "multi_label_classification":
|
|
loss_fct = BCEWithLogitsLoss()
|
|
loss = loss_fct(logits, labels)
|
|
if not return_dict:
|
|
output = (logits,) + outputs[2:]
|
|
return ((loss,) + output) if loss is not None else output
|
|
|
|
return SequenceClassifierOutput(
|
|
loss=loss,
|
|
logits=logits,
|
|
hidden_states=outputs.hidden_states,
|
|
attentions=outputs.attentions,
|
|
)
|
|
|
|
|
|
@add_start_docstrings(
|
|
"""
|
|
MPNet 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.
|
|
""",
|
|
MPNET_START_DOCSTRING,
|
|
)
|
|
class MPNetForMultipleChoice(MPNetPreTrainedModel):
|
|
def __init__(self, config):
|
|
super().__init__(config)
|
|
|
|
self.mpnet = MPNetModel(config)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
self.classifier = nn.Linear(config.hidden_size, 1)
|
|
|
|
# Initialize weights and apply final processing
|
|
self.post_init()
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=MultipleChoiceModelOutput,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
labels: Optional[torch.LongTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
) -> Union[Tuple[torch.Tensor], MultipleChoiceModelOutput]:
|
|
r"""
|
|
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)
|
|
"""
|
|
|
|
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
|
num_choices = input_ids.shape[1] if input_ids is not None else inputs_embeds.shape[1]
|
|
|
|
flat_input_ids = input_ids.view(-1, input_ids.size(-1)) if input_ids is not None else None
|
|
flat_position_ids = position_ids.view(-1, position_ids.size(-1)) if position_ids is not None else None
|
|
flat_attention_mask = attention_mask.view(-1, attention_mask.size(-1)) if attention_mask is not None else None
|
|
flat_inputs_embeds = (
|
|
inputs_embeds.view(-1, inputs_embeds.size(-2), inputs_embeds.size(-1))
|
|
if inputs_embeds is not None
|
|
else None
|
|
)
|
|
|
|
outputs = self.mpnet(
|
|
flat_input_ids,
|
|
position_ids=flat_position_ids,
|
|
attention_mask=flat_attention_mask,
|
|
head_mask=head_mask,
|
|
inputs_embeds=flat_inputs_embeds,
|
|
output_attentions=output_attentions,
|
|
output_hidden_states=output_hidden_states,
|
|
return_dict=return_dict,
|
|
)
|
|
pooled_output = outputs[1]
|
|
|
|
pooled_output = self.dropout(pooled_output)
|
|
logits = self.classifier(pooled_output)
|
|
reshaped_logits = logits.view(-1, num_choices)
|
|
|
|
loss = None
|
|
if labels is not None:
|
|
loss_fct = CrossEntropyLoss()
|
|
loss = loss_fct(reshaped_logits, labels)
|
|
|
|
if not return_dict:
|
|
output = (reshaped_logits,) + outputs[2:]
|
|
return ((loss,) + output) if loss is not None else output
|
|
|
|
return MultipleChoiceModelOutput(
|
|
loss=loss,
|
|
logits=reshaped_logits,
|
|
hidden_states=outputs.hidden_states,
|
|
attentions=outputs.attentions,
|
|
)
|
|
|
|
|
|
@add_start_docstrings(
|
|
"""
|
|
MPNet 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.
|
|
""",
|
|
MPNET_START_DOCSTRING,
|
|
)
|
|
class MPNetForTokenClassification(MPNetPreTrainedModel):
|
|
def __init__(self, config):
|
|
super().__init__(config)
|
|
self.num_labels = config.num_labels
|
|
|
|
self.mpnet = MPNetModel(config, add_pooling_layer=False)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
self.classifier = nn.Linear(config.hidden_size, config.num_labels)
|
|
|
|
# Initialize weights and apply final processing
|
|
self.post_init()
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=TokenClassifierOutput,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
labels: Optional[torch.LongTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
) -> Union[Tuple[torch.Tensor], TokenClassifierOutput]:
|
|
r"""
|
|
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]`.
|
|
"""
|
|
|
|
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
|
|
|
outputs = self.mpnet(
|
|
input_ids,
|
|
attention_mask=attention_mask,
|
|
position_ids=position_ids,
|
|
head_mask=head_mask,
|
|
inputs_embeds=inputs_embeds,
|
|
output_attentions=output_attentions,
|
|
output_hidden_states=output_hidden_states,
|
|
return_dict=return_dict,
|
|
)
|
|
|
|
sequence_output = outputs[0]
|
|
|
|
sequence_output = self.dropout(sequence_output)
|
|
logits = self.classifier(sequence_output)
|
|
|
|
loss = None
|
|
if labels is not None:
|
|
loss_fct = CrossEntropyLoss()
|
|
loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
|
|
|
|
if not return_dict:
|
|
output = (logits,) + outputs[2:]
|
|
return ((loss,) + output) if loss is not None else output
|
|
|
|
return TokenClassifierOutput(
|
|
loss=loss,
|
|
logits=logits,
|
|
hidden_states=outputs.hidden_states,
|
|
attentions=outputs.attentions,
|
|
)
|
|
|
|
|
|
class MPNetClassificationHead(nn.Module):
|
|
"""Head for sentence-level classification tasks."""
|
|
|
|
def __init__(self, config):
|
|
super().__init__()
|
|
self.dense = nn.Linear(config.hidden_size, config.hidden_size)
|
|
self.dropout = nn.Dropout(config.hidden_dropout_prob)
|
|
self.out_proj = nn.Linear(config.hidden_size, config.num_labels)
|
|
|
|
def forward(self, features, **kwargs):
|
|
x = features[:, 0, :] # take <s> token (equiv. to BERT's [CLS] token)
|
|
x = self.dropout(x)
|
|
x = self.dense(x)
|
|
x = torch.tanh(x)
|
|
x = self.dropout(x)
|
|
x = self.out_proj(x)
|
|
return x
|
|
|
|
|
|
@add_start_docstrings(
|
|
"""
|
|
MPNet 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`).
|
|
""",
|
|
MPNET_START_DOCSTRING,
|
|
)
|
|
class MPNetForQuestionAnswering(MPNetPreTrainedModel):
|
|
def __init__(self, config):
|
|
super().__init__(config)
|
|
|
|
self.num_labels = config.num_labels
|
|
self.mpnet = MPNetModel(config, add_pooling_layer=False)
|
|
self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)
|
|
|
|
# Initialize weights and apply final processing
|
|
self.post_init()
|
|
|
|
@add_start_docstrings_to_model_forward(MPNET_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
|
|
@add_code_sample_docstrings(
|
|
checkpoint=_CHECKPOINT_FOR_DOC,
|
|
output_type=QuestionAnsweringModelOutput,
|
|
config_class=_CONFIG_FOR_DOC,
|
|
)
|
|
def forward(
|
|
self,
|
|
input_ids: Optional[torch.LongTensor] = None,
|
|
attention_mask: Optional[torch.FloatTensor] = None,
|
|
position_ids: Optional[torch.LongTensor] = None,
|
|
head_mask: Optional[torch.FloatTensor] = None,
|
|
inputs_embeds: Optional[torch.FloatTensor] = None,
|
|
start_positions: Optional[torch.LongTensor] = None,
|
|
end_positions: Optional[torch.LongTensor] = None,
|
|
output_attentions: Optional[bool] = None,
|
|
output_hidden_states: Optional[bool] = None,
|
|
return_dict: Optional[bool] = None,
|
|
) -> Union[Tuple[torch.Tensor], QuestionAnsweringModelOutput]:
|
|
r"""
|
|
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.
|
|
"""
|
|
|
|
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
|
|
|
outputs = self.mpnet(
|
|
input_ids,
|
|
attention_mask=attention_mask,
|
|
position_ids=position_ids,
|
|
head_mask=head_mask,
|
|
inputs_embeds=inputs_embeds,
|
|
output_attentions=output_attentions,
|
|
output_hidden_states=output_hidden_states,
|
|
return_dict=return_dict,
|
|
)
|
|
|
|
sequence_output = outputs[0]
|
|
|
|
logits = self.qa_outputs(sequence_output)
|
|
start_logits, end_logits = logits.split(1, dim=-1)
|
|
start_logits = start_logits.squeeze(-1).contiguous()
|
|
end_logits = end_logits.squeeze(-1).contiguous()
|
|
|
|
total_loss = None
|
|
if start_positions is not None and end_positions is not None:
|
|
# If we are on multi-GPU, split add a dimension
|
|
if len(start_positions.size()) > 1:
|
|
start_positions = start_positions.squeeze(-1)
|
|
if len(end_positions.size()) > 1:
|
|
end_positions = end_positions.squeeze(-1)
|
|
# sometimes the start/end positions are outside our model inputs, we ignore these terms
|
|
ignored_index = start_logits.size(1)
|
|
start_positions = start_positions.clamp(0, ignored_index)
|
|
end_positions = end_positions.clamp(0, ignored_index)
|
|
|
|
loss_fct = CrossEntropyLoss(ignore_index=ignored_index)
|
|
start_loss = loss_fct(start_logits, start_positions)
|
|
end_loss = loss_fct(end_logits, end_positions)
|
|
total_loss = (start_loss + end_loss) / 2
|
|
|
|
if not return_dict:
|
|
output = (start_logits, end_logits) + outputs[2:]
|
|
return ((total_loss,) + output) if total_loss is not None else output
|
|
|
|
return QuestionAnsweringModelOutput(
|
|
loss=total_loss,
|
|
start_logits=start_logits,
|
|
end_logits=end_logits,
|
|
hidden_states=outputs.hidden_states,
|
|
attentions=outputs.attentions,
|
|
)
|
|
|
|
|
|
def create_position_ids_from_input_ids(input_ids, padding_idx):
|
|
"""
|
|
Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols
|
|
are ignored. This is modified from fairseq's `utils.make_positions`. :param torch.Tensor x: :return torch.Tensor:
|
|
"""
|
|
# The series of casts and type-conversions here are carefully balanced to both work with ONNX export and XLA.
|
|
mask = input_ids.ne(padding_idx).int()
|
|
incremental_indices = torch.cumsum(mask, dim=1).type_as(mask) * mask
|
|
return incremental_indices.long() + padding_idx
|