modeling_roberta_svo_graph.py

# coding=utf-8
# Copyright 2018 The Google AI Language Team Authors and The HuggingFace Inc. team.
# 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 RoBERTa model. """

import math
import dgl
import torch
import torch.nn as nn
import numpy as np
from torch.nn import init
import torch.utils.checkpoint
from torch.nn import CrossEntropyLoss, MSELoss
import torch.nn.functional as F
from gcn import RGCNModel

from transformers.activations import ACT2FN, gelu
from transformers.file_utils import (
    add_code_sample_docstrings,
    add_start_docstrings,
    add_start_docstrings_to_model_forward,
    replace_return_docstrings,
)
from transformers.modeling_outputs import (
    BaseModelOutputWithPastAndCrossAttentions,
    BaseModelOutputWithPoolingAndCrossAttentions,
    CausalLMOutputWithCrossAttentions,
    MaskedLMOutput,
    MultipleChoiceModelOutput,
    QuestionAnsweringModelOutput,
    SequenceClassifierOutput,
    TokenClassifierOutput,
)
from transformers.modeling_utils import (
    PreTrainedModel,
    apply_chunking_to_forward,
    find_pruneable_heads_and_indices,
    prune_linear_layer,
)
from transformers.utils import logging
from transformers import RobertaConfig


logger = logging.get_logger(__name__)

_CHECKPOINT_FOR_DOC = "roberta-base"
_CONFIG_FOR_DOC = "RobertaConfig"
_TOKENIZER_FOR_DOC = "RobertaTokenizer"

ROBERTA_PRETRAINED_MODEL_ARCHIVE_LIST = [
    "roberta-base",
    "roberta-large",
    "roberta-large-mnli",
    "distilroberta-base",
    "roberta-base-openai-detector",
    "roberta-large-openai-detector",
    # See all RoBERTa models at https://huggingface.co/models?filter=roberta
]


class TriLinear(nn.Module):

    def __init__(self, input_size):
        super(TriLinear, self).__init__()
        self.w1 = nn.Parameter(torch.FloatTensor(1, input_size))
        self.w2 = nn.Parameter(torch.FloatTensor(1, input_size))
        self.w3 = nn.Parameter(torch.FloatTensor(1, input_size))

        self.init_param()

    def forward(self, query, key):
        ndim = query.dim()
        q_logit = F.linear(query, self.w1)
        k_logit = F.linear(key, self.w2)

        shape = [1] * (ndim - 1) + [-1]
        dot_k = self.w3.view(shape) * key
        dot_logit = torch.matmul(query, torch.transpose(dot_k, -1, -2))

        logit = q_logit + torch.transpose(k_logit, -1, -2) + dot_logit
        return logit

    def init_param(self):
        init.normal_(self.w1, 0., 0.02)
        init.normal_(self.w2, 0., 0.02)
        init.normal_(self.w3, 0., 0.02)


class SCAttention(nn.Module) :
    def __init__(self, input_size, hidden_size) :
        super(SCAttention, self).__init__()
        self.hidden_size = hidden_size
        self.W = nn.Linear(input_size, hidden_size)
        self.map_linear = nn.Linear(hidden_size, hidden_size)
        #self.map_linear = nn.Linear(4 * hidden_size, hidden_size)
        #self.map_linear = nn.Linear(4 * hidden_size, hidden_size)
        self.init_weights()

    def init_weights(self) :
        nn.init.xavier_uniform_(self.W.weight.data)
        self.W.bias.data.fill_(0.1)

    #W_wq_relu_cox4_nolastRelu
    def forward(self, passage, p_mask, question, q_mask):
        Wp = F.relu(self.W(passage))
        Wq = F.relu(self.W(question))
        scores = torch.bmm(Wp, Wq.transpose(2, 1))
        # mask = q_mask.unsqueeze(1).repeat(1, passage.size(1), 1)
        q_mask = torch.transpose(q_mask, 1, 2)
        mask = q_mask.repeat(1, passage.size(1), 1)
        alpha = masked_softmax(scores, mask)
        output = torch.bmm(alpha, Wq)
        output = self.map_linear(output)
        return output


class GeneralContrastiveLoss(nn.Module):
    def __init__(self, nmax=1, margin=0.2):
        super(GeneralContrastiveLoss, self).__init__()
        self.margin = margin
        self.nmax = nmax

    def forward(self, logits_pos,logits_neg):
        hingeloss = torch.sum(torch.clamp(logits_neg + (0.2 - logits_pos).view(-1, 1).expand_as(logits_neg), min=0))
        loss = hingeloss

        return loss


class Attention(nn.Module):

    def __init__(self, sim):
        super(Attention, self).__init__()
        self.sim = sim

    def forward(self, query, key, value, query_mask=None, key_mask=None):
        ndim = query.dim()
        logit = self.sim(query, key)
        if query_mask is not None and key_mask is not None:
            mask = query_mask.unsqueeze(ndim - 1) * key_mask.unsqueeze(ndim - 2)
            logit = logit.masked_fill(~mask, -float('inf'))

        attn_weight = F.softmax(logit, dim=-1)
        if query_mask is not None and key_mask is not None:
            attn_weight = attn_weight.masked_fill(~mask, 0.)

        attn = torch.matmul(attn_weight, value)

        kq_weight = F.softmax(logit, dim=1)
        if query_mask is not None and key_mask is not None:
            kq_weight = kq_weight.masked_fill(~mask, 0.)

        co_weight = torch.matmul(attn_weight, torch.transpose(kq_weight, -1, -2))
        co_attn = torch.matmul(co_weight, query)

        return (attn, attn_weight), (co_attn, co_weight)

class AttentivePooling(nn.Module):

    def __init__(self, input_size):
        super(AttentivePooling, self).__init__()
        self.fc = nn.Linear(input_size, 1)

    def forward(self, input, mask):
        bsz, length, size = input.size()
        score = self.fc(input.contiguous().view(-1, size)).view(bsz, length)
        score = score.masked_fill(~mask, -float('inf'))
        prob = F.softmax(score, dim=-1)
        attn = torch.bmm(prob.unsqueeze(1), input)
        return attn


def masked_softmax(vector: torch.Tensor,
                   mask: torch.Tensor,
                   dim: int = -1,
                   memory_efficient: bool = False,
                   mask_fill_value: float = -1e32) -> torch.Tensor:
    if mask is None:
        result = torch.nn.functional.softmax(vector, dim=dim)
    else:
        mask = mask.float()
        #mask = mask.half()

        while mask.dim() < vector.dim():
            mask = mask.unsqueeze(1)
        if not memory_efficient:
            # To limit numerical errors from large vector elements outside the mask, we zero these out.
            result = torch.nn.functional.softmax(vector * mask, dim=dim)
            result = result * mask
            result = result / (result.sum(dim=dim, keepdim=True) + 1e-13)
        else:
            masked_vector = vector.masked_fill((1 - mask).byte(), mask_fill_value)
            result = torch.nn.functional.softmax(masked_vector, dim=dim)
    return result


class RobertaEmbeddings(nn.Module):
    """
    Same as BertEmbeddings with a tiny tweak for positional embeddings indexing.
    """

    # Copied from transformers.models.bert.modeling_bert.BertEmbeddings.__init__
    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.token_type_embeddings = nn.Embedding(config.type_vocab_size, 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)))
        self.position_embedding_type = getattr(config, "position_embedding_type", "absolute")

        # End copy
        self.padding_idx = config.pad_token_id
        self.position_embeddings = nn.Embedding(
            config.max_position_embeddings, config.hidden_size, padding_idx=self.padding_idx
        )

    def forward(
        self, input_ids=None, token_type_ids=None, position_ids=None, inputs_embeds=None, past_key_values_length=0
    ):
        if position_ids is None:
            if input_ids is not None:
                # Create the position ids from the input token ids. Any padded tokens remain padded.
                position_ids = create_position_ids_from_input_ids(
                    input_ids, self.padding_idx, past_key_values_length
                ).to(input_ids.device)
            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]

        if token_type_ids is None:
            token_type_ids = torch.zeros(input_shape, dtype=torch.long, device=self.position_ids.device)

        if inputs_embeds is None:
            inputs_embeds = self.word_embeddings(input_ids)
        token_type_embeddings = self.token_type_embeddings(token_type_ids)

        embeddings = inputs_embeds + token_type_embeddings
        if self.position_embedding_type == "absolute":
            position_embeddings = self.position_embeddings(position_ids)
            embeddings += 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)


# Copied from transformers.models.bert.modeling_bert.BertSelfAttention with Bert->Roberta
class RobertaSelfAttention(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.query = nn.Linear(config.hidden_size, self.all_head_size)
        self.key = nn.Linear(config.hidden_size, self.all_head_size)
        self.value = nn.Linear(config.hidden_size, self.all_head_size)

        self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
        self.position_embedding_type = getattr(config, "position_embedding_type", "absolute")
        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
            self.max_position_embeddings = config.max_position_embeddings
            self.distance_embedding = nn.Embedding(2 * config.max_position_embeddings - 1, self.attention_head_size)

        self.is_decoder = config.is_decoder

    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,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_value=None,
        output_attentions=False,
    ):
        mixed_query_layer = self.query(hidden_states)

        # If this is instantiated as a cross-attention module, the keys
        # and values come from an encoder; the attention mask needs to be
        # such that the encoder's padding tokens are not attended to.
        is_cross_attention = encoder_hidden_states is not None

        if is_cross_attention and past_key_value is not None:
            # reuse k,v, cross_attentions
            key_layer = past_key_value[0]
            value_layer = past_key_value[1]
            attention_mask = encoder_attention_mask
        elif is_cross_attention:
            key_layer = self.transpose_for_scores(self.key(encoder_hidden_states))
            value_layer = self.transpose_for_scores(self.value(encoder_hidden_states))
            attention_mask = encoder_attention_mask
        elif past_key_value is not None:
            key_layer = self.transpose_for_scores(self.key(hidden_states))
            value_layer = self.transpose_for_scores(self.value(hidden_states))
            key_layer = torch.cat([past_key_value[0], key_layer], dim=2)
            value_layer = torch.cat([past_key_value[1], value_layer], dim=2)
        else:
            key_layer = self.transpose_for_scores(self.key(hidden_states))
            value_layer = self.transpose_for_scores(self.value(hidden_states))

        query_layer = self.transpose_for_scores(mixed_query_layer)

        if self.is_decoder:
            # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states.
            # Further calls to cross_attention layer can then reuse all cross-attention
            # key/value_states (first "if" case)
            # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of
            # all previous decoder key/value_states. Further calls to uni-directional self-attention
            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
            # if encoder bi-directional self-attention `past_key_value` is always `None`
            past_key_value = (key_layer, value_layer)

        # Take the dot product between "query" and "key" to get the raw attention scores.
        attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))

        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
            seq_length = hidden_states.size()[1]
            position_ids_l = torch.arange(seq_length, dtype=torch.long, device=hidden_states.device).view(-1, 1)
            position_ids_r = torch.arange(seq_length, dtype=torch.long, device=hidden_states.device).view(1, -1)
            distance = position_ids_l - position_ids_r
            positional_embedding = self.distance_embedding(distance + self.max_position_embeddings - 1)
            positional_embedding = positional_embedding.to(dtype=query_layer.dtype)  # fp16 compatibility

            if self.position_embedding_type == "relative_key":
                relative_position_scores = torch.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
                attention_scores = attention_scores + relative_position_scores
            elif self.position_embedding_type == "relative_key_query":
                relative_position_scores_query = torch.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
                relative_position_scores_key = torch.einsum("bhrd,lrd->bhlr", key_layer, positional_embedding)
                attention_scores = attention_scores + relative_position_scores_query + relative_position_scores_key

        attention_scores = attention_scores / math.sqrt(self.attention_head_size)
        if attention_mask is not None:
            # Apply the attention mask is (precomputed for all layers in RobertaModel forward() function)
            attention_scores = attention_scores + attention_mask

        # Normalize the attention scores to probabilities.
        attention_probs = nn.Softmax(dim=-1)(attention_scores)

        # This is actually dropping out entire tokens to attend to, which might
        # seem a bit unusual, but is taken from the original Transformer paper.
        attention_probs = self.dropout(attention_probs)

        # Mask heads if we want to
        if head_mask is not None:
            attention_probs = attention_probs * head_mask

        context_layer = torch.matmul(attention_probs, value_layer)

        context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
        new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,)
        context_layer = context_layer.view(*new_context_layer_shape)

        outputs = (context_layer, attention_probs) if output_attentions else (context_layer,)

        if self.is_decoder:
            outputs = outputs + (past_key_value,)
        return outputs


# Copied from transformers.models.bert.modeling_bert.BertSelfOutput
class RobertaSelfOutput(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.dense = nn.Linear(config.hidden_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, input_tensor):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.dropout(hidden_states)
        hidden_states = self.LayerNorm(hidden_states + input_tensor)
        return hidden_states


# Copied from transformers.models.bert.modeling_bert.BertAttention with Bert->Roberta
class RobertaAttention(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.self = RobertaSelfAttention(config)
        self.output = RobertaSelfOutput(config)
        self.pruned_heads = set()

    def prune_heads(self, heads):
        if len(heads) == 0:
            return
        heads, index = find_pruneable_heads_and_indices(
            heads, self.self.num_attention_heads, self.self.attention_head_size, self.pruned_heads
        )

        # Prune linear layers
        self.self.query = prune_linear_layer(self.self.query, index)
        self.self.key = prune_linear_layer(self.self.key, index)
        self.self.value = prune_linear_layer(self.self.value, index)
        self.output.dense = prune_linear_layer(self.output.dense, index, dim=1)

        # Update hyper params and store pruned heads
        self.self.num_attention_heads = self.self.num_attention_heads - len(heads)
        self.self.all_head_size = self.self.attention_head_size * self.self.num_attention_heads
        self.pruned_heads = self.pruned_heads.union(heads)

    def forward(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_value=None,
        output_attentions=False,
    ):
        self_outputs = self.self(
            hidden_states,
            attention_mask,
            head_mask,
            encoder_hidden_states,
            encoder_attention_mask,
            past_key_value,
            output_attentions,
        )
        attention_output = self.output(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 RobertaIntermediate(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):
        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 RobertaOutput(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, input_tensor):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.dropout(hidden_states)
        hidden_states = self.LayerNorm(hidden_states + input_tensor)
        return hidden_states


# Copied from transformers.models.bert.modeling_bert.BertLayer with Bert->Roberta
class RobertaLayer(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.chunk_size_feed_forward = config.chunk_size_feed_forward
        self.seq_len_dim = 1
        self.attention = RobertaAttention(config)
        self.is_decoder = config.is_decoder
        self.add_cross_attention = config.add_cross_attention
        if self.add_cross_attention:
            assert self.is_decoder, f"{self} should be used as a decoder model if cross attention is added"
            self.crossattention = RobertaAttention(config)
        self.intermediate = RobertaIntermediate(config)
        self.output = RobertaOutput(config)

    def forward(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_value=None,
        output_attentions=False,
    ):
        # decoder uni-directional self-attention cached key/values tuple is at positions 1,2
        self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None
        self_attention_outputs = self.attention(
            hidden_states,
            attention_mask,
            head_mask,
            output_attentions=output_attentions,
            past_key_value=self_attn_past_key_value,
        )
        attention_output = self_attention_outputs[0]

        # if decoder, the last output is tuple of self-attn cache
        if self.is_decoder:
            outputs = self_attention_outputs[1:-1]
            present_key_value = self_attention_outputs[-1]
        else:
            outputs = self_attention_outputs[1:]  # add self attentions if we output attention weights

        cross_attn_present_key_value = None
        if self.is_decoder and encoder_hidden_states is not None:
            assert hasattr(
                self, "crossattention"
            ), f"If `encoder_hidden_states` are passed, {self} has to be instantiated with cross-attention layers by setting `config.add_cross_attention=True`"

            # cross_attn cached key/values tuple is at positions 3,4 of past_key_value tuple
            cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None
            cross_attention_outputs = self.crossattention(
                attention_output,
                attention_mask,
                head_mask,
                encoder_hidden_states,
                encoder_attention_mask,
                cross_attn_past_key_value,
                output_attentions,
            )
            attention_output = cross_attention_outputs[0]
            outputs = outputs + cross_attention_outputs[1:-1]  # add cross attentions if we output attention weights

            # add cross-attn cache to positions 3,4 of present_key_value tuple
            cross_attn_present_key_value = cross_attention_outputs[-1]
            present_key_value = present_key_value + cross_attn_present_key_value

        layer_output = apply_chunking_to_forward(
            self.feed_forward_chunk, self.chunk_size_feed_forward, self.seq_len_dim, attention_output
        )
        outputs = (layer_output,) + outputs

        # if decoder, return the attn key/values as the last output
        if self.is_decoder:
            outputs = outputs + (present_key_value,)

        return outputs

    def feed_forward_chunk(self, attention_output):
        intermediate_output = self.intermediate(attention_output)
        layer_output = self.output(intermediate_output, attention_output)
        return layer_output


# Copied from transformers.models.bert.modeling_bert.BertEncoder with Bert->Roberta
class RobertaEncoder(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.config = config
        self.layer = nn.ModuleList([RobertaLayer(config) for _ in range(config.num_hidden_layers)])

    def forward(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_values=None,
        use_cache=None,
        output_attentions=False,
        output_hidden_states=False,
        return_dict=True,
    ):
        all_hidden_states = () if output_hidden_states else None
        all_self_attentions = () if output_attentions else None
        all_cross_attentions = () if output_attentions and self.config.add_cross_attention else None

        next_decoder_cache = () if use_cache else None
        for i, layer_module in enumerate(self.layer):
            if output_hidden_states:
                all_hidden_states = all_hidden_states + (hidden_states,)

            layer_head_mask = head_mask[i] if head_mask is not None else None
            past_key_value = past_key_values[i] if past_key_values is not None else None

            if getattr(self.config, "gradient_checkpointing", False) and self.training:

                if use_cache:
                    logger.warn(
                        "`use_cache=True` is incompatible with `config.gradient_checkpointing=True`. Setting "
                        "`use_cache=False`..."
                    )
                    use_cache = False

                def create_custom_forward(module):
                    def custom_forward(*inputs):
                        return module(*inputs, past_key_value, output_attentions)

                    return custom_forward

                layer_outputs = torch.utils.checkpoint.checkpoint(
                    create_custom_forward(layer_module),
                    hidden_states,
                    attention_mask,
                    layer_head_mask,
                    encoder_hidden_states,
                    encoder_attention_mask,
                )
            else:
                layer_outputs = layer_module(
                    hidden_states,
                    attention_mask,
                    layer_head_mask,
                    encoder_hidden_states,
                    encoder_attention_mask,
                    past_key_value,
                    output_attentions,
                )

            hidden_states = layer_outputs[0]
            if use_cache:
                next_decoder_cache += (layer_outputs[-1],)
            if output_attentions:
                all_self_attentions = all_self_attentions + (layer_outputs[1],)
                if self.config.add_cross_attention:
                    all_cross_attentions = all_cross_attentions + (layer_outputs[2],)

        if output_hidden_states:
            all_hidden_states = all_hidden_states + (hidden_states,)

        if not return_dict:
            return tuple(
                v
                for v in [
                    hidden_states,
                    next_decoder_cache,
                    all_hidden_states,
                    all_self_attentions,
                    all_cross_attentions,
                ]
                if v is not None
            )
        return BaseModelOutputWithPastAndCrossAttentions(
            last_hidden_state=hidden_states,
            past_key_values=next_decoder_cache,
            hidden_states=all_hidden_states,
            attentions=all_self_attentions,
            cross_attentions=all_cross_attentions,
        )


# Copied from transformers.models.bert.modeling_bert.BertPooler
class RobertaPooler(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):
        # 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


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

    config_class = RobertaConfig
    base_model_prefix = "roberta"

    # Copied from transformers.models.bert.modeling_bert.BertPreTrainedModel._init_weights
    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)


ROBERTA_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.RobertaConfig`): 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.
"""

ROBERTA_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.RobertaTokenizer`. See
            :meth:`transformers.PreTrainedTokenizer.encode` and :meth:`transformers.PreTrainedTokenizer.__call__` 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>`__
        token_type_ids (:obj:`torch.LongTensor` of shape :obj:`({0})`, `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? <../glossary.html#token-type-ids>`_
        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.
"""


@add_start_docstrings(
    "The bare RoBERTa Model transformer outputting raw hidden-states without any specific head on top.",
    ROBERTA_START_DOCSTRING,
)
class RobertaModel(RobertaPreTrainedModel):
    """
    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`_ by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz
    Kaiser and Illia Polosukhin.
    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.
    .. _`Attention is all you need`: https://arxiv.org/abs/1706.03762
    """

    _keys_to_ignore_on_load_missing = [r"position_ids"]

    # Copied from transformers.models.bert.modeling_bert.BertModel.__init__ with Bert->Roberta
    def __init__(self, config, add_pooling_layer=True):
        super().__init__(config)
        self.config = config

        self.embeddings = RobertaEmbeddings(config)
        self.encoder = RobertaEncoder(config)

        self.pooler = RobertaPooler(config) if add_pooling_layer else None


        self.init_weights()

    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(ROBERTA_INPUTS_DOCSTRING.format("(batch_size, sequence_length)"))
    @add_code_sample_docstrings(
        #tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=BaseModelOutputWithPoolingAndCrossAttentions,
        config_class=_CONFIG_FOR_DOC,
    )
    # Copied from transformers.models.bert.modeling_bert.BertModel.forward
    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        token_type_ids=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 tokens that are **masked**.
        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()
            batch_size, seq_length = input_shape
        elif inputs_embeds is not None:
            input_shape = inputs_embeds.size()[:-1]
            batch_size, seq_length = input_shape
        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

        # 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)
        if token_type_ids is None:
            token_type_ids = torch.zeros(input_shape, dtype=torch.long, 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: 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,
            token_type_ids=token_type_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]
        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 BaseModelOutputWithPoolingAndCrossAttentions(
            last_hidden_state=sequence_output,
            pooler_output=pooled_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,
        )


@add_start_docstrings(
    """RoBERTa Model with a `language modeling` head on top for CLM fine-tuning. """, ROBERTA_START_DOCSTRING
)
class RobertaForCausalLM(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_missing = [r"position_ids", r"lm_head.decoder.bias"]
    _keys_to_ignore_on_load_unexpected = [r"pooler"]

    def __init__(self, config):
        super().__init__(config)

        if not config.is_decoder:
            logger.warning("If you want to use `RobertaLMHeadModel` as a standalone, add `is_decoder=True.`")

        self.roberta = RobertaModel(config, add_pooling_layer=False)
        self.lm_head = RobertaLMHead(config)

        self.init_weights()

    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(ROBERTA_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,
        token_type_ids=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 RobertaTokenizer, RobertaForCausalLM, RobertaConfig
            >>> import torch
            >>> tokenizer = RobertaTokenizer.from_pretrained('roberta-base')
            >>> config = RobertaConfig.from_pretrained("roberta-base")
            >>> config.is_decoder = True
            >>> model = RobertaForCausalLM.from_pretrained('roberta-base', config=config)
            >>> inputs = tokenizer("Hello, my dog is cute", 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.roberta(
            input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            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[2:]
            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


@add_start_docstrings("""RoBERTa Model with a `language modeling` head on top. """, ROBERTA_START_DOCSTRING)
class RobertaForMaskedLM(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_missing = [r"position_ids", r"lm_head.decoder.bias"]
    _keys_to_ignore_on_load_unexpected = [r"pooler"]

    def __init__(self, config):
        super().__init__(config)

        if config.is_decoder:
            logger.warning(
                "If you want to use `RobertaForMaskedLM` make sure `config.is_decoder=False` for "
                "bi-directional self-attention."
            )

        self.roberta = RobertaModel(config, add_pooling_layer=False)
        self.lm_head = RobertaLMHead(config)

        self.init_weights()

    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(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        #tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=MaskedLMOutput,
        config_class=_CONFIG_FOR_DOC,
        mask="<mask>",
    )
    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        token_type_ids=None,
        position_ids=None,
        head_mask=None,
        inputs_embeds=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        labels=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
    ):
        r"""
        labels (:obj:`torch.LongTensor` of shape :obj:`(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]``
        kwargs (:obj:`Dict[str, any]`, optional, defaults to `{}`):
            Used to hide legacy arguments that have been deprecated.
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.roberta(
            input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            encoder_hidden_states=encoder_hidden_states,
            encoder_attention_mask=encoder_attention_mask,
            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 RobertaLMHead(nn.Module):
    """Roberta Head for masked 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(
    """
    RoBERTa Model transformer with a sequence classification/regression head on top (a linear layer on top of the
    pooled output) e.g. for GLUE tasks.
    """,
    ROBERTA_START_DOCSTRING,
)
class RobertaForSequenceClassification(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels

        self.roberta = RobertaModel(config, add_pooling_layer=False)
        self.classifier = RobertaClassificationHead(config)

        self.init_weights()

    @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        #tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=SequenceClassifierOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        token_type_ids=None,
        position_ids=None,
        head_mask=None,
        inputs_embeds=None,
        labels=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
    ):
        r"""
        labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`):
            Labels for computing the sequence classification/regression loss. Indices should be in :obj:`[0, ...,
            config.num_labels - 1]`. If :obj:`config.num_labels == 1` a regression loss is computed (Mean-Square loss),
            If :obj:`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.roberta(
            input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            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.num_labels == 1:
                #  We are doing regression
                loss_fct = MSELoss()
                loss = loss_fct(logits.view(-1), labels.view(-1))
            else:
                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 SequenceClassifierOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )


@add_start_docstrings(
    """
    Roberta 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.
    """,
    ROBERTA_START_DOCSTRING,
)
class RobertaForMultipleChoice(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config, max_doc_len, max_query_len, max_option_len):
        super(RobertaForMultipleChoice, self).__init__(config)

        self.roberta = RobertaModel(config)
        self.num_labels = 4
        self.max_doc_len = max_doc_len
        self.max_query_len = max_query_len
        self.max_option_len = max_option_len
        self.dropout1 = nn.Dropout(config.hidden_dropout_prob)
        self.dropout2 = nn.Dropout(config.hidden_dropout_prob)
        self.classifier = nn.Linear(config.hidden_size, 1)

        self.hidden_size = config.hidden_size
        self.score_fc = nn.Linear(config.hidden_size, 1)
        self.attn_sim = TriLinear(config.hidden_size)
        self.attention = Attention(sim=self.attn_sim)
        self.attn_fc = nn.Linear(config.hidden_size * 3, config.hidden_size, bias=True)
        self.GCN = RGCNModel(config.hidden_size, 5, 1, True)
        self.GCN.to(self.device)
        
        self.opt_attn_sim = TriLinear(config.hidden_size)
        self.comp_fc = nn.Linear(config.hidden_size*7, config.hidden_size, bias=True)
        self.opt_attention = Attention(sim=self.opt_attn_sim)
        
        self.opt_selfattn_sim = TriLinear(config.hidden_size)
        self.opt_self_attention = Attention(sim=self.opt_selfattn_sim)
        self.opt_selfattn_fc = nn.Linear(config.hidden_size * 4, config.hidden_size, bias=True)

        self.gate_fc = nn.Linear(config.hidden_size * 3, config.hidden_size, bias=True)
        self.query_attentive_pooling = AttentivePooling(input_size=config.hidden_size)

        self.w_selfattn = nn.Linear(config.hidden_size, 1, bias=True)
        self.w_output = nn.Linear(config.hidden_size, 1, bias=True)
        self.dropout = nn.Dropout(config.hidden_dropout_prob)

        self.proj_attention = SCAttention(config.hidden_size, config.hidden_size)
        self.pooler = RobertaPooler(config)

        self.init_weights()
        self.svo_weight = 0.5
    
    # def option_attention(self, U, V):
    #     S = torch.zeros(U.size(0), V.size(0))
    #     for i in range(U.size(0)):
    #         for j in range(V.size(0)):
    #             cat_result = torch.cat((U[i], V[j]))
    #             cat_result = torch.cat((cat_result, U[i]*V[j])).unsqueeze(1)    
    #             S[i][j] = torch.matmul(self.v,cat_result)
    #     A = torch.softmax(S, dim=0)
    #     assert A.shape == S.shape
    #     return A.to(self.device)             # [U.shape[0], V.shape[0]]

    # def gather_option(self, subset):
    #     subset_example_option = []
    #     for idx_i, i in enumerate(subset):
    #         cur_option = i
    #         for idx_j, j in enumerate(subset):
    #             if idx_j != idx_i:
    #                 attention_result = torch.matmul(j.T, self.option_attention(j, i)).T
    #                 attention_result = torch.cat((i-attention_result, i * attention_result), dim=1)
    #                 cur_option = torch.cat((cur_option, attention_result), dim=1)
    #             else:
    #                 continue
    #         subset_example_option.append(torch.tanh(self.W_c(cur_option)))
    #     return subset_example_option


    # def correlated_option(self, option_rep, num_choice=4):
    #     correlated_option = []
    #     for i in range(0, len(option_rep), num_choice):
    #         subset_per_example = option_rep[i: i + num_choice]
    #         subset_example_option = self.gather_option(subset_per_example)
    #         correlated_option.append(item for item in subset_example_option)
    #     return correlated_option

    def split_bert_sequence(self, seq, seq1_lengths, max_seq1_length, seq2_lengths, max_seq2_length, pad = 0, has_cls = True):
        if has_cls:
            cls = seq[:, 0, :]
        else:
            cls = None
        begin_index = 1 if has_cls else 0
        seq1 = seq[:, begin_index: max_seq1_length + begin_index, :]
        seq1_mask = self.sequence_mask(seq1_lengths, max_seq1_length)
        seq1 = seq1.float().masked_fill(
            ~seq1_mask.unsqueeze(2),
            float(pad),
        ).type_as(seq1)
        seq_range = torch.arange(0, max_seq2_length).long().unsqueeze(0)
        if seq1_lengths.is_cuda:
            seq_range = seq_range.cuda()
        seq_index = seq_range + seq1_lengths.unsqueeze(1) + 1 + begin_index
        batch_size, index_len = seq_index.size()
        dim = seq.size()[2]
        seq2 = torch.gather(seq, dim=1, index=seq_index.unsqueeze(2).expand(batch_size, index_len, dim))
        seq2_mask = self.sequence_mask(seq2_lengths, max_seq2_length)
        seq2 = seq2.float().masked_fill(
            ~seq2_mask.unsqueeze(2),
            float(pad),
        ).type_as(seq2)
        return cls, seq1, seq2

    def sequence_mask(self, sequence_length, max_len=None):
        if max_len is None:
            max_len = sequence_length.data.max()
        batch_size = sequence_length.size(0)
        seq_range = torch.arange(0, max_len).long()
        seq_range_expand = seq_range.unsqueeze(0).expand(batch_size, max_len)
        if sequence_length.is_cuda:
            seq_range_expand = seq_range_expand.cuda()
        seq_length_expand = (sequence_length.unsqueeze(1).expand_as(seq_range_expand))
        return seq_range_expand < seq_length_expand

    def logic_op(self, input, input_mask):
        selfattn_unmask = self.w_selfattn(self.dropout(input))
        selfattn_unmask.masked_fill_(~input_mask, -float('inf'))
        selfattn_weight = F.softmax(selfattn_unmask, dim=1)
        selfattn = torch.sum(selfattn_weight * input, dim=1)
        score = self.w_output(self.dropout(selfattn))
        return score


    @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length"))
    @add_code_sample_docstrings(
        #tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=MultipleChoiceModelOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids=None,        #[batch_size, num_choice, max_seq_len]
        token_type_ids=None,
        attention_mask=None,
        labels=None,
        position_ids=None,
        head_mask=None,
        inputs_embeds=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
        # choice_ids=None,
        query_len=None,
        opt_len=None,
        svo_ids=None,
    ):
        r"""
        labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`):
            Labels for computing the multiple choice classification loss. Indices should be in ``[0, ...,
            num_choices-1]`` where :obj:`num_choices` is the size of the second dimension of the input tensors. (See
            :obj:`input_ids` above)
        """
        max_svo_len = svo_ids.size(2)
        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]
        bsz = input_ids.size(0)
        flat_input_ids = input_ids.view(-1, input_ids.size(-1)) if input_ids is not None else None    
        svo_graph_ids = svo_ids.view(bsz * self.num_labels, max_svo_len, 3, 2)
        flat_svo_ids = svo_ids.view(bsz * self.num_labels, -1, 2)
        flat_position_ids = position_ids.view(-1, position_ids.size(-1)) if position_ids is not None else None
        flat_token_type_ids = token_type_ids.view(-1, token_type_ids.size(-1)) if token_type_ids is not None else None
        # flat_choice_ids = choice_ids.view(-1, choice_ids.size(-1)) if choice_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
        )
        # after flat: [bz*num_choice, max_seq_length]

        outputs = self.roberta(
            flat_input_ids,
            position_ids=flat_position_ids,
            token_type_ids=flat_token_type_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,
        )
        sequence_output = outputs[0] # outputs[0].shape: [bz*num_choice, max_seq_length, dim]
        pooled_output = outputs[1]
        # sequence_output = pooled_output
        # assert sequence_output.size(0) == flat_choice_ids.size(0)
        # assert sequence_output.size(1) == flat_choice_ids.size(1)
        sequence_len = sequence_output.size(1)


        ### SVO triplet implementation
        batch_id = 0
        batch_svo_ids = []

        for batch in flat_svo_ids:
            offset = batch_id * sequence_len + 1
            svo_seqs = [offset + int(item[0]) for item in batch if int(item[0]) != -100]
            while (len(svo_seqs) < max_svo_len * 3):
                svo_seqs.append(0)
            batch_svo_ids.append(svo_seqs)
            batch_id += 1

        batch_svo_ids = torch.tensor(batch_svo_ids)
        batch_svo_ids = batch_svo_ids.view(-1)


        ### averaged svo_loss
        # for batch in flat_svo_ids:
        #     offset = batch_id * sequence_len + 1
        #     svo_seqs = [[offset + item[0], item[1]] for item in batch if item[0] != -100]
        #     while (len(svo_seqs) < max_svo_len * 3):
        #         svo_seqs.append([0, 0])
        #     batch_svo_ids.append(svo_seqs)
        #     batch_id += 1
        # batch_svo_ids = torch.tensor(batch_svo_ids)  # [bz*num_choice*max_svo_len *3]
        # batch_svo_ids = batch_svo_ids.view(-1, 2)

        sequence_output_svo = sequence_output.view(-1, self.hidden_size)
        sequence_output_svo = torch.cat([sequence_output_svo.new_zeros((1, self.hidden_size)), sequence_output_svo], dim=0)

        batch_svo_ids = batch_svo_ids.cuda()
        # svo_features = []
        # for item in batch_svo_ids:
        #     index = torch.tensor([i for i in range(item[0], item[0] + item[1])], dtype=torch.long).to(self.device)
        #     value = sequence_output_svo.index_select(0, index)
        #     if value.size(0) == 0:
        #         print(value.shape)
        #         value = torch.tensor([value])
        #     if value.size(0) != 0 and value.size(0) != 1:
        #         value = torch.sum(value, dim=0)/int(item[1]) 
        #     svo_features.append(value)
        # svo_features = torch.cat(svo_features,dim=0).cuda()
        svo_features = sequence_output_svo.index_select(0, batch_svo_ids)
        svo_features = svo_features.view(bsz * self.num_labels, max_svo_len, 3, self.hidden_size)
        # eou_features = eou_features.cuda()

        sv_features = svo_features[:, :, 0, :] + svo_features[:, :, 1, :]
        o_features = svo_features[:, :, 2, :]

        ### SVO graph implementation
        svo_rep = []
        svo_mask = []
        for choice_idx, svo in enumerate(svo_graph_ids):
            G = dgl.DGLGraph().to(self.device)
            edge_type = []
            edges = []
            nodes_all = []
            for item in svo.view(-1,  2):
                if list(item) not in nodes_all:
                    nodes_all.append(list(item))
            if [-100, 0] in nodes_all:
                nodes_all.remove([-100,0])
            G.add_nodes(len(nodes_all))

            if nodes_all == []:
                svo = torch.tensor([[[0, 1], [self.max_doc_len +2, 1], [self.max_doc_len + self.max_query_len + self.max_option_len +4, 1]]]).to(self.device)
                nodes_all = [list(item) for item in svo.view(-1, 2)]
            

            for item in svo:
                if int(item[0][0]) == -100:
                    break
                else:
                    # add default_in and default_out edges
                    G.add_edges(nodes_all.index(list(item[0])), nodes_all.index(list(item[1])))
                    edges.append([nodes_all.index(list(item[0])), nodes_all.index(list(item[1]))])
                    edge_type.append(0)

                    G.add_edges(nodes_all.index(list(item[1])), nodes_all.index(list(item[0])))
                    edges.append([nodes_all.index(list(item[1])), nodes_all.index(list(item[0]))])
                    edge_type.append(1)

                    G.add_edges(nodes_all.index(list(item[1])), nodes_all.index(list(item[2])))
                    edges.append([nodes_all.index(list(item[1])), nodes_all.index(list(item[2]))])
                    edge_type.append(2)

                    G.add_edges(nodes_all.index(list(item[2])), nodes_all.index(list(item[1])))
                    edges.append([nodes_all.index(list(item[2])), nodes_all.index(list(item[1]))])
                    edge_type.append(3)
            # add self edges
            for item in range(len(nodes_all)):
                G.add_edge(item, item)
                edge_type.append(4)
                edges.append([item, item])

            # add node feature
            for i in range(len(nodes_all)):
                index = torch.tensor([i for i in range(nodes_all[i][0], nodes_all[i][0] + nodes_all[i][1])], dtype=torch.long).to(self.device)
                selected_item = torch.index_select(sequence_output[choice_idx], 0, index)
                if selected_item.size(0) == 1:
                    value = selected_item
                else:
                    value = (torch.sum(selected_item, dim=0)/nodes_all[i][1]).unsqueeze(0)
                G.nodes[[i]].data['h'] = value
            
            edge_norm = []
            for e1, e2 in edges:
                if e1 == e2:
                    edge_norm.append(1)
                else:
                    edge_norm.append(1/(G.in_degrees(e2)-1))
            
            edge_type = torch.from_numpy(np.array(edge_type)).to(self.device)
            edge_norm = torch.from_numpy(np.array(edge_norm)).unsqueeze(1).float().to(self.device)

            G.edata.update({'rel_type':edge_type,})
            G.edata.update({'norm': edge_norm})
            X = self.GCN(G)[0]
            svo_rep.append(X)
            svo_mask.append(torch.tensor([1] * X.size(0), dtype=torch.bool))
        svo_rep = torch.nn.utils.rnn.pad_sequence(svo_rep).to(self.device)
        svo_rep = torch.transpose(svo_rep, 0, 1).contiguous()
        svo_mask = torch.nn.utils.rnn.pad_sequence(svo_mask).to(self.device).unsqueeze(2)
        svo_mask = torch.transpose(svo_mask, 0, 1).contiguous()
        # final_rep = self.logic_op(svo_rep, svo_mask)
        output = self.proj_attention(sequence_output, flat_attention_mask.unsqueeze(2), svo_rep, svo_mask)
        final_rep = self.pooler(output)

        ### OCN implementation
        doc_enc = sequence_output[:, 0 : self.max_doc_len + 2, :]
        opt_enc = sequence_output[:, self.max_doc_len + 2:, :]

        opt_len = opt_len.view(-1)
        query_len = query_len.view(-1)

        _, query_enc, _ = self.split_bert_sequence(opt_enc, query_len, self.max_query_len, opt_len, self.max_option_len, has_cls=True)

        # options_rep = []
        # for idx, item in enumerate(sequence_output):
        #     select_index =[]
        #     for id, items in enumerate(flat_choice_ids[idx]):
        #         if items == 1:
        #             select_index.append(id)
        #         else:
        #             continue
        #     select_index = torch.LongTensor(select_index).cuda()
        #     selected_option = torch.index_select(item, 0, select_index)
        #     options_rep.append(selected_option)
        # correlated_option_rep = self.correlated_option(options_rep, num_choices)

        query_total_len = query_enc.size(1)
        query_mask = self.sequence_mask(query_len, query_total_len)
        query_attn = self.query_attentive_pooling(query_enc, query_mask)

        opt_total_len = opt_enc.size(1)
        opt_mask = flat_attention_mask[:, self.max_doc_len + 2:] > 0
        doc_mask = flat_attention_mask[:, 0 : self.max_doc_len + 2] > 0
        
        opt_mask = opt_mask.view(bsz, self.num_labels, opt_total_len)
        opt_enc = opt_enc.view(bsz, self.num_labels, opt_total_len, self.hidden_size)

        # Option Comparison
        correlation_list = []
        for i in range(self.num_labels):
            cur_opt = opt_enc[:, i, :, :]
            cur_mask = opt_mask[:, i, :]

            comp_info = []
            for j in range(self.num_labels):
                if j == i:
                    continue

                tmp_opt = opt_enc[:, j, :, :]
                tmp_mask = opt_mask[:, j, :]

                (attn, _), _ = self.opt_attention(cur_opt, tmp_opt, tmp_opt, cur_mask, tmp_mask)
                comp_info.append(cur_opt * attn)
                comp_info.append(cur_opt - attn)

            correlation = torch.tanh(self.comp_fc(torch.cat([cur_opt] + comp_info, dim=-1)))
            correlation_list.append(correlation)

        correlation_list = [correlation.unsqueeze(1) for correlation in correlation_list]
        opt_correlation = torch.cat(correlation_list, dim=1)

        opt_mask = opt_mask.view(bsz * self.num_labels, opt_total_len)
        opt_enc = opt_enc.view(bsz * self.num_labels, opt_total_len, self.hidden_size)
        opt_correlation = opt_correlation.contiguous().view(bsz * self.num_labels, opt_total_len, self.hidden_size)
        gate = torch.sigmoid(self.gate_fc(torch.cat((opt_enc, opt_correlation, query_attn.expand_as(opt_enc)), -1)))
        option = opt_enc * gate + opt_correlation * (1.0 - gate)

        (attn, _), (coattn, _) = self.attention(option, doc_enc, doc_enc, opt_mask, doc_mask)
        fusion = self.attn_fc(torch.cat((option, attn, coattn), -1))
        fusion = F.relu(fusion)

        (attn, _), _ = self.opt_self_attention(fusion, fusion, fusion, opt_mask, opt_mask)
        fusion = self.opt_selfattn_fc(torch.cat((fusion, attn, fusion * attn, fusion - attn), -1))
        fusion = F.relu(fusion)

        fusion = fusion.masked_fill(
            ~opt_mask.unsqueeze(-1).expand(bsz * self.num_labels, opt_total_len, self.hidden_size),
            -float('inf'))
        fusion, _ = fusion.max(dim=1)

        fusion = self.dropout2(fusion)
        reshaped_logits = self.score_fc(fusion).view(bsz, self.num_labels)

        # pooled_output = self.dropout1(pooled_output)
        # logits = self.classifier(pooled_output)
        # reshaped_logits = logits.view(-1, num_choices)
        # svo_logits = final_rep.view(-1, num_choices)
        # reshaped_logits = (svo_logits.view(-1, num_choices) + reshaped_logits)/2
        logits = self.classifier(final_rep)
        reshaped_logits = logits.view(-1, num_choices)

        loss = None
        if labels is not None:
            # binary_labels = torch.zeros_like(reshaped_logits)
            # for bid, choice in enumerate(labels):
            #     binary_labels[bid][choice] = 1
            
            loss_fct = CrossEntropyLoss()
            cls_loss = loss_fct(reshaped_logits, labels)
            criterion = nn.CosineSimilarity()
            svo_sim = criterion(sv_features.view(-1, self.hidden_size), o_features.view(-1, self.hidden_size))
            svo_inds = torch.nonzero(svo_sim)
            svo_sim = svo_sim[svo_inds]
            svo_loss = 1. - svo_sim
            svo_loss = torch.mean(svo_loss)
            # svo_graph_loss = loss_fct(svo_logits, labels)
            total_loss = cls_loss + self.svo_weight*svo_loss
            # print(svo_loss)
            # loss_binary = nn.BCEWithLogitsLoss()
            # loss_contrastive = loss_fct(contrastive_scores, labels)
            # loss = loss_fct(reshaped_logits, labels) + loss_binary(reshaped_logits, binary_labels)
            # loss = loss_binary(reshaped_logits, binary_labels)

        if not return_dict:
            output = (reshaped_logits,) + outputs[2:]
            return ((total_loss,) + output) if total_loss is not None else output

        return MultipleChoiceModelOutput(
            loss=total_loss,
            logits=reshaped_logits,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )


@add_start_docstrings(
    """
    Roberta 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.
    """,
    ROBERTA_START_DOCSTRING,
)
class RobertaForTokenClassification(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_unexpected = [r"pooler"]
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels

        self.roberta = RobertaModel(config, add_pooling_layer=False)
        self.dropout = nn.Dropout(config.hidden_dropout_prob)
        self.classifier = nn.Linear(config.hidden_size, config.num_labels)

        self.init_weights()

    @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        #tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=TokenClassifierOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        token_type_ids=None,
        position_ids=None,
        head_mask=None,
        inputs_embeds=None,
        labels=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
    ):
        r"""
        labels (:obj:`torch.LongTensor` of shape :obj:`(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.roberta(
            input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            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()
            # Only keep active parts of the loss
            if attention_mask is not None:
                active_loss = attention_mask.view(-1) == 1
                active_logits = logits.view(-1, self.num_labels)
                active_labels = torch.where(
                    active_loss, labels.view(-1), torch.tensor(loss_fct.ignore_index).type_as(labels)
                )
                loss = loss_fct(active_logits, active_labels)
            else:
                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 RobertaClassificationHead(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 [CLS])
        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(
    """
    Roberta 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`).
    """,
    ROBERTA_START_DOCSTRING,
)
class RobertaForQuestionAnswering(RobertaPreTrainedModel):
    _keys_to_ignore_on_load_unexpected = [r"pooler"]
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels

        self.roberta = RobertaModel(config, add_pooling_layer=False)
        self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)

        self.init_weights()

    @add_start_docstrings_to_model_forward(ROBERTA_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
       # tokenizer_class=_TOKENIZER_FOR_DOC,
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=QuestionAnsweringModelOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        token_type_ids=None,
        position_ids=None,
        head_mask=None,
        inputs_embeds=None,
        start_positions=None,
        end_positions=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
    ):
        r"""
        start_positions (:obj:`torch.LongTensor` of shape :obj:`(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 (:obj:`sequence_length`). Position outside of the
            sequence are not taken into account for computing the loss.
        end_positions (:obj:`torch.LongTensor` of shape :obj:`(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 (:obj:`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.roberta(
            input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            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)
        end_logits = end_logits.squeeze(-1)

        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.clamp_(0, ignored_index)
            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, past_key_values_length=0):
    """
    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`.
    Args:
        x: torch.Tensor x:
    Returns: 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) + past_key_values_length) * mask
    return incremental_indices.long() + padding_idx