image-match/python/py/Lib/site-packages/transformers/models/led/modeling_led.py

# Copyright 2021 Iz Beltagy, Matthew E. Peters, Arman Cohan and The HuggingFace Inc. team. 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 LED model."""

import math
from dataclasses import dataclass

import torch
from torch import nn
from torch.nn import CrossEntropyLoss

from ... import initialization as init
from ...activations import ACT2FN
from ...cache_utils import Cache, DynamicCache, EncoderDecoderCache
from ...generation import GenerationMixin
from ...masking_utils import create_bidirectional_mask, create_causal_mask
from ...modeling_layers import GradientCheckpointingLayer
from ...modeling_outputs import BaseModelOutputWithPastAndCrossAttentions
from ...modeling_utils import PreTrainedModel
from ...utils import ModelOutput, auto_docstring, logging
from .configuration_led import LEDConfig


logger = logging.get_logger(__name__)


def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start_token_id: int):
    """
    Shift input ids one token to the right.
    """
    shifted_input_ids = input_ids.new_zeros(input_ids.shape)
    shifted_input_ids[:, 1:] = input_ids[:, :-1].clone()
    shifted_input_ids[:, 0] = decoder_start_token_id

    if pad_token_id is None:
        raise ValueError("config.pad_token_id has to be defined.")
    # replace possible -100 values in labels by `pad_token_id`
    shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id)

    return shifted_input_ids


def _prepare_4d_attention_mask_inverted(mask: torch.Tensor, dtype: torch.dtype, tgt_len: int | None = None):
    """
    Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.
    """
    bsz, src_len = mask.size()
    tgt_len = tgt_len if tgt_len is not None else src_len

    expanded_mask = mask[:, None, None, :].expand(bsz, 1, tgt_len, src_len).to(dtype)

    inverted_mask = 1.0 - expanded_mask
    expanded_attention_mask = inverted_mask.masked_fill(inverted_mask.bool(), torch.finfo(dtype).min)

    # make sure that global_attn_mask is positive
    expanded_attention_mask = expanded_attention_mask * inverted_mask

    return expanded_attention_mask


class LEDLearnedPositionalEmbedding(nn.Embedding):
    """
    This module learns positional embeddings up to a fixed maximum size.
    """

    def __init__(self, num_embeddings: int, embedding_dim: int):
        super().__init__(num_embeddings, embedding_dim)

    def forward(self, input_ids_shape: torch.Size, past_key_values_length: int = 0):
        """`input_ids_shape` is expected to be [bsz x seqlen]."""
        bsz, seq_len = input_ids_shape[:2]
        positions = torch.arange(
            past_key_values_length, past_key_values_length + seq_len, dtype=torch.long, device=self.weight.device
        )
        return super().forward(positions)


# Copied from transformers.models.longformer.modeling_longformer.LongformerSelfAttention with Longformer->LEDEncoder
class LEDEncoderSelfAttention(nn.Module):
    def __init__(self, config, layer_id):
        super().__init__()
        if config.hidden_size % config.num_attention_heads != 0:
            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_heads = config.num_attention_heads
        self.head_dim = int(config.hidden_size / config.num_attention_heads)
        self.embed_dim = config.hidden_size

        self.query = nn.Linear(config.hidden_size, self.embed_dim)
        self.key = nn.Linear(config.hidden_size, self.embed_dim)
        self.value = nn.Linear(config.hidden_size, self.embed_dim)

        # separate projection layers for tokens with global attention
        self.query_global = nn.Linear(config.hidden_size, self.embed_dim)
        self.key_global = nn.Linear(config.hidden_size, self.embed_dim)
        self.value_global = nn.Linear(config.hidden_size, self.embed_dim)

        self.dropout = config.attention_probs_dropout_prob

        self.layer_id = layer_id
        attention_window = config.attention_window[self.layer_id]
        assert attention_window % 2 == 0, (
            f"`attention_window` for layer {self.layer_id} has to be an even value. Given {attention_window}"
        )
        assert attention_window > 0, (
            f"`attention_window` for layer {self.layer_id} has to be positive. Given {attention_window}"
        )

        self.one_sided_attn_window_size = attention_window // 2

        self.config = config

    def forward(
        self,
        hidden_states,
        attention_mask=None,
        is_index_masked=None,
        is_index_global_attn=None,
        is_global_attn=None,
        output_attentions=False,
    ):
        """
        [`LEDEncoderSelfAttention`] expects *len(hidden_states)* to be multiple of *attention_window*. Padding to
        *attention_window* happens in [`LEDEncoderModel.forward`] to avoid redoing the padding on each layer.

        The *attention_mask* is changed in [`LEDEncoderModel.forward`] from 0, 1, 2 to:

            - -10000: no attention
            - 0: local attention
            - +10000: global attention
        """
        hidden_states = hidden_states.transpose(0, 1)

        # project hidden states
        query_vectors = self.query(hidden_states)
        key_vectors = self.key(hidden_states)
        value_vectors = self.value(hidden_states)

        seq_len, batch_size, embed_dim = hidden_states.size()
        assert embed_dim == self.embed_dim, (
            f"hidden_states should have embed_dim = {self.embed_dim}, but has {embed_dim}"
        )

        # normalize query
        query_vectors /= math.sqrt(self.head_dim)

        query_vectors = query_vectors.view(seq_len, batch_size, self.num_heads, self.head_dim).transpose(0, 1)
        key_vectors = key_vectors.view(seq_len, batch_size, self.num_heads, self.head_dim).transpose(0, 1)

        attn_scores = self._sliding_chunks_query_key_matmul(
            query_vectors, key_vectors, self.one_sided_attn_window_size
        )

        # values to pad for attention probs
        remove_from_windowed_attention_mask = (attention_mask != 0)[:, :, None, None]

        # cast to fp32/fp16 then replace 1's with -inf
        float_mask = remove_from_windowed_attention_mask.type_as(query_vectors).masked_fill(
            remove_from_windowed_attention_mask, torch.finfo(query_vectors.dtype).min
        )
        # diagonal mask with zeros everywhere and -inf inplace of padding
        diagonal_mask = self._sliding_chunks_query_key_matmul(
            float_mask.new_ones(size=float_mask.size()), float_mask, self.one_sided_attn_window_size
        )

        # pad local attention probs
        attn_scores += diagonal_mask

        assert list(attn_scores.size()) == [
            batch_size,
            seq_len,
            self.num_heads,
            self.one_sided_attn_window_size * 2 + 1,
        ], (
            f"local_attn_probs should be of size ({batch_size}, {seq_len}, {self.num_heads},"
            f" {self.one_sided_attn_window_size * 2 + 1}), but is of size {attn_scores.size()}"
        )

        # compute local attention probs from global attention keys and contact over window dim
        if is_global_attn:
            # compute global attn indices required through out forward fn
            (
                max_num_global_attn_indices,
                is_index_global_attn_nonzero,
                is_local_index_global_attn_nonzero,
                is_local_index_no_global_attn_nonzero,
            ) = self._get_global_attn_indices(is_index_global_attn)
            # calculate global attn probs from global key

            global_key_attn_scores = self._concat_with_global_key_attn_probs(
                query_vectors=query_vectors,
                key_vectors=key_vectors,
                max_num_global_attn_indices=max_num_global_attn_indices,
                is_index_global_attn_nonzero=is_index_global_attn_nonzero,
                is_local_index_global_attn_nonzero=is_local_index_global_attn_nonzero,
                is_local_index_no_global_attn_nonzero=is_local_index_no_global_attn_nonzero,
            )
            # concat to local_attn_probs
            # (batch_size, seq_len, num_heads, extra attention count + 2*window+1)
            attn_scores = torch.cat((global_key_attn_scores, attn_scores), dim=-1)

            # free memory
            del global_key_attn_scores

        attn_probs = nn.functional.softmax(
            attn_scores, dim=-1, dtype=torch.float32
        )  # use fp32 for numerical stability

        # softmax sometimes inserts NaN if all positions are masked, replace them with 0
        attn_probs = torch.masked_fill(attn_probs, is_index_masked[:, :, None, None], 0.0)
        attn_probs = attn_probs.type_as(attn_scores)

        # free memory
        del attn_scores

        # apply dropout
        attn_probs = nn.functional.dropout(attn_probs, p=self.dropout, training=self.training)

        value_vectors = value_vectors.view(seq_len, batch_size, self.num_heads, self.head_dim).transpose(0, 1)

        # compute local attention output with global attention value and add
        if is_global_attn:
            # compute sum of global and local attn
            attn_output = self._compute_attn_output_with_global_indices(
                value_vectors=value_vectors,
                attn_probs=attn_probs,
                max_num_global_attn_indices=max_num_global_attn_indices,
                is_index_global_attn_nonzero=is_index_global_attn_nonzero,
                is_local_index_global_attn_nonzero=is_local_index_global_attn_nonzero,
            )
        else:
            # compute local attn only
            attn_output = self._sliding_chunks_matmul_attn_probs_value(
                attn_probs, value_vectors, self.one_sided_attn_window_size
            )

        assert attn_output.size() == (batch_size, seq_len, self.num_heads, self.head_dim), "Unexpected size"
        attn_output = attn_output.transpose(0, 1).reshape(seq_len, batch_size, embed_dim).contiguous()

        # compute value for global attention and overwrite to attention output
        # TODO: remove the redundant computation
        if is_global_attn:
            global_attn_output, global_attn_probs = self._compute_global_attn_output_from_hidden(
                hidden_states=hidden_states,
                max_num_global_attn_indices=max_num_global_attn_indices,
                is_local_index_global_attn_nonzero=is_local_index_global_attn_nonzero,
                is_index_global_attn_nonzero=is_index_global_attn_nonzero,
                is_local_index_no_global_attn_nonzero=is_local_index_no_global_attn_nonzero,
                is_index_masked=is_index_masked,
            )

            # get only non zero global attn output
            nonzero_global_attn_output = global_attn_output[
                is_local_index_global_attn_nonzero[0], :, is_local_index_global_attn_nonzero[1]
            ]

            # overwrite values with global attention
            attn_output[is_index_global_attn_nonzero[::-1]] = nonzero_global_attn_output.view(
                len(is_local_index_global_attn_nonzero[0]), -1
            )
            # The attention weights for tokens with global attention are
            # just filler values, they were never used to compute the output.
            # Fill with 0 now, the correct values are in 'global_attn_probs'.
            attn_probs[is_index_global_attn_nonzero] = 0

        outputs = (attn_output.transpose(0, 1),)

        if output_attentions:
            outputs += (attn_probs,)

        return outputs + (global_attn_probs,) if (is_global_attn and output_attentions) else outputs

    @staticmethod
    def _pad_and_transpose_last_two_dims(hidden_states_padded, padding):
        """pads rows and then flips rows and columns"""
        hidden_states_padded = nn.functional.pad(
            hidden_states_padded, padding
        )  # padding value is not important because it will be overwritten
        hidden_states_padded = hidden_states_padded.view(
            *hidden_states_padded.size()[:-2], hidden_states_padded.size(-1), hidden_states_padded.size(-2)
        )
        return hidden_states_padded

    @staticmethod
    def _pad_and_diagonalize(chunked_hidden_states):
        """
        shift every row 1 step right, converting columns into diagonals.

        Example:

        ```python
        chunked_hidden_states: [
            0.4983,
            2.6918,
            -0.0071,
            1.0492,
            -1.8348,
            0.7672,
            0.2986,
            0.0285,
            -0.7584,
            0.4206,
            -0.0405,
            0.1599,
            2.0514,
            -1.1600,
            0.5372,
            0.2629,
        ]
        window_overlap = num_rows = 4
        ```

                     (pad & diagonalize) => [ 0.4983, 2.6918, -0.0071, 1.0492, 0.0000, 0.0000, 0.0000
                       0.0000, -1.8348, 0.7672, 0.2986, 0.0285, 0.0000, 0.0000 0.0000, 0.0000, -0.7584, 0.4206,
                       -0.0405, 0.1599, 0.0000 0.0000, 0.0000, 0.0000, 2.0514, -1.1600, 0.5372, 0.2629 ]
        """
        total_num_heads, num_chunks, window_overlap, hidden_dim = chunked_hidden_states.size()
        chunked_hidden_states = nn.functional.pad(
            chunked_hidden_states, (0, window_overlap + 1)
        )  # total_num_heads x num_chunks x window_overlap x (hidden_dim+window_overlap+1). Padding value is not important because it'll be overwritten
        chunked_hidden_states = chunked_hidden_states.view(
            total_num_heads, num_chunks, -1
        )  # total_num_heads x num_chunks x window_overlap*window_overlap+window_overlap
        chunked_hidden_states = chunked_hidden_states[
            :, :, :-window_overlap
        ]  # total_num_heads x num_chunks x window_overlap*window_overlap
        chunked_hidden_states = chunked_hidden_states.view(
            total_num_heads, num_chunks, window_overlap, window_overlap + hidden_dim
        )
        chunked_hidden_states = chunked_hidden_states[:, :, :, :-1]
        return chunked_hidden_states

    @staticmethod
    def _chunk(hidden_states, window_overlap, onnx_export: bool = False):
        """convert into overlapping chunks. Chunk size = 2w, overlap size = w"""
        if not onnx_export:
            # non-overlapping chunks of size = 2w
            hidden_states = hidden_states.view(
                hidden_states.size(0),
                torch.div(hidden_states.size(1), (window_overlap * 2), rounding_mode="trunc"),
                window_overlap * 2,
                hidden_states.size(2),
            )
            # use `as_strided` to make the chunks overlap with an overlap size = window_overlap
            chunk_size = list(hidden_states.size())
            chunk_size[1] = chunk_size[1] * 2 - 1

            chunk_stride = list(hidden_states.stride())
            chunk_stride[1] = chunk_stride[1] // 2
            return hidden_states.as_strided(size=chunk_size, stride=chunk_stride)

        # When exporting to ONNX, use this separate logic
        # have to use slow implementation since as_strided, unfold and 2d-tensor indexing aren't supported (yet) in ONNX export

        # TODO replace this with
        # > return hidden_states.unfold(dimension=1, size=window_overlap * 2, step=window_overlap).transpose(2, 3)
        # once `unfold` is supported
        # the case hidden_states.size(1) == window_overlap * 2 can also simply return hidden_states.unsqueeze(1), but that's control flow

        chunk_size = [
            hidden_states.size(0),
            torch.div(hidden_states.size(1), window_overlap, rounding_mode="trunc") - 1,
            window_overlap * 2,
            hidden_states.size(2),
        ]

        overlapping_chunks = torch.empty(chunk_size, device=hidden_states.device)
        for chunk in range(chunk_size[1]):
            overlapping_chunks[:, chunk, :, :] = hidden_states[
                :, chunk * window_overlap : chunk * window_overlap + 2 * window_overlap, :
            ]
        return overlapping_chunks

    @staticmethod
    def _mask_invalid_locations(input_tensor, affected_seq_len) -> torch.Tensor:
        beginning_mask_2d = input_tensor.new_ones(affected_seq_len, affected_seq_len + 1).tril().flip(dims=[0])
        beginning_mask = beginning_mask_2d[None, :, None, :]
        ending_mask = beginning_mask.flip(dims=(1, 3))
        beginning_input = input_tensor[:, :affected_seq_len, :, : affected_seq_len + 1]
        beginning_mask = beginning_mask.expand(beginning_input.size())
        input_tensor[:, :affected_seq_len, :, : affected_seq_len + 1] = torch.full_like(
            beginning_input, -float("inf")
        ).where(beginning_mask.bool(), beginning_input)
        ending_input = input_tensor[:, -affected_seq_len:, :, -(affected_seq_len + 1) :]
        ending_mask = ending_mask.expand(ending_input.size())
        input_tensor[:, -affected_seq_len:, :, -(affected_seq_len + 1) :] = torch.full_like(
            ending_input, -float("inf")
        ).where(ending_mask.bool(), ending_input)

    def _sliding_chunks_query_key_matmul(self, query: torch.Tensor, key: torch.Tensor, window_overlap: int):
        """
        Matrix multiplication of query and key tensors using with a sliding window attention pattern. This
        implementation splits the input into overlapping chunks of size 2w (e.g. 512 for pretrained LEDEncoder) with an
        overlap of size window_overlap
        """
        batch_size, seq_len, num_heads, head_dim = query.size()
        assert seq_len % (window_overlap * 2) == 0, (
            f"Sequence length should be multiple of {window_overlap * 2}. Given {seq_len}"
        )
        assert query.size() == key.size()

        chunks_count = torch.div(seq_len, window_overlap, rounding_mode="trunc") - 1

        # group batch_size and num_heads dimensions into one, then chunk seq_len into chunks of size window_overlap * 2
        query = query.transpose(1, 2).reshape(batch_size * num_heads, seq_len, head_dim)
        key = key.transpose(1, 2).reshape(batch_size * num_heads, seq_len, head_dim)

        query = self._chunk(query, window_overlap, getattr(self.config, "onnx_export", False))
        key = self._chunk(key, window_overlap, getattr(self.config, "onnx_export", False))

        # matrix multiplication
        # bcxd: batch_size * num_heads x chunks x 2window_overlap x head_dim
        # bcyd: batch_size * num_heads x chunks x 2window_overlap x head_dim
        # bcxy: batch_size * num_heads x chunks x 2window_overlap x 2window_overlap
        diagonal_chunked_attention_scores = torch.einsum("bcxd,bcyd->bcxy", (query, key))  # multiply

        # convert diagonals into columns
        diagonal_chunked_attention_scores = self._pad_and_transpose_last_two_dims(
            diagonal_chunked_attention_scores, padding=(0, 0, 0, 1)
        )

        # allocate space for the overall attention matrix where the chunks are combined. The last dimension
        # has (window_overlap * 2 + 1) columns. The first (window_overlap) columns are the window_overlap lower triangles (attention from a word to
        # window_overlap previous words). The following column is attention score from each word to itself, then
        # followed by window_overlap columns for the upper triangle.

        diagonal_attention_scores = diagonal_chunked_attention_scores.new_zeros(
            (batch_size * num_heads, chunks_count + 1, window_overlap, window_overlap * 2 + 1)
        )

        # copy parts from diagonal_chunked_attention_scores into the combined matrix of attentions
        # - copying the main diagonal and the upper triangle
        diagonal_attention_scores[:, :-1, :, window_overlap:] = diagonal_chunked_attention_scores[
            :, :, :window_overlap, : window_overlap + 1
        ]
        diagonal_attention_scores[:, -1, :, window_overlap:] = diagonal_chunked_attention_scores[
            :, -1, window_overlap:, : window_overlap + 1
        ]
        # - copying the lower triangle
        diagonal_attention_scores[:, 1:, :, :window_overlap] = diagonal_chunked_attention_scores[
            :, :, -(window_overlap + 1) : -1, window_overlap + 1 :
        ]

        diagonal_attention_scores[:, 0, 1:window_overlap, 1:window_overlap] = diagonal_chunked_attention_scores[
            :, 0, : window_overlap - 1, 1 - window_overlap :
        ]

        # separate batch_size and num_heads dimensions again
        diagonal_attention_scores = diagonal_attention_scores.view(
            batch_size, num_heads, seq_len, 2 * window_overlap + 1
        ).transpose(2, 1)

        self._mask_invalid_locations(diagonal_attention_scores, window_overlap)
        return diagonal_attention_scores

    def _sliding_chunks_matmul_attn_probs_value(
        self, attn_probs: torch.Tensor, value: torch.Tensor, window_overlap: int
    ):
        """
        Same as _sliding_chunks_query_key_matmul but for attn_probs and value tensors. Returned tensor will be of the
        same shape as `attn_probs`
        """
        batch_size, seq_len, num_heads, head_dim = value.size()

        assert seq_len % (window_overlap * 2) == 0
        assert attn_probs.size()[:3] == value.size()[:3]
        assert attn_probs.size(3) == 2 * window_overlap + 1
        chunks_count = torch.div(seq_len, window_overlap, rounding_mode="trunc") - 1
        # group batch_size and num_heads dimensions into one, then chunk seq_len into chunks of size 2 window overlap

        chunked_attn_probs = attn_probs.transpose(1, 2).reshape(
            batch_size * num_heads,
            torch.div(seq_len, window_overlap, rounding_mode="trunc"),
            window_overlap,
            2 * window_overlap + 1,
        )

        # group batch_size and num_heads dimensions into one
        value = value.transpose(1, 2).reshape(batch_size * num_heads, seq_len, head_dim)

        # pad seq_len with w at the beginning of the sequence and another window overlap at the end
        padded_value = nn.functional.pad(value, (0, 0, window_overlap, window_overlap), value=-1)

        # chunk padded_value into chunks of size 3 window overlap and an overlap of size window overlap
        chunked_value_size = (batch_size * num_heads, chunks_count + 1, 3 * window_overlap, head_dim)
        chunked_value_stride = padded_value.stride()
        chunked_value_stride = (
            chunked_value_stride[0],
            window_overlap * chunked_value_stride[1],
            chunked_value_stride[1],
            chunked_value_stride[2],
        )
        chunked_value = padded_value.as_strided(size=chunked_value_size, stride=chunked_value_stride)

        chunked_attn_probs = self._pad_and_diagonalize(chunked_attn_probs)

        context = torch.einsum("bcwd,bcdh->bcwh", (chunked_attn_probs, chunked_value))
        return context.view(batch_size, num_heads, seq_len, head_dim).transpose(1, 2)

    @staticmethod
    def _get_global_attn_indices(is_index_global_attn):
        """compute global attn indices required throughout forward pass"""
        # helper variable
        num_global_attn_indices = is_index_global_attn.long().sum(dim=1)

        # max number of global attn indices in batch
        max_num_global_attn_indices = num_global_attn_indices.max()

        # indices of global attn
        is_index_global_attn_nonzero = is_index_global_attn.nonzero(as_tuple=True)

        # helper variable
        is_local_index_global_attn = torch.arange(
            max_num_global_attn_indices, device=is_index_global_attn.device
        ) < num_global_attn_indices.unsqueeze(dim=-1)

        # location of the non-padding values within global attention indices
        is_local_index_global_attn_nonzero = is_local_index_global_attn.nonzero(as_tuple=True)

        # location of the padding values within global attention indices
        is_local_index_no_global_attn_nonzero = (is_local_index_global_attn == 0).nonzero(as_tuple=True)
        return (
            max_num_global_attn_indices,
            is_index_global_attn_nonzero,
            is_local_index_global_attn_nonzero,
            is_local_index_no_global_attn_nonzero,
        )

    def _concat_with_global_key_attn_probs(
        self,
        key_vectors,
        query_vectors,
        max_num_global_attn_indices,
        is_index_global_attn_nonzero,
        is_local_index_global_attn_nonzero,
        is_local_index_no_global_attn_nonzero,
    ):
        batch_size = key_vectors.shape[0]

        # create only global key vectors
        key_vectors_only_global = key_vectors.new_zeros(
            batch_size, max_num_global_attn_indices, self.num_heads, self.head_dim
        )

        key_vectors_only_global[is_local_index_global_attn_nonzero] = key_vectors[is_index_global_attn_nonzero]

        # (batch_size, seq_len, num_heads, max_num_global_attn_indices)
        attn_probs_from_global_key = torch.einsum("blhd,bshd->blhs", (query_vectors, key_vectors_only_global))

        # need to transpose since ONNX export only supports consecutive indexing: https://pytorch.org/docs/stable/onnx.html#writes-sets
        attn_probs_from_global_key = attn_probs_from_global_key.transpose(1, 3)
        attn_probs_from_global_key[
            is_local_index_no_global_attn_nonzero[0], is_local_index_no_global_attn_nonzero[1], :, :
        ] = torch.finfo(attn_probs_from_global_key.dtype).min
        attn_probs_from_global_key = attn_probs_from_global_key.transpose(1, 3)

        return attn_probs_from_global_key

    def _compute_attn_output_with_global_indices(
        self,
        value_vectors,
        attn_probs,
        max_num_global_attn_indices,
        is_index_global_attn_nonzero,
        is_local_index_global_attn_nonzero,
    ):
        batch_size = attn_probs.shape[0]

        # cut local attn probs to global only
        attn_probs_only_global = attn_probs.narrow(-1, 0, max_num_global_attn_indices)
        # get value vectors for global only
        value_vectors_only_global = value_vectors.new_zeros(
            batch_size, max_num_global_attn_indices, self.num_heads, self.head_dim
        )
        value_vectors_only_global[is_local_index_global_attn_nonzero] = value_vectors[is_index_global_attn_nonzero]

        # use `matmul` because `einsum` crashes sometimes with fp16
        # attn = torch.einsum('blhs,bshd->blhd', (selected_attn_probs, selected_v))
        # compute attn output only global
        attn_output_only_global = torch.matmul(
            attn_probs_only_global.transpose(1, 2).clone(), value_vectors_only_global.transpose(1, 2).clone()
        ).transpose(1, 2)

        # reshape attn probs
        attn_probs_without_global = attn_probs.narrow(
            -1, max_num_global_attn_indices, attn_probs.size(-1) - max_num_global_attn_indices
        ).contiguous()

        # compute attn output with global
        attn_output_without_global = self._sliding_chunks_matmul_attn_probs_value(
            attn_probs_without_global, value_vectors, self.one_sided_attn_window_size
        )
        return attn_output_only_global + attn_output_without_global

    def _compute_global_attn_output_from_hidden(
        self,
        hidden_states,
        max_num_global_attn_indices,
        is_local_index_global_attn_nonzero,
        is_index_global_attn_nonzero,
        is_local_index_no_global_attn_nonzero,
        is_index_masked,
    ):
        seq_len, batch_size = hidden_states.shape[:2]

        # prepare global hidden states
        global_attn_hidden_states = hidden_states.new_zeros(max_num_global_attn_indices, batch_size, self.embed_dim)
        global_attn_hidden_states[is_local_index_global_attn_nonzero[::-1]] = hidden_states[
            is_index_global_attn_nonzero[::-1]
        ]

        # global key, query, value
        global_query_vectors_only_global = self.query_global(global_attn_hidden_states)
        global_key_vectors = self.key_global(hidden_states)
        global_value_vectors = self.value_global(hidden_states)

        # normalize
        global_query_vectors_only_global /= math.sqrt(self.head_dim)

        # reshape
        global_query_vectors_only_global = (
            global_query_vectors_only_global.contiguous()
            .view(max_num_global_attn_indices, batch_size * self.num_heads, self.head_dim)
            .transpose(0, 1)
        )  # (batch_size * self.num_heads, max_num_global_attn_indices, head_dim)
        global_key_vectors = (
            global_key_vectors.contiguous().view(-1, batch_size * self.num_heads, self.head_dim).transpose(0, 1)
        )  # batch_size * self.num_heads, seq_len, head_dim)
        global_value_vectors = (
            global_value_vectors.contiguous().view(-1, batch_size * self.num_heads, self.head_dim).transpose(0, 1)
        )  # batch_size * self.num_heads, seq_len, head_dim)

        # compute attn scores
        global_attn_scores = torch.bmm(global_query_vectors_only_global, global_key_vectors.transpose(1, 2))

        assert list(global_attn_scores.size()) == [
            batch_size * self.num_heads,
            max_num_global_attn_indices,
            seq_len,
        ], (
            "global_attn_scores have the wrong size. Size should be"
            f" {(batch_size * self.num_heads, max_num_global_attn_indices, seq_len)}, but is"
            f" {global_attn_scores.size()}."
        )

        global_attn_scores = global_attn_scores.view(batch_size, self.num_heads, max_num_global_attn_indices, seq_len)

        # need to transpose since ONNX export only supports consecutive indexing: https://pytorch.org/docs/stable/onnx.html#writes-sets
        global_attn_scores = global_attn_scores.transpose(1, 2)
        global_attn_scores[
            is_local_index_no_global_attn_nonzero[0], is_local_index_no_global_attn_nonzero[1], :, :
        ] = torch.finfo(global_attn_scores.dtype).min
        global_attn_scores = global_attn_scores.transpose(1, 2)

        global_attn_scores = global_attn_scores.masked_fill(
            is_index_masked[:, None, None, :],
            torch.finfo(global_attn_scores.dtype).min,
        )

        global_attn_scores = global_attn_scores.view(batch_size * self.num_heads, max_num_global_attn_indices, seq_len)

        # compute global attn probs
        global_attn_probs_float = nn.functional.softmax(
            global_attn_scores, dim=-1, dtype=torch.float32
        )  # use fp32 for numerical stability

        global_attn_probs = nn.functional.dropout(
            global_attn_probs_float.type_as(global_attn_scores), p=self.dropout, training=self.training
        )

        # global attn output
        global_attn_output = torch.bmm(global_attn_probs, global_value_vectors)

        assert list(global_attn_output.size()) == [
            batch_size * self.num_heads,
            max_num_global_attn_indices,
            self.head_dim,
        ], (
            "global_attn_output tensor has the wrong size. Size should be"
            f" {(batch_size * self.num_heads, max_num_global_attn_indices, self.head_dim)}, but is"
            f" {global_attn_output.size()}."
        )

        global_attn_probs = global_attn_probs.view(batch_size, self.num_heads, max_num_global_attn_indices, seq_len)
        global_attn_output = global_attn_output.view(
            batch_size, self.num_heads, max_num_global_attn_indices, self.head_dim
        )
        return global_attn_output, global_attn_probs


class LEDEncoderAttention(nn.Module):
    def __init__(self, config, layer_id):
        super().__init__()
        self.longformer_self_attn = LEDEncoderSelfAttention(config, layer_id=layer_id)
        self.output = nn.Linear(config.d_model, config.d_model)

    def forward(
        self,
        hidden_states: torch.Tensor,
        attention_mask: torch.Tensor | None = None,
        is_index_masked: torch.Tensor | None = None,
        is_index_global_attn: torch.Tensor | None = None,
        is_global_attn: bool | None = None,
        output_attentions: bool = False,
    ) -> tuple[torch.Tensor, torch.Tensor | None, tuple[torch.Tensor] | None]:
        """Input shape: Batch x Time x Channel"""

        self_outputs = self.longformer_self_attn(
            hidden_states=hidden_states,
            attention_mask=attention_mask,
            is_index_masked=is_index_masked,
            is_index_global_attn=is_index_global_attn,
            is_global_attn=is_global_attn,
            output_attentions=output_attentions,
        )

        attn_output = self.output(self_outputs[0])
        outputs = (attn_output,) + self_outputs[1:]

        return outputs


class LEDDecoderAttention(nn.Module):
    """Multi-headed attention from 'Attention Is All You Need' paper"""

    def __init__(
        self,
        embed_dim: int,
        num_heads: int,
        dropout: float | None = 0.0,
        is_decoder: bool | None = False,
        bias: bool | None = True,
        layer_idx: bool | None = None,
    ):
        super().__init__()
        self.embed_dim = embed_dim
        self.num_heads = num_heads
        self.dropout = dropout
        self.head_dim = embed_dim // num_heads
        if self.head_dim * num_heads != self.embed_dim:
            raise ValueError(
                f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim} and `num_heads`:"
                f" {num_heads})."
            )
        self.scaling = self.head_dim**-0.5
        self.is_decoder = is_decoder
        self.layer_idx = layer_idx

        self.k_proj = nn.Linear(embed_dim, embed_dim, bias=bias)
        self.v_proj = nn.Linear(embed_dim, embed_dim, bias=bias)
        self.q_proj = nn.Linear(embed_dim, embed_dim, bias=bias)
        self.out_proj = nn.Linear(embed_dim, embed_dim, bias=bias)

    def forward(
        self,
        hidden_states: torch.Tensor,
        key_value_states: torch.Tensor | None = None,
        past_key_values: Cache | None = None,
        attention_mask: torch.Tensor | None = None,
        output_attentions: bool = False,
    ) -> tuple[torch.Tensor, torch.Tensor | None, Cache | None]:
        """Input shape: Batch x Time x Channel"""

        # if key_value_states are provided this layer is used as a cross-attention layer
        # for the decoder
        is_cross_attention = key_value_states is not None
        bsz, tgt_len, embed_dim = hidden_states.size()

        # get query proj
        query_states = self.q_proj(hidden_states) * self.scaling

        is_updated = False
        if past_key_values is not None:
            if isinstance(past_key_values, EncoderDecoderCache):
                is_updated = past_key_values.is_updated.get(self.layer_idx)
                if is_cross_attention:
                    # after the first generated id, we can subsequently re-use all key/value_states from cache
                    curr_past_key_values = past_key_values.cross_attention_cache
                else:
                    curr_past_key_values = past_key_values.self_attention_cache
            else:
                curr_past_key_values = past_key_values

        current_states = key_value_states if is_cross_attention else hidden_states
        if is_cross_attention and past_key_values is not None and is_updated:
            # reuse k,v, cross_attentions
            key_states = curr_past_key_values.layers[self.layer_idx].keys
            value_states = curr_past_key_values.layers[self.layer_idx].values
        else:
            key_states = self.k_proj(current_states)
            value_states = self.v_proj(current_states)
            key_states = key_states.view(bsz, -1, self.num_heads, self.head_dim).transpose(1, 2)
            value_states = value_states.view(bsz, -1, self.num_heads, self.head_dim).transpose(1, 2)

            if past_key_values is not None:
                # save all key/value_states to cache to be re-used for fast auto-regressive generation
                key_states, value_states = curr_past_key_values.update(key_states, value_states, self.layer_idx)
                # set flag that curr layer for cross-attn is already updated so we can re-use in subsequent calls
                if is_cross_attention and isinstance(past_key_values, EncoderDecoderCache):
                    past_key_values.is_updated[self.layer_idx] = True

        proj_shape = (bsz * self.num_heads, -1, self.head_dim)
        query_states = query_states.view(bsz, tgt_len, self.num_heads, self.head_dim).transpose(1, 2)
        query_states = query_states.reshape(*proj_shape)
        key_states = key_states.reshape(*proj_shape)
        value_states = value_states.reshape(*proj_shape)

        src_len = key_states.size(1)
        attn_weights = torch.bmm(query_states, key_states.transpose(1, 2))

        if attn_weights.size() != (bsz * self.num_heads, tgt_len, src_len):
            raise ValueError(
                f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is"
                f" {attn_weights.size()}"
            )

        if attention_mask is not None:
            if attention_mask.size() != (bsz, 1, tgt_len, src_len):
                raise ValueError(
                    f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is {attention_mask.size()}"
                )
            attn_weights = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attention_mask
            attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len)

        attn_weights = nn.functional.softmax(attn_weights, dim=-1)

        if output_attentions:
            # this operation is a bit awkward, but it's required to
            # make sure that attn_weights keeps its gradient.
            # In order to do so, attn_weights have to be reshaped
            # twice and have to be reused in the following
            attn_weights_reshaped = attn_weights.view(bsz, self.num_heads, tgt_len, src_len)
            attn_weights = attn_weights_reshaped.view(bsz * self.num_heads, tgt_len, src_len)
        else:
            attn_weights_reshaped = None

        attn_probs = nn.functional.dropout(attn_weights, p=self.dropout, training=self.training)

        attn_output = torch.bmm(attn_probs, value_states)

        if attn_output.size() != (bsz * self.num_heads, tgt_len, self.head_dim):
            raise ValueError(
                f"`attn_output` should be of size {(bsz, self.num_heads, tgt_len, self.head_dim)}, but is"
                f" {attn_output.size()}"
            )

        attn_output = (
            attn_output.view(bsz, self.num_heads, tgt_len, self.head_dim)
            .transpose(1, 2)
            .reshape(bsz, tgt_len, embed_dim)
        )

        attn_output = self.out_proj(attn_output)

        return attn_output, attn_weights_reshaped, past_key_values


class LEDEncoderLayer(GradientCheckpointingLayer):
    def __init__(self, config: LEDConfig, layer_id: int):
        super().__init__()
        self.embed_dim = config.d_model
        self.self_attn = LEDEncoderAttention(config, layer_id)
        self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim)
        self.dropout = config.dropout
        self.activation_fn = ACT2FN[config.activation_function]
        self.activation_dropout = config.activation_dropout
        self.fc1 = nn.Linear(self.embed_dim, config.encoder_ffn_dim)
        self.fc2 = nn.Linear(config.encoder_ffn_dim, self.embed_dim)
        self.final_layer_norm = nn.LayerNorm(self.embed_dim)

    def forward(
        self,
        hidden_states: torch.Tensor,
        attention_mask: torch.Tensor,
        is_index_masked=None,
        is_index_global_attn=None,
        is_global_attn=None,
        output_attentions=False,
    ):
        """
        Args:
            hidden_states (`torch.FloatTensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
            attention_mask (`torch.FloatTensor`): attention mask of size
                *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
        """
        residual = hidden_states
        attn_outputs = self.self_attn(
            hidden_states=hidden_states,
            attention_mask=attention_mask,
            is_index_masked=is_index_masked,
            is_index_global_attn=is_index_global_attn,
            is_global_attn=is_global_attn,
            output_attentions=output_attentions,
        )
        hidden_states = attn_outputs[0]
        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
        hidden_states = residual + hidden_states
        hidden_states = self.self_attn_layer_norm(hidden_states)

        residual = hidden_states
        hidden_states = self.activation_fn(self.fc1(hidden_states))
        hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training=self.training)
        hidden_states = self.fc2(hidden_states)
        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
        hidden_states = residual + hidden_states
        hidden_states = self.final_layer_norm(hidden_states)

        if hidden_states.dtype == torch.float16 and not torch.isfinite(hidden_states).all():
            clamp_value = torch.finfo(hidden_states.dtype).max - 1000
            hidden_states = torch.clamp(hidden_states, min=-clamp_value, max=clamp_value)
        return (hidden_states,) + attn_outputs[1:]


class LEDDecoderLayer(GradientCheckpointingLayer):
    def __init__(self, config: LEDConfig, layer_idx=None):
        super().__init__()
        self.embed_dim = config.d_model

        self.self_attn = LEDDecoderAttention(
            embed_dim=self.embed_dim,
            num_heads=config.decoder_attention_heads,
            dropout=config.attention_dropout,
            is_decoder=True,
            layer_idx=layer_idx,
        )
        self.dropout = config.dropout
        self.activation_fn = ACT2FN[config.activation_function]
        self.activation_dropout = config.activation_dropout

        self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim)
        self.encoder_attn = LEDDecoderAttention(
            self.embed_dim,
            config.decoder_attention_heads,
            dropout=config.attention_dropout,
            is_decoder=True,
            layer_idx=layer_idx,
        )
        self.encoder_attn_layer_norm = nn.LayerNorm(self.embed_dim)
        self.fc1 = nn.Linear(self.embed_dim, config.decoder_ffn_dim)
        self.fc2 = nn.Linear(config.decoder_ffn_dim, self.embed_dim)
        self.final_layer_norm = nn.LayerNorm(self.embed_dim)

    def forward(
        self,
        hidden_states: torch.Tensor,
        attention_mask: torch.Tensor | None = None,
        encoder_hidden_states: torch.Tensor | None = None,
        encoder_attention_mask: torch.Tensor | None = None,
        past_key_values: Cache | None = None,
        output_attentions: bool | None = False,
        use_cache: bool | None = True,
        **kwargs,
    ):
        """
        Args:
            hidden_states (`torch.FloatTensor`): input to the layer of shape *(batch, seq_len, embed_dim)*
            attention_mask (`torch.FloatTensor`): attention mask of size
                *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
            encoder_hidden_states (`torch.FloatTensor`):
                cross attention input to the layer of shape *(batch, seq_len, embed_dim)*
            encoder_attention_mask (`torch.FloatTensor`): encoder attention mask of size
                *(batch, 1, tgt_len, src_len)* where padding elements are indicated by very large negative values.
            past_key_values (`Cache`): cached past key and value projection states
            output_attentions (`bool`): Whether the base model outputs attentions.
                This requires the attentions tensor to be reshaped in this function.
        """
        residual = hidden_states

        # Self-Attention
        hidden_states, self_attn_weights, present_key_value = self.self_attn(
            hidden_states=hidden_states,
            past_key_values=past_key_values,
            attention_mask=attention_mask,
            output_attentions=output_attentions,
        )
        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
        hidden_states = residual + hidden_states
        hidden_states = self.self_attn_layer_norm(hidden_states)

        # Cross-Attention Block
        cross_attn_present_key_value = None
        cross_attn_weights = None
        if encoder_hidden_states is not None:
            residual = hidden_states

            hidden_states, cross_attn_weights, cross_attn_present_key_value = self.encoder_attn(
                hidden_states=hidden_states,
                key_value_states=encoder_hidden_states,
                attention_mask=encoder_attention_mask,
                past_key_values=past_key_values,
                output_attentions=output_attentions,
            )
            hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
            hidden_states = residual + hidden_states
            hidden_states = self.encoder_attn_layer_norm(hidden_states)

        # Fully Connected
        residual = hidden_states
        hidden_states = self.activation_fn(self.fc1(hidden_states))
        hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training=self.training)
        hidden_states = self.fc2(hidden_states)
        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
        hidden_states = residual + hidden_states
        hidden_states = self.final_layer_norm(hidden_states)

        outputs = (hidden_states,)

        if output_attentions:
            outputs += (self_attn_weights, cross_attn_weights)

        if use_cache:
            outputs += (past_key_values,)

        return outputs


class LEDClassificationHead(nn.Module):
    """Head for sentence-level classification tasks."""

    def __init__(
        self,
        input_dim: int,
        inner_dim: int,
        num_classes: int,
        pooler_dropout: float,
    ):
        super().__init__()
        self.dense = nn.Linear(input_dim, inner_dim)
        self.dropout = nn.Dropout(p=pooler_dropout)
        self.out_proj = nn.Linear(inner_dim, num_classes)

    def forward(self, hidden_states: torch.Tensor):
        hidden_states = self.dropout(hidden_states)
        hidden_states = self.dense(hidden_states)
        hidden_states = torch.tanh(hidden_states)
        hidden_states = self.dropout(hidden_states)
        hidden_states = self.out_proj(hidden_states)
        return hidden_states


@auto_docstring
class LEDPreTrainedModel(PreTrainedModel):
    config: LEDConfig
    base_model_prefix = "led"
    supports_gradient_checkpointing = True

    @property
    def dummy_inputs(self):
        pad_token = self.config.pad_token_id
        input_ids = torch.tensor([[0, 6, 10, 4, 2], [0, 8, 12, 2, pad_token]], device=self.device)
        dummy_inputs = {
            "attention_mask": input_ids.ne(pad_token),
            "input_ids": input_ids,
        }
        return dummy_inputs

    def _init_weights(self, module):
        super()._init_weights(module)
        if isinstance(module, LEDForConditionalGeneration):
            init.zeros_(module.final_logits_bias)


@dataclass
@auto_docstring(
    custom_intro="""
    Base class for LEDEncoder's outputs, with potential hidden states, local and global attentions.
    """
)
# Copied from transformers.models.longformer.modeling_longformer.LongformerBaseModelOutput with Longformer->LEDEncoder
class LEDEncoderBaseModelOutput(ModelOutput):
    r"""
    attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x +
        attention_window + 1)`, where `x` is the number of tokens with global attention mask.

        Local attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token in the sequence to every token with
        global attention (first `x` values) and to every token in the attention window (remaining `attention_window
        + 1` values). Note that the first `x` values refer to tokens with fixed positions in the text, but the
        remaining `attention_window + 1` values refer to tokens with relative positions: the attention weight of a
        token to itself is located at index `x + attention_window / 2` and the `attention_window / 2` preceding
        (succeeding) values are the attention weights to the `attention_window / 2` preceding (succeeding) tokens.
        If the attention window contains a token with global attention, the attention weight at the corresponding
        index is set to 0; the value should be accessed from the first `x` attention weights. If a token has global
        attention, the attention weights to all other tokens in `attentions` is set to 0, the values should be
        accessed from `global_attentions`.
    global_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x)`,
        where `x` is the number of tokens with global attention mask.

        Global attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token with global attention to every token
        in the sequence.
    """

    last_hidden_state: torch.FloatTensor
    hidden_states: tuple[torch.FloatTensor, ...] | None = None
    attentions: tuple[torch.FloatTensor, ...] | None = None
    global_attentions: tuple[torch.FloatTensor, ...] | None = None


@dataclass
@auto_docstring(
    custom_intro="""
    Base class for model encoder's outputs that also contains : pre-computed hidden states that can speed up sequential
    decoding.
    """
)
class LEDSeq2SeqModelOutput(ModelOutput):
    r"""
    last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
        Sequence of hidden-states at the output of the last layer of the decoder of the model.

        If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
        hidden_size)` is output.
    past_key_values (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
        It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

        Contains pre-computed hidden-states (key and values in the attention blocks) of the decoder that can be
        used (see `past_key_values` input) to speed up sequential decoding.
    encoder_global_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x)`,
        where `x` is the number of tokens with global attention mask.

        Global attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token with global attention to every token
        in the sequence.
    """

    last_hidden_state: torch.FloatTensor | None = None
    past_key_values: Cache | None = None
    decoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    decoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    cross_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_last_hidden_state: torch.FloatTensor | None = None
    encoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    encoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_global_attentions: tuple[torch.FloatTensor, ...] | None = None


@dataclass
@auto_docstring(
    custom_intro="""
    Base class for sequence-to-sequence language models outputs.
    """
)
class LEDSeq2SeqLMOutput(ModelOutput):
    r"""
    loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
        Language modeling loss.
    logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
        Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
    past_key_values (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
        It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

        Contains pre-computed hidden-states (key and values in the attention blocks) of the decoder that can be
        used (see `past_key_values` input) to speed up sequential decoding.
    encoder_global_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x)`,
        where `x` is the number of tokens with global attention mask.

        Global attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token with global attention to every token
        in the sequence.
    """

    loss: torch.FloatTensor | None = None
    logits: torch.FloatTensor | None = None
    past_key_values: Cache | None = None
    decoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    decoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    cross_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_last_hidden_state: torch.FloatTensor | None = None
    encoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    encoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_global_attentions: tuple[torch.FloatTensor, ...] | None = None


@dataclass
@auto_docstring(
    custom_intro="""
    Base class for outputs of sequence-to-sequence sentence classification models.
    """
)
class LEDSeq2SeqSequenceClassifierOutput(ModelOutput):
    r"""
    loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `label` is provided):
        Classification (or regression if config.num_labels==1) loss.
    logits (`torch.FloatTensor` of shape `(batch_size, config.num_labels)`):
        Classification (or regression if config.num_labels==1) scores (before SoftMax).
    past_key_values (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
        It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

        Contains pre-computed hidden-states (key and values in the attention blocks) of the decoder that can be
        used (see `past_key_values` input) to speed up sequential decoding.
    encoder_global_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x)`,
        where `x` is the number of tokens with global attention mask.

        Global attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token with global attention to every token
        in the sequence.
    """

    loss: torch.FloatTensor | None = None
    logits: torch.FloatTensor | None = None
    past_key_values: Cache | None = None
    decoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    decoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    cross_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_last_hidden_state: torch.FloatTensor | None = None
    encoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    encoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_global_attentions: tuple[torch.FloatTensor, ...] | None = None


@dataclass
@auto_docstring(
    custom_intro="""
    Base class for outputs of sequence-to-sequence question answering models.
    """
)
class LEDSeq2SeqQuestionAnsweringModelOutput(ModelOutput):
    r"""
    loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
        Total span extraction loss is the sum of a Cross-Entropy for the start and end positions.
    past_key_values (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
        It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

        Contains pre-computed hidden-states (key and values in the attention blocks) of the decoder that can be
        used (see `past_key_values` input) to speed up sequential decoding.
    encoder_global_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
        Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, x)`,
        where `x` is the number of tokens with global attention mask.

        Global attentions weights after the attention softmax, used to compute the weighted average in the
        self-attention heads. Those are the attention weights from every token with global attention to every token
        in the sequence.
    """

    loss: torch.FloatTensor | None = None
    start_logits: torch.FloatTensor | None = None
    end_logits: torch.FloatTensor | None = None
    past_key_values: Cache | None = None
    decoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    decoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    cross_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_last_hidden_state: torch.FloatTensor | None = None
    encoder_hidden_states: tuple[torch.FloatTensor, ...] | None = None
    encoder_attentions: tuple[torch.FloatTensor, ...] | None = None
    encoder_global_attentions: tuple[torch.FloatTensor, ...] | None = None


class LEDEncoder(LEDPreTrainedModel):
    """
    Transformer encoder consisting of *config.encoder_layers* self-attention layers. Each layer is a
    [`LEDEncoderLayer`].

    Args:
        config: LEDConfig
        embed_tokens (nn.Embedding): output embedding
    """

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

        self.dropout = config.dropout
        self.layerdrop = config.encoder_layerdrop

        embed_dim = config.d_model
        self.padding_idx = config.pad_token_id
        self.max_source_positions = config.max_encoder_position_embeddings

        if isinstance(config.attention_window, int):
            if config.attention_window % 2 != 0:
                raise ValueError("`config.attention_window` has to be an even value")
            if config.attention_window <= 0:
                raise ValueError("`config.attention_window` has to be positive")
            config.attention_window = [config.attention_window] * config.num_hidden_layers  # one value per layer
        else:
            if len(config.attention_window) != config.num_hidden_layers:
                raise ValueError(
                    "`len(config.attention_window)` should equal `config.num_hidden_layers`. "
                    f"Expected {config.num_hidden_layers}, given {len(config.attention_window)}"
                )

        self.embed_tokens = nn.Embedding(config.vocab_size, embed_dim, self.padding_idx)

        self.embed_positions = LEDLearnedPositionalEmbedding(
            self.max_source_positions,
            embed_dim,
        )
        self.layers = nn.ModuleList([LEDEncoderLayer(config, i) for i in range(config.encoder_layers)])
        self.layernorm_embedding = nn.LayerNorm(embed_dim)

        self.gradient_checkpointing = False
        # Initialize weights and apply final processing
        self.post_init()

    def _merge_to_attention_mask(self, attention_mask: torch.Tensor, global_attention_mask: torch.Tensor):
        # longformer self-attention expects attention mask to have 0 (no attn), 1 (local attn), 2 (global attn)
        # (global_attention_mask + 1) => 1 for local attention, 2 for global attention
        # => final attention_mask => 0 for no attention, 1 for local attention 2 for global attention
        if attention_mask is not None:
            attention_mask = attention_mask * (global_attention_mask + 1)
        else:
            # simply use `global_attention_mask` as `attention_mask`
            # if no `attention_mask` is given
            attention_mask = global_attention_mask + 1
        return attention_mask

    def _pad_to_window_size(
        self,
        input_ids: torch.Tensor,
        attention_mask: torch.Tensor,
        inputs_embeds: torch.Tensor,
        pad_token_id: int,
    ):
        """A helper function to pad tokens and mask to work with implementation of Longformer self-attention."""
        # padding
        attention_window = (
            self.config.attention_window
            if isinstance(self.config.attention_window, int)
            else max(self.config.attention_window)
        )

        if attention_window % 2 != 0:
            raise ValueError(f"`attention_window` should be an even value. Given {attention_window}")
        input_shape = input_ids.shape if input_ids is not None else inputs_embeds.shape
        batch_size, seq_len = input_shape[:2]

        padding_len = (attention_window - seq_len % attention_window) % attention_window
        if padding_len > 0:
            logger.warning_once(
                f"Input ids are automatically padded from {seq_len} to {seq_len + padding_len} to be a multiple of "
                f"`config.attention_window`: {attention_window}"
            )
            if input_ids is not None:
                input_ids = nn.functional.pad(input_ids, (0, padding_len), value=pad_token_id)
            if inputs_embeds is not None:
                input_ids_padding = inputs_embeds.new_full(
                    (batch_size, padding_len),
                    self.config.pad_token_id,
                    dtype=torch.long,
                )
                inputs_embeds_padding = self.embed_tokens(input_ids_padding)
                inputs_embeds = torch.cat([inputs_embeds, inputs_embeds_padding], dim=-2)

            attention_mask = nn.functional.pad(
                attention_mask, (0, padding_len), value=False
            )  # no attention on the padding tokens

        return padding_len, input_ids, attention_mask, inputs_embeds

    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        global_attention_mask=None,
        inputs_embeds=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
        **kwargs,
    ):
        r"""
        Args:
            input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
                Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you
                provide it.

                Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
                [`PreTrainedTokenizer.__call__`] for details.

                [What are input IDs?](../glossary#input-ids)
            attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
                Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:

                - 1 for tokens that are **not masked**,
                - 0 for tokens that are **masked**.

                [What are attention masks?](../glossary#attention-mask)
            global_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
                Mask to decide the attention given on each token, local attention or global attention for the encoder.
                Tokens with global attention attends to all other tokens, and all other tokens attend to them. This is
                important for task-specific finetuning because it makes the model more flexible at representing the
                task. For example, for classification, the <s> token should be given global attention. For QA, all
                question tokens should also have global attention. Please refer to the [Longformer
                paper](https://huggingface.co/papers/2004.05150) for more details. Mask values selected in `[0, 1]`:

                - 0 for local attention (a sliding window attention),
                - 1 for global attention (tokens that attend to all other tokens, and all other tokens attend to them).
            inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
                Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation.
                This is useful if you want more control over how to convert `input_ids` indices into associated vectors
                than the model's internal embedding lookup matrix.
            output_attentions (`bool`, *optional*):
                Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                returned tensors for more detail.
            output_hidden_states (`bool`, *optional*):
                Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors
                for more detail.
            return_dict (`bool`, *optional*):
                Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
        """
        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.return_dict

        # check input_ids and inputs_embeds
        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 None and inputs_embeds is None:
            raise ValueError("You have to specify either input_ids or inputs_embeds")

        if inputs_embeds is None:
            inputs_embeds = self.embed_tokens(input_ids)

        # create default attention_mask
        if attention_mask is None:
            attention_mask = torch.ones(inputs_embeds.size()[:-1], device=inputs_embeds.device, dtype=torch.long)

        # merge `global_attention_mask` and `attention_mask`
        if global_attention_mask is not None:
            attention_mask = self._merge_to_attention_mask(attention_mask, global_attention_mask)

        # pad input if necessary
        padding_len, input_ids, attention_mask, inputs_embeds = self._pad_to_window_size(
            input_ids=input_ids,
            attention_mask=attention_mask,
            inputs_embeds=inputs_embeds,
            pad_token_id=self.config.pad_token_id,
        )

        # retrieve input_shape
        if input_ids is not None:
            input_shape = input_ids.size()
            input_ids = input_ids.view(-1, input_shape[-1])
        elif inputs_embeds is not None:
            input_shape = inputs_embeds.size()[:-1]

        # convert attention_mask to float
        if attention_mask is not None:
            # [bsz, seq_len] -> [bsz, seq_len]; 1 -> 0.0; 0 -> "-inf"
            attention_mask = _prepare_4d_attention_mask_inverted(attention_mask, inputs_embeds.dtype)[:, 0, 0, :]

        # get masking tensors
        is_index_masked = attention_mask < 0
        is_index_global_attn = attention_mask > 0
        is_global_attn = is_index_global_attn.flatten().any().item()

        embed_pos = self.embed_positions(input_shape)

        hidden_states = inputs_embeds + embed_pos
        hidden_states = self.layernorm_embedding(hidden_states)
        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)

        encoder_states = () if output_hidden_states else None
        all_attentions = () if output_attentions else None
        all_global_attentions = () if (output_attentions and is_global_attn) else None

        for idx, encoder_layer in enumerate(self.layers):
            if output_hidden_states:
                encoder_states = encoder_states + (hidden_states,)
            # add LayerDrop (see https://huggingface.co/papers/1909.11556 for description)
            dropout_probability = torch.rand([])

            if self.training and (dropout_probability < self.layerdrop):  # skip the layer
                layer_outputs = (None, None, None)
            else:
                layer_outputs = encoder_layer(
                    hidden_states,
                    attention_mask=attention_mask,
                    is_index_masked=is_index_masked,
                    is_index_global_attn=is_index_global_attn,
                    is_global_attn=is_global_attn,
                    output_attentions=output_attentions,
                )
                hidden_states = layer_outputs[0]

            if output_attentions:
                # bzs x seq_len x num_attn_heads x (num_global_attn + attention_window_len + 1) => bzs x num_attn_heads x seq_len x (num_global_attn + attention_window_len + 1)
                all_attentions = all_attentions + (layer_outputs[1].transpose(1, 2),)

                if is_global_attn:
                    # bzs x num_attn_heads x num_global_attn x seq_len => bzs x num_attn_heads x seq_len x num_global_attn
                    all_global_attentions = all_global_attentions + (layer_outputs[2].transpose(2, 3),)

        if output_hidden_states:
            encoder_states = encoder_states + (hidden_states,)

        # undo padding
        if padding_len > 0:
            # unpad `hidden_states` because the calling function is expecting a length == input_ids.size(1)
            hidden_states = hidden_states[:, :-padding_len]
            if output_hidden_states:
                encoder_states = tuple(state[:, :-padding_len] for state in encoder_states)

            if output_attentions:
                all_attentions = tuple(state[:, :, :-padding_len, :] for state in all_attentions)

        if not return_dict:
            return tuple(
                v for v in [hidden_states, encoder_states, all_attentions, all_global_attentions] if v is not None
            )
        return LEDEncoderBaseModelOutput(
            last_hidden_state=hidden_states,
            hidden_states=encoder_states,
            attentions=all_attentions,
            global_attentions=all_global_attentions,
        )


class LEDDecoder(LEDPreTrainedModel):
    """
    Transformer decoder consisting of *config.decoder_layers* layers. Each layer is a [`LEDDecoderLayer`]

    Args:
        config: LEDConfig
        embed_tokens (nn.Embedding): output embedding
    """

    def __init__(self, config: LEDConfig):
        super().__init__(config)
        self.dropout = config.dropout
        self.layerdrop = config.decoder_layerdrop
        self.padding_idx = config.pad_token_id
        self.max_target_positions = config.max_decoder_position_embeddings

        self.embed_tokens = nn.Embedding(config.vocab_size, config.d_model, self.padding_idx)

        self.embed_positions = LEDLearnedPositionalEmbedding(
            self.max_target_positions,
            config.d_model,
        )
        self.layers = nn.ModuleList([LEDDecoderLayer(config, layer_idx=i) for i in range(config.decoder_layers)])
        self.layernorm_embedding = nn.LayerNorm(config.d_model)

        self.gradient_checkpointing = False
        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids=None,
        attention_mask=None,
        global_attention_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_values=None,
        inputs_embeds=None,
        use_cache=None,
        output_attentions=None,
        output_hidden_states=None,
        return_dict=None,
        **kwargs,
    ):
        r"""
        Args:
            input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
                Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you
                provide it.

                Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
                [`PreTrainedTokenizer.__call__`] for details.

                [What are input IDs?](../glossary#input-ids)
            attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
                Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:

                - 1 for tokens that are **not masked**,
                - 0 for tokens that are **masked**.

                [What are attention masks?](../glossary#attention-mask)
            global_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
                Mask to decide the attention given on each token, local attention or global attention. Tokens with
                global attention attends to all other tokens, and all other tokens attend to them. This is important
                for task-specific finetuning because it makes the model more flexible at representing the task. For
                example, for classification, the <s> token should be given global attention. For QA, all question
                tokens should also have global attention. Please refer to the [Longformer
                paper](https://huggingface.co/papers/2004.05150) for more details. Mask values selected in `[0, 1]`:

                - 0 for local attention (a sliding window attention),
                - 1 for global attention (tokens that attend to all other tokens, and all other tokens attend to them).
            encoder_hidden_states (`torch.FloatTensor` of shape `(batch_size, encoder_sequence_length, hidden_size)`, *optional*):
                Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention
                of the decoder.
            encoder_attention_mask (`torch.LongTensor` of shape `(batch_size, encoder_sequence_length)`, *optional*):
                Mask to avoid performing cross-attention on padding tokens indices of encoder input_ids. 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)
            past_key_values (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
                It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

                Contains pre-computed hidden-states (key and values in the self-attention blocks and in the
                cross-attention blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.

                If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those
                that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of
                all `decoder_input_ids` of shape `(batch_size, sequence_length)`.
            inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
                Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation.
                This is useful if you want more control over how to convert `input_ids` indices into associated vectors
                than the model's internal embedding lookup matrix.
            output_attentions (`bool`, *optional*):
                Whether or not to return the attentions tensors of all attention layers. See `attentions` under
                returned tensors for more detail.
            output_hidden_states (`bool`, *optional*):
                Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors
                for more detail.
            return_dict (`bool`, *optional*):
                Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
        """
        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
        )
        use_cache = use_cache if use_cache is not None else self.config.use_cache
        return_dict = return_dict if return_dict is not None else self.config.return_dict

        # retrieve input_ids and inputs_embeds
        if input_ids is not None and inputs_embeds is not None:
            raise ValueError("You cannot specify both decoder_input_ids and decoder_inputs_embeds at the same time")
        elif input_ids is not None:
            input_shape = input_ids.size()
            input_ids = input_ids.view(-1, input_shape[-1])
        elif inputs_embeds is not None:
            input_shape = inputs_embeds.size()[:-1]
        else:
            raise ValueError("You have to specify either decoder_input_ids or decoder_inputs_embeds")

        if inputs_embeds is None:
            inputs_embeds = self.embed_tokens(input_ids)

        if self.gradient_checkpointing and self.training:
            if use_cache:
                logger.warning_once(
                    "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
                )
                use_cache = False

        if use_cache and past_key_values is None:
            past_key_values = EncoderDecoderCache(DynamicCache(config=self.config), DynamicCache(config=self.config))

        past_key_values_length = past_key_values.get_seq_length() if past_key_values is not None else 0

        combined_attention_mask = None
        if input_shape[-1] > 1:  # only create a causal mask when we go over a single token
            combined_attention_mask = create_causal_mask(
                config=self.config,
                inputs_embeds=inputs_embeds,
                attention_mask=attention_mask,
                past_key_values=past_key_values,
            )

        encoder_attention_mask = create_bidirectional_mask(
            config=self.config,
            inputs_embeds=inputs_embeds,
            attention_mask=encoder_attention_mask,
            encoder_hidden_states=encoder_hidden_states,
        )

        # embed positions
        positions = self.embed_positions(input_shape, past_key_values_length)

        hidden_states = inputs_embeds + positions
        hidden_states = self.layernorm_embedding(hidden_states)

        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)

        # decoder layers
        all_hidden_states = () if output_hidden_states else None
        all_self_attns = () if output_attentions else None
        all_cross_attentions = () if output_attentions else None

        for idx, decoder_layer in enumerate(self.layers):
            # add LayerDrop (see https://huggingface.co/papers/1909.11556 for description)
            if output_hidden_states:
                all_hidden_states += (hidden_states,)
            if self.training:
                dropout_probability = torch.rand([])
                if dropout_probability < self.layerdrop:
                    continue

            layer_outputs = decoder_layer(
                hidden_states,
                combined_attention_mask,
                encoder_hidden_states,  # as a positional argument for gradient checkpointing
                encoder_attention_mask=encoder_attention_mask,
                past_key_values=past_key_values,
                output_attentions=output_attentions,
                use_cache=use_cache,
            )

            hidden_states = layer_outputs[0]
            if output_attentions:
                all_self_attns += (layer_outputs[1],)
                all_cross_attentions += (layer_outputs[2],)

        # add hidden states from the last decoder layer
        if output_hidden_states:
            all_hidden_states += (hidden_states,)

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


@auto_docstring
class LEDModel(LEDPreTrainedModel):
    _tied_weights_keys = {
        "encoder.embed_tokens.weight": "shared.weight",
        "decoder.embed_tokens.weight": "shared.weight",
    }

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

        padding_idx, vocab_size = config.pad_token_id, config.vocab_size
        self.shared = nn.Embedding(vocab_size, config.d_model, padding_idx)

        self.encoder = LEDEncoder(config)
        self.decoder = LEDDecoder(config)

        # Initialize weights and apply final processing
        self.post_init()

    def get_input_embeddings(self):
        return self.shared

    def set_input_embeddings(self, value):
        self.shared = value
        self.encoder.embed_tokens = self.shared
        self.decoder.embed_tokens = self.shared

    @auto_docstring
    def forward(
        self,
        input_ids: torch.LongTensor | None = None,
        attention_mask: torch.Tensor | None = None,
        decoder_input_ids: torch.LongTensor | None = None,
        decoder_attention_mask: torch.LongTensor | None = None,
        encoder_outputs: tuple[tuple[torch.FloatTensor]] | None = None,
        global_attention_mask: torch.FloatTensor | None = None,
        past_key_values: Cache | None = None,
        inputs_embeds: torch.FloatTensor | None = None,
        decoder_inputs_embeds: torch.FloatTensor | None = None,
        use_cache: bool | None = None,
        output_attentions: bool | None = None,
        output_hidden_states: bool | None = None,
        return_dict: bool | None = None,
        **kwargs,
    ) -> tuple[torch.Tensor] | LEDSeq2SeqModelOutput:
        r"""
        decoder_input_ids (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Indices of decoder input sequence tokens in the vocabulary.

            Indices can be obtained using [`LedTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)

            LED uses the `eos_token_id` as the starting token for `decoder_input_ids` generation. If `past_key_values`
            is used, optionally only the last `decoder_input_ids` have to be input (see `past_key_values`).
        decoder_attention_mask (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Default behavior: generate a tensor that ignores pad tokens in `decoder_input_ids`. Causal mask will also
            be used by default.

            If you want to change padding behavior, you should read [`modeling_led._prepare_decoder_inputs`] and modify
            to your needs. See diagram 1 in [the paper](https://huggingface.co/papers/1910.13461) for more information on the
            default strategy.
        global_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to decide the attention given on each token, local attention or global attention for the encoder.
            Tokens with global attention attends to all other tokens, and all other tokens attend to them. This is
            important for task-specific finetuning because it makes the model more flexible at representing the task.
            For example, for classification, the <s> token should be given global attention. For QA, all question
            tokens should also have global attention. Please refer to the [Longformer
            paper](https://huggingface.co/papers/2004.05150) for more details. Mask values selected in `[0, 1]`:

            - 0 for local attention (a sliding window attention),
            - 1 for global attention (tokens that attend to all other tokens, and all other tokens attend to them).
        """
        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
        )
        use_cache = use_cache if use_cache is not None else self.config.use_cache
        return_dict = return_dict if return_dict is not None else self.config.return_dict

        # Using this like Bart, as LED is derived from it. So far
        # No checkpoint on the hub exists that uses that in practice.
        # https://github.com/huggingface/transformers/blob/ac3cb660cad283163f7c73cad511124e845ca388/src/transformers/models/bart/modeling_bart.py#L1153
        if decoder_input_ids is None and decoder_inputs_embeds is None:
            decoder_input_ids = shift_tokens_right(
                input_ids, self.config.pad_token_id, self.config.decoder_start_token_id
            )

        if encoder_outputs is None:
            encoder_outputs = self.encoder(
                input_ids=input_ids,
                attention_mask=attention_mask,
                global_attention_mask=global_attention_mask,
                inputs_embeds=inputs_embeds,
                output_attentions=output_attentions,
                output_hidden_states=output_hidden_states,
                return_dict=return_dict,
            )
        # If the user passed a tuple for encoder_outputs, we wrap it in a LEDEncoderBaseModelOutput when return_dict=False
        elif return_dict and not isinstance(encoder_outputs, LEDEncoderBaseModelOutput):
            encoder_outputs = LEDEncoderBaseModelOutput(
                last_hidden_state=encoder_outputs[0],
                hidden_states=encoder_outputs[1] if len(encoder_outputs) > 1 else None,
                attentions=encoder_outputs[2] if len(encoder_outputs) > 2 else None,
                global_attentions=encoder_outputs[3] if len(encoder_outputs) > 3 else None,
            )

        # decoder outputs consists of (dec_features, past_key_values, dec_hidden, dec_attn)
        decoder_outputs = self.decoder(
            input_ids=decoder_input_ids,
            attention_mask=decoder_attention_mask,
            encoder_hidden_states=encoder_outputs[0],
            encoder_attention_mask=attention_mask,
            past_key_values=past_key_values,
            inputs_embeds=decoder_inputs_embeds,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )

        if not return_dict:
            return decoder_outputs + encoder_outputs

        return LEDSeq2SeqModelOutput(
            last_hidden_state=decoder_outputs.last_hidden_state,
            past_key_values=decoder_outputs.past_key_values,
            decoder_hidden_states=decoder_outputs.hidden_states,
            decoder_attentions=decoder_outputs.attentions,
            cross_attentions=decoder_outputs.cross_attentions,
            encoder_last_hidden_state=encoder_outputs.last_hidden_state,
            encoder_hidden_states=encoder_outputs.hidden_states,
            encoder_attentions=encoder_outputs.attentions,
            encoder_global_attentions=encoder_outputs.global_attentions,
        )


@auto_docstring(
    custom_intro="""
    The LED Model with a language modeling head. Can be used for summarization.
    """
)
class LEDForConditionalGeneration(LEDPreTrainedModel, GenerationMixin):
    base_model_prefix = "led"
    _keys_to_ignore_on_load_missing = ["final_logits_bias"]
    _tied_weights_keys = {
        "lm_head.weight": "led.shared.weight",
    }

    def __init__(self, config: LEDConfig):
        super().__init__(config)
        self.led = LEDModel(config)
        self.register_buffer("final_logits_bias", torch.zeros((1, self.led.shared.num_embeddings)))
        self.lm_head = nn.Linear(config.d_model, self.led.shared.num_embeddings, bias=False)

        # Initialize weights and apply final processing
        self.post_init()

    def resize_token_embeddings(
        self, new_num_tokens: int, pad_to_multiple_of: int | None = None, mean_resizing: bool = True
    ) -> nn.Embedding:
        new_embeddings = super().resize_token_embeddings(new_num_tokens, pad_to_multiple_of, mean_resizing)
        self._resize_final_logits_bias(new_embeddings.weight.shape[0])
        return new_embeddings

    def _resize_final_logits_bias(self, new_num_tokens: int) -> None:
        old_num_tokens = self.final_logits_bias.shape[-1]
        if new_num_tokens <= old_num_tokens:
            new_bias = self.final_logits_bias[:, :new_num_tokens]
        else:
            extra_bias = torch.zeros((1, new_num_tokens - old_num_tokens), device=self.final_logits_bias.device)
            new_bias = torch.cat([self.final_logits_bias, extra_bias], dim=1)
        self.register_buffer("final_logits_bias", new_bias)

    @auto_docstring
    def forward(
        self,
        input_ids: torch.LongTensor | None = None,
        attention_mask: torch.Tensor | None = None,
        decoder_input_ids: torch.LongTensor | None = None,
        decoder_attention_mask: torch.LongTensor | None = None,
        encoder_outputs: tuple[tuple[torch.FloatTensor]] | None = None,
        global_attention_mask: torch.FloatTensor | None = None,
        past_key_values: Cache | None = None,
        inputs_embeds: torch.FloatTensor | None = None,
        decoder_inputs_embeds: torch.FloatTensor | None = None,
        labels: torch.LongTensor | None = None,
        use_cache: bool | None = None,
        output_attentions: bool | None = None,
        output_hidden_states: bool | None = None,
        return_dict: bool | None = None,
        **kwargs,
    ) -> tuple[torch.Tensor] | LEDSeq2SeqLMOutput:
        r"""
        decoder_input_ids (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Indices of decoder input sequence tokens in the vocabulary.

            Indices can be obtained using [`LedTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)

            LED uses the `eos_token_id` as the starting token for `decoder_input_ids` generation. If `past_key_values`
            is used, optionally only the last `decoder_input_ids` have to be input (see `past_key_values`).
        decoder_attention_mask (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Default behavior: generate a tensor that ignores pad tokens in `decoder_input_ids`. Causal mask will also
            be used by default.

            If you want to change padding behavior, you should read [`modeling_led._prepare_decoder_inputs`] and modify
            to your needs. See diagram 1 in [the paper](https://huggingface.co/papers/1910.13461) for more information on the
            default strategy.
        global_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to decide the attention given on each token, local attention or global attention for the encoder.
            Tokens with global attention attends to all other tokens, and all other tokens attend to them. This is
            important for task-specific finetuning because it makes the model more flexible at representing the task.
            For example, for classification, the <s> token should be given global attention. For QA, all question
            tokens should also have global attention. Please refer to the [Longformer
            paper](https://huggingface.co/papers/2004.05150) for more details. Mask values selected in `[0, 1]`:

            - 0 for local attention (a sliding window attention),
            - 1 for global attention (tokens that attend to all other tokens, and all other tokens attend to them).
        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
            config.vocab_size]` or -100 (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]`.

        Example Summarization:

        ```python
        >>> import torch
        >>> from transformers import AutoTokenizer, LEDForConditionalGeneration

        >>> model = LEDForConditionalGeneration.from_pretrained("allenai/led-large-16384-arxiv")
        >>> tokenizer = AutoTokenizer.from_pretrained("allenai/led-large-16384-arxiv")

        >>> ARTICLE_TO_SUMMARIZE = '''Transformers (Vaswani et al., 2017) have achieved state-of-the-art
        ...     results in a wide range of natural language tasks including generative language modeling
        ...     (Dai et al., 2019; Radford et al., 2019) and discriminative ... language understanding (Devlin et al., 2019).
        ...     This success is partly due to the self-attention component which enables the network to capture contextual
        ...     information from the entire sequence. While powerful, the memory and computational requirements of
        ...     self-attention grow quadratically with sequence length, making it infeasible (or very expensive) to
        ...     process long sequences. To address this limitation, we present Longformer, a modified Transformer
        ...     architecture with a self-attention operation that scales linearly with the sequence length, making it
        ...     versatile for processing long documents (Fig 1). This is an advantage for natural language tasks such as
        ...     long document classification, question answering (QA), and coreference resolution, where existing approaches
        ...     partition or shorten the long context into smaller sequences that fall within the typical 512 token limit
        ...     of BERT-style pretrained models. Such partitioning could potentially result in loss of important
        ...     cross-partition information, and to mitigate this problem, existing methods often rely on complex
        ...     architectures to address such interactions. On the other hand, our proposed Longformer is able to build
        ...     contextual representations of the entire context using multiple layers of attention, reducing the need for
        ...     task-specific architectures.'''
        >>> inputs = tokenizer.encode(ARTICLE_TO_SUMMARIZE, return_tensors="pt")

        >>> # Global attention on the first token (cf. Beltagy et al. 2020)
        >>> global_attention_mask = torch.zeros_like(inputs)
        >>> global_attention_mask[:, 0] = 1

        >>> # Generate Summary
        >>> summary_ids = model.generate(inputs, global_attention_mask=global_attention_mask, num_beams=3, max_length=32)
        >>> print(tokenizer.decode(summary_ids[0], skip_special_tokens=True, clean_up_tokenization_spaces=True))
        ```

        Example Conditional generation :

        ```python
        >>> from transformers import AutoTokenizer, LEDForConditionalGeneration

        >>> tokenizer = AutoTokenizer.from_pretrained("allenai/led-base-16384")
        >>> TXT = "My friends are <mask> but they eat too many carbs."

        >>> model = LEDForConditionalGeneration.from_pretrained("allenai/led-base-16384")
        >>> input_ids = tokenizer([TXT], return_tensors="pt")["input_ids"]

        >>> prediction = model.generate(input_ids)[0]
        >>> print(tokenizer.decode(prediction, skip_special_tokens=True))
        ```
        """
        return_dict = return_dict if return_dict is not None else self.config.return_dict

        if labels is not None:
            if use_cache:
                logger.warning("The `use_cache` argument is changed to `False` since `labels` is provided.")
            use_cache = False
            if decoder_input_ids is None and decoder_inputs_embeds is None:
                decoder_input_ids = shift_tokens_right(
                    labels, self.config.pad_token_id, self.config.decoder_start_token_id
                )

        outputs = self.led(
            input_ids,
            attention_mask=attention_mask,
            decoder_input_ids=decoder_input_ids,
            decoder_attention_mask=decoder_attention_mask,
            encoder_outputs=encoder_outputs,
            global_attention_mask=global_attention_mask,
            past_key_values=past_key_values,
            inputs_embeds=inputs_embeds,
            decoder_inputs_embeds=decoder_inputs_embeds,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )
        lm_logits = self.lm_head(outputs[0]) + self.final_logits_bias

        masked_lm_loss = None
        if labels is not None:
            loss_fct = CrossEntropyLoss()
            masked_lm_loss = loss_fct(lm_logits.view(-1, self.config.vocab_size), labels.view(-1))

        if not return_dict:
            output = (lm_logits,) + outputs[1:]
            return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output

        return LEDSeq2SeqLMOutput(
            loss=masked_lm_loss,
            logits=lm_logits,
            past_key_values=outputs.past_key_values,
            decoder_hidden_states=outputs.decoder_hidden_states,
            decoder_attentions=outputs.decoder_attentions,
            cross_attentions=outputs.cross_attentions,
            encoder_last_hidden_state=outputs.encoder_last_hidden_state,
            encoder_hidden_states=outputs.encoder_hidden_states,
            encoder_attentions=outputs.encoder_attentions,
            encoder_global_attentions=outputs.encoder_global_attentions,
        )

    def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor):
        return shift_tokens_right(labels, self.config.pad_token_id, self.config.decoder_start_token_id)


@auto_docstring
class LEDForQuestionAnswering(LEDPreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        config.num_labels = 2
        self.num_labels = config.num_labels

        self.led = LEDModel(config)
        self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    @auto_docstring
    def forward(
        self,
        input_ids: torch.LongTensor | None = None,
        attention_mask: torch.Tensor | None = None,
        decoder_input_ids: torch.LongTensor | None = None,
        decoder_attention_mask: torch.LongTensor | None = None,
        encoder_outputs: tuple[tuple[torch.FloatTensor]] | None = None,
        global_attention_mask: torch.FloatTensor | None = None,
        start_positions: torch.LongTensor | None = None,
        end_positions: torch.LongTensor | None = None,
        inputs_embeds: torch.FloatTensor | None = None,
        decoder_inputs_embeds: torch.FloatTensor | None = None,
        use_cache: bool | None = None,
        output_attentions: bool | None = None,
        output_hidden_states: bool | None = None,
        return_dict: bool | None = None,
        **kwargs,
    ) -> tuple[torch.Tensor] | LEDSeq2SeqQuestionAnsweringModelOutput:
        r"""
        decoder_input_ids (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Indices of decoder input sequence tokens in the vocabulary.

            Indices can be obtained using [`LedTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)

            LED uses the `eos_token_id` as the starting token for `decoder_input_ids` generation. If `past_key_values`
            is used, optionally only the last `decoder_input_ids` have to be input (see `past_key_values`).
        decoder_attention_mask (`torch.LongTensor` of shape `(batch_size, target_sequence_length)`, *optional*):
            Default behavior: generate a tensor that ignores pad tokens in `decoder_input_ids`. Causal mask will also
            be used by default.

            If you want to change padding behavior, you should read [`modeling_led._prepare_decoder_inputs`] and modify
            to your needs. See diagram 1 in [the paper](https://huggingface.co/papers/1910.13461) for more information on the
            default strategy.
        global_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to decide the attention given on each token, local attention or global attention for the encoder.
            Tokens with global attention attends to all other tokens, and all other tokens attend to them. This is
            important for task-specific finetuning because it makes the model more flexible at representing the task.
            For example, for classification, the <s> token should be given global attention. For QA, all question
            tokens should also have global attention. Please refer to the [Longformer
            paper](https://huggingface.co/papers/2004.05150) for more details. Mask values selected in `[0, 1]`:

            - 0 for local attention (a sliding window attention),
            - 1 for global attention (tokens that attend to all other tokens, and all other tokens attend to them).
        """
        return_dict = return_dict if return_dict is not None else self.config.return_dict
        if start_positions is not None and end_positions is not None:
            use_cache = False

        outputs = self.led(
            input_ids,
            attention_mask=attention_mask,
            decoder_input_ids=decoder_input_ids,
            decoder_attention_mask=decoder_attention_mask,
            global_attention_mask=global_attention_mask,
            encoder_outputs=encoder_outputs,
            inputs_embeds=inputs_embeds,
            decoder_inputs_embeds=decoder_inputs_embeds,
            use_cache=use_cache,
            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[1:]
            return ((total_loss,) + output) if total_loss is not None else output

        return LEDSeq2SeqQuestionAnsweringModelOutput(
            loss=total_loss,
            start_logits=start_logits,
            end_logits=end_logits,
            past_key_values=outputs.past_key_values,
            decoder_hidden_states=outputs.decoder_hidden_states,
            decoder_attentions=outputs.decoder_attentions,
            cross_attentions=outputs.cross_attentions,
            encoder_last_hidden_state=outputs.encoder_last_hidden_state,
            encoder_hidden_states=outputs.encoder_hidden_states,
            encoder_attentions=outputs.encoder_attentions,
            encoder_global_attentions=outputs.encoder_global_attentions,
        )


__all__ = [
    "LEDForConditionalGeneration",
    "LEDForQuestionAnswering",
    "LEDModel",
    "LEDPreTrainedModel",
]