refactor: reorganize project structure and update import paths

mineru/model/mfr/unimernet/unimernet_hf/unimer_mbart/modeling_unimer_mbart.py

+# coding=utf-8
+# Copyright 2021, The Facebook AI Research Team 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 UnimerMBART model."""
+
+import copy
+import math
+from dataclasses import dataclass
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.nn.functional as F
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
+
+from transformers.activations import ACT2FN
+from transformers.modeling_attn_mask_utils import (
+    _prepare_4d_attention_mask,
+    _prepare_4d_attention_mask_for_sdpa,
+    _prepare_4d_causal_attention_mask,
+    _prepare_4d_causal_attention_mask_for_sdpa,
+)
+from transformers.modeling_outputs import (
+    BaseModelOutput,
+    BaseModelOutputWithPastAndCrossAttentions,
+    CausalLMOutputWithCrossAttentions,
+    Seq2SeqLMOutput,
+    Seq2SeqModelOutput,
+    Seq2SeqQuestionAnsweringModelOutput,
+    Seq2SeqSequenceClassifierOutput,
+)
+from transformers import GenerationMixin, PreTrainedModel
+from transformers.utils import (
+    add_code_sample_docstrings,
+    add_end_docstrings,
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    is_flash_attn_2_available,
+    is_flash_attn_greater_or_equal_2_10,
+    logging,
+    replace_return_docstrings,
+)
+from .configuration_unimer_mbart import UnimerMBartConfig
+
+
+if is_flash_attn_2_available():
+    from flash_attn import flash_attn_func, flash_attn_varlen_func
+    from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input  # noqa
+
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "facebook/mbart-large-cc25"
+_CONFIG_FOR_DOC = "MBartConfig"
+
+# Base model docstring
+_EXPECTED_OUTPUT_SHAPE = [1, 8, 1024]
+
+
+# Copied from transformers.models.llama.modeling_llama._get_unpad_data
+def _get_unpad_data(attention_mask):
+    seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32)
+    indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten()
+    max_seqlen_in_batch = seqlens_in_batch.max().item()
+    cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.int32), (1, 0))
+    return (
+        indices,
+        cu_seqlens,
+        max_seqlen_in_batch,
+    )
+
+
+def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int):
+    """
+    Shift input ids one token to the right, and wrap the last non pad token (the <LID> token) Note that MBart does not
+    have a single `decoder_start_token_id` in contrast to other Bart-like models.
+    """
+    prev_output_tokens = input_ids.clone()
+
+    if pad_token_id is None:
+        raise ValueError("self.model.config.pad_token_id has to be defined.")
+    # replace possible -100 values in labels by `pad_token_id`
+    prev_output_tokens.masked_fill_(prev_output_tokens == -100, pad_token_id)
+
+    index_of_eos = (prev_output_tokens.ne(pad_token_id).sum(dim=1) - 1).unsqueeze(-1)
+    decoder_start_tokens = prev_output_tokens.gather(1, index_of_eos).squeeze()
+    prev_output_tokens[:, 1:] = prev_output_tokens[:, :-1].clone()
+    prev_output_tokens[:, 0] = decoder_start_tokens
+
+    return prev_output_tokens
+
+@dataclass
+class CausalLMOutputWithCrossAttentionsAndCounting(CausalLMOutputWithCrossAttentions):
+    """
+    Base class for causal language model (or autoregressive) outputs.
+
+    Args:
+        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
+            Language modeling loss (for next-token prediction).
+        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).
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        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,
+            sequence_length)`.
+
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        cross_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,
+            sequence_length)`.
+
+            Cross attentions weights after the attention softmax, used to compute the weighted average in the
+            cross-attention heads.
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `torch.FloatTensor` tuples of length `config.n_layers`, with each tuple containing the cached key,
+            value states of the self-attention and the cross-attention layers if model is used in encoder-decoder
+            setting. Only relevant if `config.is_decoder = True`.
+
+            Contains pre-computed hidden-states (key and values in the attention blocks) that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+        counting:
+            Counting
+    """
+    counting: Optional[torch.FloatTensor] = None
+
+# Copied from transformers.models.bart.modeling_bart.BartLearnedPositionalEmbedding with Bart->MBart
+class UnimerMBartLearnedPositionalEmbedding(nn.Embedding):
+    """
+    This module learns positional embeddings up to a fixed maximum size.
+    """
+
+    def __init__(self, num_embeddings: int, embedding_dim: int):
+        # MBart is set up so that if padding_idx is specified then offset the embedding ids by 2
+        # and adjust num_embeddings appropriately. Other models don't have this hack
+        self.offset = 2
+        super().__init__(num_embeddings + self.offset, embedding_dim)
+
+    def forward(self, input_ids: torch.Tensor, 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
+        ).expand(bsz, -1)
+
+        return super().forward(positions + self.offset)
+
+
+# Copied from transformers.models.bart.modeling_bart.BartScaledWordEmbedding with Bart->MBart
+class UnimerMBartScaledWordEmbedding(nn.Embedding):
+    """
+    This module overrides nn.Embeddings' forward by multiplying with embeddings scale.
+    """
+
+    def __init__(self, num_embeddings: int, embedding_dim: int, padding_idx: int, embed_scale: Optional[float] = 1.0):
+        super().__init__(num_embeddings, embedding_dim, padding_idx)
+        self.embed_scale = embed_scale
+
+    def forward(self, input_ids: torch.Tensor):
+        return super().forward(input_ids) * self.embed_scale
+
+
+# Copied from transformers.models.bart.modeling_bart.BartAttention with Bart->MBart
+class UnimerMBartAttention(nn.Module):
+    """Multi-headed attention from 'Attention Is All You Need' paper, with qk_squeeze"""
+
+    def __init__(
+        self,
+        embed_dim: int,
+        num_heads: int,
+        dropout: float = 0.0,
+        is_decoder: bool = False,
+        bias: bool = True,
+        is_causal: bool = False,
+        *,
+        config: UnimerMBartConfig,
+    ):
+        super().__init__()
+        self.embed_dim = embed_dim
+        self.num_heads = num_heads
+        self.dropout = dropout
+        self.head_dim = embed_dim // num_heads
+        self.config = config
+
+        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}"
+                f" and `num_heads`: {num_heads})."
+            )
+        
+        self.squeeze_dim = embed_dim // config.qk_squeeze
+        self.squeeze_head_dim = self.squeeze_dim // num_heads
+        self.scaling = self.squeeze_head_dim**-0.5
+        self.is_decoder = is_decoder
+        self.is_causal = is_causal
+
+        self.q_proj = nn.Linear(embed_dim, self.squeeze_dim, bias=bias)
+        self.k_proj = nn.Linear(embed_dim, self.squeeze_dim, bias=bias)
+        self.v_proj = nn.Linear(embed_dim, embed_dim, bias=bias)
+        self.out_proj = nn.Linear(embed_dim, embed_dim, bias=bias)
+
+    def _shape_qk(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+        return tensor.view(bsz, seq_len, self.num_heads, self.squeeze_head_dim).transpose(1, 2).contiguous()
+
+    def _shape_v(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+        return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous()
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        key_value_states: Optional[torch.Tensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        layer_head_mask: Optional[torch.Tensor] = None,
+        output_attentions: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        """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, _ = hidden_states.size()
+
+        # get query proj
+        query_states = self.q_proj(hidden_states) * self.scaling
+        # get key, value proj
+        # `past_key_value[0].shape[2] == key_value_states.shape[1]`
+        # is checking that the `sequence_length` of the `past_key_value` is the same as
+        # the provided `key_value_states` to support prefix tuning
+        if (
+            is_cross_attention
+            and past_key_value is not None
+            and past_key_value[0].shape[2] == key_value_states.shape[1]
+        ):
+            # reuse k,v, cross_attentions
+            key_states = past_key_value[0]
+            value_states = past_key_value[1]
+        elif is_cross_attention:
+            # cross_attentions
+            key_states = self._shape_qk(self.k_proj(key_value_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(key_value_states), -1, bsz)
+        elif past_key_value is not None:
+            # reuse k, v, self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+            key_states = torch.cat([past_key_value[0], key_states], dim=2)
+            value_states = torch.cat([past_key_value[1], value_states], dim=2)
+        else:
+            # self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+
+        if self.is_decoder:
+            # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states.
+            # Further calls to cross_attention layer can then reuse all cross-attention
+            # key/value_states (first "if" case)
+            # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of
+            # all previous decoder key/value_states. Further calls to uni-directional self-attention
+            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
+            # if encoder bi-directional self-attention `past_key_value` is always `None`
+            past_key_value = (key_states, value_states)
+
+        proj_shape = (bsz * self.num_heads, -1, self.squeeze_head_dim)
+        value_shape = (bsz * self.num_heads, -1, self.head_dim)
+        query_states = self._shape_qk(query_states, tgt_len, bsz).view(*proj_shape)
+        key_states = key_states.reshape(*proj_shape)
+        value_states = value_states.reshape(*value_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 layer_head_mask is not None:
+            if layer_head_mask.size() != (self.num_heads,):
+                raise ValueError(
+                    f"Head mask for a single layer should be of size {(self.num_heads,)}, but is"
+                    f" {layer_head_mask.size()}"
+                )
+            attn_weights = layer_head_mask.view(1, -1, 1, 1) * attn_weights.view(bsz, self.num_heads, tgt_len, src_len)
+            attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len)
+
+        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)
+        attn_output = attn_output.transpose(1, 2)
+
+        # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be
+        # partitioned across GPUs when using tensor-parallelism.
+        attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim)
+
+        attn_output = self.out_proj(attn_output)
+
+        return attn_output, attn_weights_reshaped, past_key_value
+
+
+# Copied from transformers.models.bart.modeling_bart.BartFlashAttention2 with Bart->MBart
+class UnimerMBartFlashAttention2(UnimerMBartAttention):
+    """
+    MBart flash attention module. This module inherits from `MBartSqueezeAttention` as the weights of the module stays
+    untouched. The only required change would be on the forward pass where it needs to correctly call the public API of
+    flash attention and deal with padding tokens in case the input contains any of them.
+    """
+
+    # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2.__init__
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+        # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1.
+        # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0.
+        # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left).
+        self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10()
+
+    # def _reshape(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+    #     return tensor.view(bsz, seq_len, self.num_heads, self.head_dim)
+
+    def _shape_qk(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+        return tensor.view(bsz, seq_len, self.num_heads, self.squeeze_head_dim)
+
+    def _shape_v(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+        return tensor.view(bsz, seq_len, self.num_heads, self.head_dim)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        key_value_states: Optional[torch.Tensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        layer_head_mask: Optional[torch.Tensor] = None,
+        output_attentions: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        # MBartFlashAttention2 attention does not support output_attentions
+        if output_attentions:
+            raise ValueError("MBartFlashAttention2 attention does not support output_attentions")
+
+        # 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, q_len, _ = hidden_states.size()
+
+        # get query proj
+        query_states = self._shape_qk(self.q_proj(hidden_states), -1, bsz)
+
+        # get key, value proj
+        # `past_key_value[0].shape[2] == key_value_states.shape[1]`
+        # is checking that the `sequence_length` of the `past_key_value` is the same as
+        # the provided `key_value_states` to support prefix tuning
+        if (
+            is_cross_attention
+            and past_key_value is not None
+            and past_key_value[0].shape[2] == key_value_states.shape[1]
+        ):
+            # reuse k,v, cross_attentions
+            key_states = past_key_value[0].transpose(1, 2)
+            value_states = past_key_value[1].transpose(1, 2)
+        elif is_cross_attention:
+            # cross_attentions
+            key_states = self._shape_qk(self.k_proj(key_value_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(key_value_states), -1, bsz)
+        elif past_key_value is not None:
+            # reuse k, v, self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+            key_states = torch.cat([past_key_value[0].transpose(1, 2), key_states], dim=1)
+            value_states = torch.cat([past_key_value[1].transpose(1, 2), value_states], dim=1)
+        else:
+            # self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+
+        if self.is_decoder:
+            # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states.
+            # Further calls to cross_attention layer can then reuse all cross-attention
+            # key/value_states (first "if" case)
+            # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of
+            # all previous decoder key/value_states. Further calls to uni-directional self-attention
+            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
+            # if encoder bi-directional self-attention `past_key_value` is always `None`
+            past_key_value = (key_states.transpose(1, 2), value_states.transpose(1, 2))
+
+        kv_seq_len = key_states.shape[-2]
+        if past_key_value is not None:
+            kv_seq_len += past_key_value[0].shape[-2]
+
+        # In PEFT, usually we cast the layer norms in float32 for training stability reasons
+        # therefore the input hidden states gets silently casted in float32. Hence, we need
+        # cast them back in the correct dtype just to be sure everything works as expected.
+        # This might slowdown training & inference so it is recommended to not cast the LayerNorms
+        # in fp32. (LlamaRMSNorm handles it correctly)
+
+        input_dtype = query_states.dtype
+        if input_dtype == torch.float32:
+            if torch.is_autocast_enabled():
+                target_dtype = torch.get_autocast_gpu_dtype()
+            # Handle the case where the model is quantized
+            elif hasattr(self.config, "_pre_quantization_dtype"):
+                target_dtype = self.config._pre_quantization_dtype
+            else:
+                target_dtype = self.q_proj.weight.dtype
+
+            logger.warning_once(
+                f"The input hidden states seems to be silently casted in float32, this might be related to"
+                f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in"
+                f" {target_dtype}."
+            )
+
+            query_states = query_states.to(target_dtype)
+            key_states = key_states.to(target_dtype)
+            value_states = value_states.to(target_dtype)
+
+        attn_output = self._flash_attention_forward(
+            query_states, key_states, value_states, attention_mask, q_len, dropout=self.dropout
+        )
+
+        attn_output = attn_output.reshape(bsz, q_len, -1)
+        attn_output = self.out_proj(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+    # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._flash_attention_forward
+    def _flash_attention_forward(
+        self, query_states, key_states, value_states, attention_mask, query_length, dropout=0.0, softmax_scale=None
+    ):
+        """
+        Calls the forward method of Flash Attention - if the input hidden states contain at least one padding token
+        first unpad the input, then computes the attention scores and pad the final attention scores.
+
+        Args:
+            query_states (`torch.Tensor`):
+                Input query states to be passed to Flash Attention API
+            key_states (`torch.Tensor`):
+                Input key states to be passed to Flash Attention API
+            value_states (`torch.Tensor`):
+                Input value states to be passed to Flash Attention API
+            attention_mask (`torch.Tensor`):
+                The padding mask - corresponds to a tensor of size `(batch_size, seq_len)` where 0 stands for the
+                position of padding tokens and 1 for the position of non-padding tokens.
+            dropout (`float`):
+                Attention dropout
+            softmax_scale (`float`, *optional*):
+                The scaling of QK^T before applying softmax. Default to 1 / sqrt(head_dim)
+        """
+        if not self._flash_attn_uses_top_left_mask:
+            causal = self.is_causal
+        else:
+            # TODO: Remove the `query_length != 1` check once Flash Attention for RoCm is bumped to 2.1. For details, please see the comment in LlamaFlashAttention2 __init__.
+            causal = self.is_causal and query_length != 1
+
+        # Contains at least one padding token in the sequence
+        if attention_mask is not None:
+            batch_size = query_states.shape[0]
+
+            query_states, key_states, value_states, indices_q, cu_seq_lens, max_seq_lens = self._upad_input(
+                query_states, key_states, value_states, attention_mask, query_length
+            )
+
+            cu_seqlens_q, cu_seqlens_k = cu_seq_lens
+            max_seqlen_in_batch_q, max_seqlen_in_batch_k = max_seq_lens
+
+            attn_output_unpad = flash_attn_varlen_func(
+                query_states,
+                key_states,
+                value_states,
+                cu_seqlens_q=cu_seqlens_q,
+                cu_seqlens_k=cu_seqlens_k,
+                max_seqlen_q=max_seqlen_in_batch_q,
+                max_seqlen_k=max_seqlen_in_batch_k,
+                dropout_p=dropout,
+                softmax_scale=softmax_scale,
+                causal=causal,
+            )
+
+            attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length)
+        else:
+            attn_output = flash_attn_func(
+                query_states, key_states, value_states, dropout, softmax_scale=softmax_scale, causal=causal
+            )
+
+        return attn_output
+
+    # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._upad_input
+    def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length):
+        indices_k, cu_seqlens_k, max_seqlen_in_batch_k = _get_unpad_data(attention_mask)
+        batch_size, kv_seq_len, num_key_value_heads, head_dim = key_layer.shape
+
+        key_layer = index_first_axis(
+            key_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+        )
+        value_layer = index_first_axis(
+            value_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+        )
+        if query_length == kv_seq_len:
+            query_layer = index_first_axis(
+                query_layer.reshape(batch_size * kv_seq_len, self.num_heads, head_dim), indices_k
+            )
+            cu_seqlens_q = cu_seqlens_k
+            max_seqlen_in_batch_q = max_seqlen_in_batch_k
+            indices_q = indices_k
+        elif query_length == 1:
+            max_seqlen_in_batch_q = 1
+            cu_seqlens_q = torch.arange(
+                batch_size + 1, dtype=torch.int32, device=query_layer.device
+            )  # There is a memcpy here, that is very bad.
+            indices_q = cu_seqlens_q[:-1]
+            query_layer = query_layer.squeeze(1)
+        else:
+            # The -q_len: slice assumes left padding.
+            attention_mask = attention_mask[:, -query_length:]
+            query_layer, indices_q, cu_seqlens_q, max_seqlen_in_batch_q = unpad_input(query_layer, attention_mask)
+
+        return (
+            query_layer,
+            key_layer,
+            value_layer,
+            indices_q,
+            (cu_seqlens_q, cu_seqlens_k),
+            (max_seqlen_in_batch_q, max_seqlen_in_batch_k),
+        )
+
+class UnimerMBartSdpaAttention(UnimerMBartAttention):
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        key_value_states: Optional[torch.Tensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        layer_head_mask: Optional[torch.Tensor] = None,
+        output_attentions: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        """Input shape: Batch x Time x Channel"""
+        if output_attentions or layer_head_mask is not None:
+            # TODO: Improve this warning with e.g. `model.config._attn_implementation = "manual"` once this is implemented.
+            logger.warning(
+                "BartModel is using BartSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True` or `layer_head_mask` not None. Falling back to the manual attention"
+                ' implementation, but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.'
+            )
+            return super().forward(
+                hidden_states,
+                key_value_states=key_value_states,
+                past_key_value=past_key_value,
+                attention_mask=attention_mask,
+                layer_head_mask=layer_head_mask,
+                output_attentions=output_attentions,
+            )
+
+        # 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, _ = hidden_states.size()
+
+        # get query proj
+        query_states = self.q_proj(hidden_states)
+        # get key, value proj
+        # `past_key_value[0].shape[2] == key_value_states.shape[1]`
+        # is checking that the `sequence_length` of the `past_key_value` is the same as
+        # the provided `key_value_states` to support prefix tuning
+        if (
+            is_cross_attention
+            and past_key_value is not None
+            and past_key_value[0].shape[2] == key_value_states.shape[1]
+        ):
+            # reuse k,v, cross_attentions
+            key_states = past_key_value[0]
+            value_states = past_key_value[1]
+        elif is_cross_attention:
+            # cross_attentions
+            key_states = self._shape_qk(self.k_proj(key_value_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(key_value_states), -1, bsz)
+        elif past_key_value is not None:
+            # reuse k, v, self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+            key_states = torch.cat([past_key_value[0], key_states], dim=2)
+            value_states = torch.cat([past_key_value[1], value_states], dim=2)
+        else:
+            # self_attention
+            key_states = self._shape_qk(self.k_proj(hidden_states), -1, bsz)
+            value_states = self._shape_v(self.v_proj(hidden_states), -1, bsz)
+
+        if self.is_decoder:
+            # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states.
+            # Further calls to cross_attention layer can then reuse all cross-attention
+            # key/value_states (first "if" case)
+            # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of
+            # all previous decoder key/value_states. Further calls to uni-directional self-attention
+            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
+            # if encoder bi-directional self-attention `past_key_value` is always `None`
+            past_key_value = (key_states, value_states)
+
+        query_states = self._shape_qk(query_states, tgt_len, bsz)
+
+        # We dispatch to SDPA's Flash Attention or Efficient kernels via this `is_causal` if statement instead of an inline conditional assignment
+        # in SDPA to support both torch.compile's dynamic shapes and full graph options. An inline conditional prevents dynamic shapes from compiling.
+        # The tgt_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case tgt_len == 1.
+        is_causal = True if self.is_causal and attention_mask is None and tgt_len > 1 else False
+
+        # NOTE: SDPA with memory-efficient backend is currently (torch==2.1.2) bugged when using non-contiguous inputs and a custom attn_mask,
+        # but we are fine here as `_shape` do call `.contiguous()`. Reference: https://github.com/pytorch/pytorch/issues/112577
+        attn_output = torch.nn.functional.scaled_dot_product_attention(
+            query_states,
+            key_states,
+            value_states,
+            attn_mask=attention_mask,
+            dropout_p=self.dropout if self.training else 0.0,
+            is_causal=is_causal,
+        )
+
+        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.transpose(1, 2)
+
+        # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be
+        # partitioned across GPUs when using tensor-parallelism.
+        attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim)
+
+        attn_output = self.out_proj(attn_output)
+
+        return attn_output, None, past_key_value
+
+UNIMER_MBART_ATTENTION_CLASSES = {
+    "eager": UnimerMBartAttention,
+    "flash_attention_2": UnimerMBartFlashAttention2,
+    "sdpa": UnimerMBartSdpaAttention,
+}
+
+
+class UnimerMBartEncoderLayer(nn.Module):
+    def __init__(self, config: UnimerMBartConfig):
+        super().__init__()
+        self.embed_dim = config.d_model
+
+        self.self_attn = UNIMER_MBART_ATTENTION_CLASSES[config._attn_implementation](
+            embed_dim=self.embed_dim,
+            num_heads=config.encoder_attention_heads,
+            dropout=config.attention_dropout,
+            config=config,
+        )
+        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,
+        layer_head_mask: torch.Tensor,
+        output_attentions: bool = False,
+    ) -> torch.Tensor:
+        """
+        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.
+            layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size
+                `(encoder_attention_heads,)`.
+            output_attentions (`bool`, *optional*):
+                Whether or not to return the attentions tensors of all attention layers. See `attentions` under
+                returned tensors for more detail.
+        """
+        residual = hidden_states
+        hidden_states = self.self_attn_layer_norm(hidden_states)
+        hidden_states, attn_weights, _ = self.self_attn(
+            hidden_states=hidden_states,
+            attention_mask=attention_mask,
+            layer_head_mask=layer_head_mask,
+            output_attentions=output_attentions,
+        )
+        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
+        hidden_states = residual + hidden_states
+
+        residual = hidden_states
+        hidden_states = self.final_layer_norm(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
+
+        if hidden_states.dtype == torch.float16 and (
+            torch.isinf(hidden_states).any() or torch.isnan(hidden_states).any()
+        ):
+            clamp_value = torch.finfo(hidden_states.dtype).max - 1000
+            hidden_states = torch.clamp(hidden_states, min=-clamp_value, max=clamp_value)
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (attn_weights,)
+
+        return outputs
+
+
+class UnimerMBartDecoderLayer(nn.Module):
+    def __init__(self, config: UnimerMBartConfig):
+        super().__init__()
+        self.embed_dim = config.d_model
+
+        self.self_attn = UNIMER_MBART_ATTENTION_CLASSES[config._attn_implementation](
+            embed_dim=self.embed_dim,
+            num_heads=config.decoder_attention_heads,
+            dropout=config.attention_dropout,
+            is_decoder=True,
+            is_causal=True,
+            config=config,
+        )
+        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 = UNIMER_MBART_ATTENTION_CLASSES[config._attn_implementation](
+            self.embed_dim,
+            config.decoder_attention_heads,
+            dropout=config.attention_dropout,
+            is_decoder=True,
+            config=config,
+        )
+        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: Optional[torch.Tensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.Tensor] = None,
+        layer_head_mask: Optional[torch.Tensor] = None,
+        cross_attn_layer_head_mask: Optional[torch.Tensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = True,
+    ) -> torch.Tensor:
+        """
+        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.
+            layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size
+                `(encoder_attention_heads,)`.
+            cross_attn_layer_head_mask (`torch.FloatTensor`): mask for cross-attention heads in a given layer of
+                size `(decoder_attention_heads,)`.
+            past_key_value (`Tuple(torch.FloatTensor)`): cached past key and value projection states
+            output_attentions (`bool`, *optional*):
+                Whether or not to return the attentions tensors of all attention layers. See `attentions` under
+                returned tensors for more detail.
+        """
+        residual = hidden_states
+        hidden_states = self.self_attn_layer_norm(hidden_states)
+
+        # Self Attention
+        # decoder uni-directional self-attention cached key/values tuple is at positions 1,2
+        self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None
+        # add present self-attn cache to positions 1,2 of present_key_value tuple
+        hidden_states, self_attn_weights, present_key_value = self.self_attn(
+            hidden_states=hidden_states,
+            past_key_value=self_attn_past_key_value,
+            attention_mask=attention_mask,
+            layer_head_mask=layer_head_mask,
+            output_attentions=output_attentions,
+        )
+        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
+        hidden_states = residual + 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 = self.encoder_attn_layer_norm(hidden_states)
+
+            # cross_attn cached key/values tuple is at positions 3,4 of present_key_value tuple
+            cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None
+            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,
+                layer_head_mask=cross_attn_layer_head_mask,
+                past_key_value=cross_attn_past_key_value,
+                output_attentions=output_attentions,
+            )
+            hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
+            hidden_states = residual + hidden_states
+
+            # add cross-attn to positions 3,4 of present_key_value tuple
+            present_key_value = present_key_value + cross_attn_present_key_value
+
+        # Fully Connected
+        residual = hidden_states
+        hidden_states = self.final_layer_norm(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
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (self_attn_weights, cross_attn_weights)
+
+        if use_cache:
+            outputs += (present_key_value,)
+
+        return outputs
+
+
+# Copied from transformers.models.bart.modeling_bart.BartClassificationHead with Bart->MBart
+class UnimerMBartClassificationHead(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) -> 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
+
+
+class UnimerMBartPreTrainedModel(PreTrainedModel):
+    config_class = UnimerMBartConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["MBartDecoderLayer", "MBartSqueezeAttention"]
+    _supports_flash_attn_2 = True
+    _supports_sdpa = True
+
+    def _init_weights(self, module):
+        std = self.config.init_std
+        if isinstance(module, nn.Linear):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+    @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
+
+
+MBART_START_DOCSTRING = r"""
+    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
+    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
+    etc.)
+
+    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
+    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
+    and behavior.
+
+    Parameters:
+        config ([`MBartConfig`]):
+            Model configuration class with all the parameters of the model. Initializing with a config file does not
+            load the weights associated with the model, only the configuration. Check out the
+            [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+MBART_GENERATION_EXAMPLE = r"""
+    Translation example:
+
+    ```python
+    >>> from transformers import AutoTokenizer, MBartForConditionalGeneration
+
+    >>> model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-en-ro")
+    >>> tokenizer = AutoTokenizer.from_pretrained("facebook/mbart-large-en-ro")
+
+    >>> example_english_phrase = "42 is the answer"
+    >>> inputs = tokenizer(example_english_phrase, return_tensors="pt")
+
+    >>> # Translate
+    >>> generated_ids = model.generate(**inputs, num_beams=4, max_length=5)
+    >>> tokenizer.batch_decode(generated_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
+    '42 este răspuns'
+    ```
+
+    Mask filling example:
+
+    ```python
+    >>> from transformers import AutoTokenizer, MBartForConditionalGeneration
+
+    >>> model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-cc25")
+    >>> tokenizer = AutoTokenizer.from_pretrained("facebook/mbart-large-cc25")
+
+    >>> # de_DE is the language symbol id <LID> for German
+    >>> TXT = "</s> Meine Freunde sind <mask> nett aber sie essen zu viel Kuchen. </s> de_DE"
+
+    >>> input_ids = tokenizer([TXT], add_special_tokens=False, return_tensors="pt")["input_ids"]
+    >>> logits = model(input_ids).logits
+
+    >>> masked_index = (input_ids[0] == tokenizer.mask_token_id).nonzero().item()
+    >>> probs = logits[0, masked_index].softmax(dim=0)
+    >>> values, predictions = probs.topk(5)
+
+    >>> tokenizer.decode(predictions).split()
+    ['nett', 'sehr', 'ganz', 'nicht', 'so']
+    ```
+"""
+
+MBART_INPUTS_DOCSTRING = 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)
+        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 [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            [What are decoder input IDs?](../glossary#decoder-input-ids)
+
+            MBart uses a specific language id token as the starting token for `decoder_input_ids` generation that
+            varies according to source and target language, *e.g.* 25004 for *en_XX*, and 25003 for *de_DE*. If
+            `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
+            `past_key_values`).
+
+            For translation and summarization training, `decoder_input_ids` should be provided. If no
+            `decoder_input_ids` is provided, the model will create this tensor by shifting the `input_ids` to the right
+            for denoising pre-training following the paper.
+        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.
+        head_mask (`torch.Tensor` of shape `(encoder_layers, encoder_attention_heads)`, *optional*):
+            Mask to nullify selected heads of the attention modules in the encoder. Mask values selected in `[0, 1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+
+        decoder_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+            Mask to nullify selected heads of the attention modules in the decoder. Mask values selected in `[0, 1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+
+        cross_attn_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+            Mask to nullify selected heads of the cross-attention modules in the decoder. Mask values selected in `[0,
+            1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+
+        encoder_outputs (`tuple(tuple(torch.FloatTensor)`, *optional*):
+            Tuple consists of (`last_hidden_state`, *optional*: `hidden_states`, *optional*: `attentions`)
+            `last_hidden_state` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) is a sequence of
+            hidden-states at the output of the last layer of the encoder. Used in the cross-attention of the decoder.
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
+            `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
+
+            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.
+        decoder_inputs_embeds (`torch.FloatTensor` of shape `(batch_size, target_sequence_length, hidden_size)`, *optional*):
+            Optionally, instead of passing `decoder_input_ids` you can choose to directly pass an embedded
+            representation. If `past_key_values` is used, optionally only the last `decoder_inputs_embeds` have to be
+            input (see `past_key_values`). This is useful if you want more control over how to convert
+            `decoder_input_ids` indices into associated vectors than the model's internal embedding lookup matrix.
+
+            If `decoder_input_ids` and `decoder_inputs_embeds` are both unset, `decoder_inputs_embeds` takes the value
+            of `inputs_embeds`.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        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.
+"""
+
+
+class UnimerMBartEncoder(UnimerMBartPreTrainedModel):
+    """
+    Transformer encoder consisting of *config.encoder_layers* self attention layers. Each layer is a
+    [`MBartEncoderLayer`].
+
+    Args:
+        config: MBartConfig
+        embed_tokens (nn.Embedding): output embedding
+    """
+
+    def __init__(self, config: UnimerMBartConfig, embed_tokens: Optional[nn.Embedding] = None):
+        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_position_embeddings
+        embed_scale = math.sqrt(embed_dim) if config.scale_embedding else 1.0
+
+        self.embed_tokens = UnimerMBartScaledWordEmbedding(
+            config.vocab_size, embed_dim, self.padding_idx, embed_scale=embed_scale
+        )
+
+        if embed_tokens is not None:
+            self.embed_tokens.weight = embed_tokens.weight
+
+        self.embed_positions = UnimerMBartLearnedPositionalEmbedding(
+            config.max_position_embeddings,
+            embed_dim,
+        )
+        self.layers = nn.ModuleList([UnimerMBartEncoderLayer(config) for _ in range(config.encoder_layers)])
+        self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2"
+        self._use_sdpa = config._attn_implementation == "sdpa"
+        self.layernorm_embedding = nn.LayerNorm(embed_dim)
+        self.layer_norm = nn.LayerNorm(config.d_model)
+
+        self.gradient_checkpointing = False
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def _backward_compatibility_gradient_checkpointing(self):
+        # Override to not delete the attribute from the config
+        if self.supports_gradient_checkpointing and getattr(self.config, "gradient_checkpointing", False):
+            self.gradient_checkpointing_enable()
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, BaseModelOutput]:
+        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)
+            head_mask (`torch.Tensor` of shape `(encoder_layers, encoder_attention_heads)`, *optional*):
+                Mask to nullify selected heads of the attention modules. Mask values selected in `[0, 1]`:
+
+                - 1 indicates the head is **not masked**,
+                - 0 indicates the head is **masked**.
+
+            inputs_embeds (`torch.FloatTensor` of shape `(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.use_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 input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input = input_ids
+            input_shape = input.shape
+            input_ids = input_ids.view(-1, input_shape[-1])
+        elif inputs_embeds is not None:
+            input = inputs_embeds[:, :, -1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        embed_pos = self.embed_positions(input)
+
+        hidden_states = inputs_embeds + embed_pos.to(inputs_embeds.device)
+        hidden_states = self.layernorm_embedding(hidden_states)
+        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
+
+        # expand attention_mask
+        if attention_mask is not None:
+            # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+            if self._use_flash_attention_2:
+                attention_mask = attention_mask if 0 in attention_mask else None
+            elif self._use_sdpa and head_mask is None and not output_attentions:
+                # output_attentions=True & head_mask can not be supported when using SDPA, fall back to
+                # the manual implementation that requires a 4D causal mask in all cases.
+                # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+                attention_mask = _prepare_4d_attention_mask_for_sdpa(attention_mask, inputs_embeds.dtype)
+            else:
+                # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+                attention_mask = _prepare_4d_attention_mask(attention_mask, inputs_embeds.dtype)
+
+        encoder_states = () if output_hidden_states else None
+        all_attentions = () if output_attentions else None
+
+        # check if head_mask has a correct number of layers specified if desired
+        if head_mask is not None:
+            if head_mask.size()[0] != len(self.layers):
+                raise ValueError(
+                    f"The head_mask should be specified for {len(self.layers)} layers, but it is for"
+                    f" {head_mask.size()[0]}."
+                )
+        for idx, encoder_layer in enumerate(self.layers):
+            if output_hidden_states:
+                encoder_states = encoder_states + (hidden_states,)
+            # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description)
+            to_drop = False
+            if self.training:
+                dropout_probability = torch.rand([])
+                if dropout_probability < self.layerdrop:  # skip the layer
+                    to_drop = True
+
+            if to_drop:
+                layer_outputs = (None, None)
+            else:
+                if self.gradient_checkpointing and self.training:
+                    layer_outputs = self._gradient_checkpointing_func(
+                        encoder_layer.__call__,
+                        hidden_states,
+                        attention_mask,
+                        (head_mask[idx] if head_mask is not None else None),
+                        output_attentions,
+                    )
+                else:
+                    layer_outputs = encoder_layer(
+                        hidden_states,
+                        attention_mask,
+                        layer_head_mask=(head_mask[idx] if head_mask is not None else None),
+                        output_attentions=output_attentions,
+                    )
+
+                hidden_states = layer_outputs[0]
+
+            if output_attentions:
+                all_attentions = all_attentions + (layer_outputs[1],)
+
+        hidden_states = self.layer_norm(hidden_states)
+
+        if output_hidden_states:
+            encoder_states = encoder_states + (hidden_states,)
+
+        if not return_dict:
+            return tuple(v for v in [hidden_states, encoder_states, all_attentions] if v is not None)
+        return BaseModelOutput(
+            last_hidden_state=hidden_states, hidden_states=encoder_states, attentions=all_attentions
+        )
+
+
+class UnimerMBartDecoder(UnimerMBartPreTrainedModel):
+    """
+    Transformer decoder consisting of *config.decoder_layers* layers. Each layer is a [`MBartDecoderLayer`]
+
+    Args:
+        config: MBartConfig
+        embed_tokens (nn.Embedding): output embedding
+    """
+
+    def __init__(self, config: UnimerMBartConfig, embed_tokens: Optional[nn.Embedding] = None):
+        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_position_embeddings
+        embed_scale = math.sqrt(config.d_model) if config.scale_embedding else 1.0
+
+        self.embed_tokens = UnimerMBartScaledWordEmbedding(
+            config.vocab_size, config.d_model, self.padding_idx, embed_scale=embed_scale
+        )
+
+        if embed_tokens is not None:
+            self.embed_tokens.weight = embed_tokens.weight
+
+        self.embed_positions = UnimerMBartLearnedPositionalEmbedding(
+            config.max_position_embeddings,
+            config.d_model,
+        )
+        self.layers = nn.ModuleList([UnimerMBartDecoderLayer(config) for _ in range(config.decoder_layers)])
+        self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2"
+        self._use_sdpa = config._attn_implementation == "sdpa"
+        self.layernorm_embedding = nn.LayerNorm(config.d_model)
+        self.layer_norm = nn.LayerNorm(config.d_model)
+
+        self.gradient_checkpointing = False
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.embed_tokens
+
+    def set_input_embeddings(self, value):
+        self.embed_tokens = value
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        count_pred: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.FloatTensor] = None,
+        encoder_attention_mask: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, BaseModelOutputWithPastAndCrossAttentions]:
+        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)
+            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)
+            head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+                Mask to nullify selected heads of the attention modules. Mask values selected in `[0, 1]`:
+
+                - 1 indicates the head is **not masked**,
+                - 0 indicates the head is **masked**.
+
+            cross_attn_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+                Mask to nullify selected heads of the cross-attention modules in the decoder to avoid performing
+                cross-attention on hidden heads. Mask values selected in `[0, 1]`:
+
+                - 1 indicates the head is **not masked**,
+                - 0 indicates the head is **masked**.
+
+            past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+                Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of
+                shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of
+                shape `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
+
+                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.use_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 = input_ids
+            input_shape = input.size()
+            input_ids = input_ids.view(-1, input_shape[-1])
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+            input = inputs_embeds[:, :, -1]
+        else:
+            raise ValueError("You have to specify either decoder_input_ids or decoder_inputs_embeds")
+
+        # past_key_values_length
+        past_key_values_length = past_key_values[0][0].shape[2] if past_key_values is not None else 0
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        if self._use_flash_attention_2:
+            # 2d mask is passed through the layers
+            attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None
+        elif self._use_sdpa and not output_attentions and cross_attn_head_mask is None:
+            # output_attentions=True & cross_attn_head_mask can not be supported when using SDPA, and we fall back on
+            # the manual implementation that requires a 4D causal mask in all cases.
+            attention_mask = _prepare_4d_causal_attention_mask_for_sdpa(
+                attention_mask,
+                input_shape,
+                inputs_embeds,
+                past_key_values_length,
+            )
+        else:
+            # 4d mask is passed through the layers
+            attention_mask = _prepare_4d_causal_attention_mask(
+                attention_mask, input_shape, inputs_embeds, past_key_values_length
+            )
+
+        # expand encoder attention mask
+        if encoder_hidden_states is not None and encoder_attention_mask is not None:
+            if self._use_flash_attention_2:
+                encoder_attention_mask = encoder_attention_mask if 0 in encoder_attention_mask else None
+            elif self._use_sdpa and cross_attn_head_mask is None and not output_attentions:
+                # output_attentions=True & cross_attn_head_mask can not be supported when using SDPA, and we fall back on
+                # the manual implementation that requires a 4D causal mask in all cases.
+                # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+                encoder_attention_mask = _prepare_4d_attention_mask_for_sdpa(
+                    encoder_attention_mask,
+                    inputs_embeds.dtype,
+                    tgt_len=input_shape[-1],
+                )
+            else:
+                # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+                encoder_attention_mask = _prepare_4d_attention_mask(
+                    encoder_attention_mask, inputs_embeds.dtype, tgt_len=input_shape[-1]
+                )
+
+        # embed positions
+        positions = self.embed_positions(input, past_key_values_length)
+
+        hidden_states = inputs_embeds + positions.to(inputs_embeds.device)
+
+        # TODO: add counting context weight to hidden_states
+        if count_pred is not None:
+            count_context_weight = self.counting_context_weight(count_pred)
+            hidden_states = hidden_states + 0.5 * count_context_weight.unsqueeze(1)
+
+        hidden_states = self.layernorm_embedding(hidden_states)
+        hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training)
+
+        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
+
+        # 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 and encoder_hidden_states is not None) else None
+        next_decoder_cache = () if use_cache else None
+
+        # check if head_mask/cross_attn_head_mask has a correct number of layers specified if desired
+        for attn_mask, mask_name in zip([head_mask, cross_attn_head_mask], ["head_mask", "cross_attn_head_mask"]):
+            if attn_mask is not None:
+                if attn_mask.size()[0] != len(self.layers):
+                    raise ValueError(
+                        f"The `{mask_name}` should be specified for {len(self.layers)} layers, but it is for"
+                        f" {attn_mask.size()[0]}."
+                    )
+        for idx, decoder_layer in enumerate(self.layers):
+            # add LayerDrop (see https://arxiv.org/abs/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
+
+            past_key_value = past_key_values[idx] if past_key_values is not None else None
+
+            if self.gradient_checkpointing and self.training:
+                layer_outputs = self._gradient_checkpointing_func(
+                    decoder_layer.__call__,
+                    hidden_states,
+                    attention_mask,
+                    encoder_hidden_states,
+                    encoder_attention_mask,
+                    head_mask[idx] if head_mask is not None else None,
+                    cross_attn_head_mask[idx] if cross_attn_head_mask is not None else None,
+                    None,
+                    output_attentions,
+                    use_cache,
+                )
+            else:
+                layer_outputs = decoder_layer(
+                    hidden_states,
+                    attention_mask=attention_mask,
+                    encoder_hidden_states=encoder_hidden_states,
+                    encoder_attention_mask=encoder_attention_mask,
+                    layer_head_mask=(head_mask[idx] if head_mask is not None else None),
+                    cross_attn_layer_head_mask=(
+                        cross_attn_head_mask[idx] if cross_attn_head_mask is not None else None
+                    ),
+                    past_key_value=past_key_value,
+                    output_attentions=output_attentions,
+                    use_cache=use_cache,
+                )
+            hidden_states = layer_outputs[0]
+
+            if use_cache:
+                next_decoder_cache += (layer_outputs[3 if output_attentions else 1],)
+
+            if output_attentions:
+                all_self_attns += (layer_outputs[1],)
+
+                if encoder_hidden_states is not None:
+                    all_cross_attentions += (layer_outputs[2],)
+
+        hidden_states = self.layer_norm(hidden_states)
+
+        # add hidden states from the last decoder layer
+        if output_hidden_states:
+            all_hidden_states += (hidden_states,)
+
+        next_cache = next_decoder_cache if use_cache else None
+        if not return_dict:
+            return tuple(
+                v
+                for v in [hidden_states, next_cache, all_hidden_states, all_self_attns, all_cross_attentions]
+                if v is not None
+            )
+        return BaseModelOutputWithPastAndCrossAttentions(
+            last_hidden_state=hidden_states,
+            past_key_values=next_cache,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attns,
+            cross_attentions=all_cross_attentions,
+        )
+
+
+@add_start_docstrings(
+    "The bare MBART Model outputting raw hidden-states without any specific head on top.",
+    MBART_START_DOCSTRING,
+)
+class UnimerMBartModel(UnimerMBartPreTrainedModel):
+    _tied_weights_keys = ["encoder.embed_tokens.weight", "decoder.embed_tokens.weight"]
+
+    def __init__(self, config: UnimerMBartConfig):
+        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 = UnimerMBartEncoder(config, self.shared)
+        self.decoder = UnimerMBartDecoder(config, self.shared)
+
+        # 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
+
+    def get_encoder(self):
+        return self.encoder
+
+    def get_decoder(self):
+        return self.decoder
+
+    def _tie_weights(self):
+        if self.config.tie_word_embeddings:
+            self._tie_or_clone_weights(self.encoder.embed_tokens, self.get_input_embeddings())
+            self._tie_or_clone_weights(self.decoder.embed_tokens, self.get_input_embeddings())
+
+    @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=Seq2SeqModelOutput,
+        config_class=_CONFIG_FOR_DOC,
+        expected_output=_EXPECTED_OUTPUT_SHAPE,
+    )
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        decoder_input_ids: Optional[torch.LongTensor] = None,
+        decoder_attention_mask: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        decoder_head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        encoder_outputs: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Seq2SeqModelOutput, Tuple[torch.FloatTensor]]:
+        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.use_return_dict
+
+        # different to other models, MBart automatically creates decoder_input_ids from
+        # input_ids if no decoder_input_ids are provided
+        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)
+
+        if encoder_outputs is None:
+            encoder_outputs = self.encoder(
+                input_ids=input_ids,
+                attention_mask=attention_mask,
+                head_mask=head_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 BaseModelOutput when return_dict=True
+        elif return_dict and not isinstance(encoder_outputs, BaseModelOutput):
+            encoder_outputs = BaseModelOutput(
+                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,
+            )
+
+        # decoder outputs consists of (dec_features, past_key_value, 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,
+            head_mask=decoder_head_mask,
+            cross_attn_head_mask=cross_attn_head_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 Seq2SeqModelOutput(
+            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,
+        )
+
+
+@add_start_docstrings(
+    "The MBART Model with a language modeling head. Can be used for summarization, after fine-tuning the pretrained models.",
+    MBART_START_DOCSTRING,
+)
+class UnimerMBartForConditionalGeneration(UnimerMBartPreTrainedModel, GenerationMixin):
+    base_model_prefix = "model"
+    _keys_to_ignore_on_load_missing = ["final_logits_bias"]
+    _tied_weights_keys = ["model.encoder.embed_tokens.weight", "model.decoder.embed_tokens.weight", "lm_head.weight"]
+
+    def __init__(self, config: UnimerMBartConfig):
+        super().__init__(config)
+        self.model = UnimerMBartModel(config)
+        self.register_buffer("final_logits_bias", torch.zeros((1, self.model.shared.num_embeddings)))
+        self.lm_head = nn.Linear(config.d_model, self.model.shared.num_embeddings, bias=False)
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_encoder(self):
+        return self.model.get_encoder()
+
+    def get_decoder(self):
+        return self.model.get_decoder()
+
+    def resize_token_embeddings(self, new_num_tokens: int, pad_to_multiple_of: Optional[int] = None) -> nn.Embedding:
+        new_embeddings = super().resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
+        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)
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
+    @replace_return_docstrings(output_type=Seq2SeqLMOutput, config_class=_CONFIG_FOR_DOC)
+    @add_end_docstrings(MBART_GENERATION_EXAMPLE)
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        decoder_input_ids: Optional[torch.LongTensor] = None,
+        decoder_attention_mask: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        decoder_head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        encoder_outputs: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Seq2SeqLMOutput, Tuple[torch.FloatTensor]]:
+        r"""
+        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]`.
+
+        Returns:
+
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_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)
+
+        outputs = self.model(
+            input_ids,
+            attention_mask=attention_mask,
+            decoder_input_ids=decoder_input_ids,
+            encoder_outputs=encoder_outputs,
+            decoder_attention_mask=decoder_attention_mask,
+            head_mask=head_mask,
+            decoder_head_mask=decoder_head_mask,
+            cross_attn_head_mask=cross_attn_head_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 Seq2SeqLMOutput(
+            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,
+        )
+
+    def prepare_inputs_for_generation(
+        self,
+        decoder_input_ids,
+        past_key_values=None,
+        attention_mask=None,
+        head_mask=None,
+        decoder_head_mask=None,
+        cross_attn_head_mask=None,
+        use_cache=None,
+        encoder_outputs=None,
+        **kwargs,
+    ):
+        # cut decoder_input_ids if past is used
+        if past_key_values is not None:
+            past_length = past_key_values[0][0].shape[2]
+
+            # Some generation methods already pass only the last input ID
+            if decoder_input_ids.shape[1] > past_length:
+                remove_prefix_length = past_length
+            else:
+                # Default to old behavior: keep only final ID
+                remove_prefix_length = decoder_input_ids.shape[1] - 1
+
+            decoder_input_ids = decoder_input_ids[:, remove_prefix_length:]
+
+        return {
+            "input_ids": None,  # encoder_outputs is defined. input_ids not needed
+            "encoder_outputs": encoder_outputs,
+            "past_key_values": past_key_values,
+            "decoder_input_ids": decoder_input_ids,
+            "attention_mask": attention_mask,
+            "head_mask": head_mask,
+            "decoder_head_mask": decoder_head_mask,
+            "cross_attn_head_mask": cross_attn_head_mask,
+            "use_cache": use_cache,  # change this to avoid caching (presumably for debugging)
+        }
+
+    def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor):
+        return shift_tokens_right(labels, self.config.pad_token_id)
+
+    @staticmethod
+    def _reorder_cache(past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            # cached cross_attention states don't have to be reordered -> they are always the same
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past[:2])
+                + layer_past[2:],
+            )
+        return reordered_past
+
+
+@add_start_docstrings(
+    """
+    MBart model with a sequence classification/head on top (a linear layer on top of the pooled output) e.g. for GLUE
+    tasks.
+    """,
+    MBART_START_DOCSTRING,
+)
+class UnimerMBartForSequenceClassification(UnimerMBartPreTrainedModel):
+    _tied_weights_keys = ["model.encoder.embed_tokens.weight", "model.decoder.embed_tokens.weight"]
+
+    def __init__(self, config: UnimerMBartConfig, **kwargs):
+        super().__init__(config, **kwargs)
+        self.model = UnimerMBartModel(config)
+        self.classification_head = UnimerMBartClassificationHead(
+            config.d_model,
+            config.d_model,
+            config.num_labels,
+            config.classifier_dropout,
+        )
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=Seq2SeqSequenceClassifierOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    # Copied from transformers.models.bart.modeling_bart.BartForSequenceClassification.forward
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        decoder_input_ids: Optional[torch.LongTensor] = None,
+        decoder_attention_mask: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        decoder_head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        encoder_outputs: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, Seq2SeqSequenceClassifierOutput]:
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
+            config.num_labels - 1]`. If `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        if labels is not None:
+            use_cache = False
+
+        if input_ids is None and inputs_embeds is not None:
+            raise NotImplementedError(
+                f"Passing input embeddings is currently not supported for {self.__class__.__name__}"
+            )
+
+        outputs = self.model(
+            input_ids,
+            attention_mask=attention_mask,
+            decoder_input_ids=decoder_input_ids,
+            decoder_attention_mask=decoder_attention_mask,
+            head_mask=head_mask,
+            decoder_head_mask=decoder_head_mask,
+            cross_attn_head_mask=cross_attn_head_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,
+        )
+        hidden_states = outputs[0]  # last hidden state
+
+        eos_mask = input_ids.eq(self.config.eos_token_id).to(hidden_states.device)
+
+        if len(torch.unique_consecutive(eos_mask.sum(1))) > 1:
+            raise ValueError("All examples must have the same number of <eos> tokens.")
+        sentence_representation = hidden_states[eos_mask, :].view(hidden_states.size(0), -1, hidden_states.size(-1))[
+            :, -1, :
+        ]
+        logits = self.classification_head(sentence_representation)
+
+        loss = None
+        if labels is not None:
+            labels = labels.to(logits.device)
+            if self.config.problem_type is None:
+                if self.config.num_labels == 1:
+                    self.config.problem_type = "regression"
+                elif self.config.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
+                    self.config.problem_type = "single_label_classification"
+                else:
+                    self.config.problem_type = "multi_label_classification"
+
+            if self.config.problem_type == "regression":
+                loss_fct = MSELoss()
+                if self.config.num_labels == 1:
+                    loss = loss_fct(logits.squeeze(), labels.squeeze())
+                else:
+                    loss = loss_fct(logits, labels)
+            elif self.config.problem_type == "single_label_classification":
+                loss_fct = CrossEntropyLoss()
+                loss = loss_fct(logits.view(-1, self.config.num_labels), labels.view(-1))
+            elif self.config.problem_type == "multi_label_classification":
+                loss_fct = BCEWithLogitsLoss()
+                loss = loss_fct(logits, labels)
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return ((loss,) + output) if loss is not None else output
+
+        return Seq2SeqSequenceClassifierOutput(
+            loss=loss,
+            logits=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,
+        )
+
+
+@add_start_docstrings(
+    """
+    MBART Model with a span classification head on top for extractive question-answering tasks like SQuAD (a linear
+    layer on top of the hidden-states output to compute `span start logits` and `span end logits`).
+    """,
+    MBART_START_DOCSTRING,
+)
+class UnimerMBartForQuestionAnswering(UnimerMBartPreTrainedModel):
+    _tied_weights_keys = ["model.encoder.embed_tokens.weight", "model.decoder.embed_tokens.weight"]
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        config.num_labels = 2
+        self.num_labels = config.num_labels
+
+        self.model = UnimerMBartModel(config)
+        self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @add_start_docstrings_to_model_forward(MBART_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=Seq2SeqQuestionAnsweringModelOutput,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    # Copied from transformers.models.bart.modeling_bart.BartForQuestionAnswering.forward
+    def forward(
+        self,
+        input_ids: torch.Tensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        decoder_input_ids: Optional[torch.LongTensor] = None,
+        decoder_attention_mask: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        decoder_head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        encoder_outputs: Optional[List[torch.FloatTensor]] = None,
+        start_positions: Optional[torch.LongTensor] = None,
+        end_positions: Optional[torch.LongTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        decoder_inputs_embeds: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, Seq2SeqQuestionAnsweringModelOutput]:
+        r"""
+        start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for position (index) of the start of the labelled span for computing the token classification loss.
+            Positions are clamped to the length of the sequence (*sequence_length*). Position outside of the sequence
+            are not taken into account for computing the loss.
+        end_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+            Labels for position (index) of the end of the labelled span for computing the token classification loss.
+            Positions are clamped to the length of the sequence (*sequence_length*). Position outside of the sequence
+            are not taken into account for computing the loss.
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        if start_positions is not None and end_positions is not None:
+            use_cache = False
+
+        outputs = self.model(
+            input_ids,
+            attention_mask=attention_mask,
+            decoder_input_ids=decoder_input_ids,
+            decoder_attention_mask=decoder_attention_mask,
+            head_mask=head_mask,
+            decoder_head_mask=decoder_head_mask,
+            cross_attn_head_mask=cross_attn_head_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 Seq2SeqQuestionAnsweringModelOutput(
+            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,
+        )
+
+
+# Copied from transformers.models.bart.modeling_bart.BartDecoderWrapper with Bart->MBart
+class UnimerMBartDecoderWrapper(UnimerMBartPreTrainedModel):
+    """
+    This wrapper class is a helper class to correctly load pretrained checkpoints when the causal language model is
+    used in combination with the [`EncoderDecoderModel`] framework.
+    """
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.decoder = UnimerMBartDecoder(config)
+
+    def forward(self, *args, **kwargs):
+        return self.decoder(*args, **kwargs)
+
+
+# Copied from transformers.models.bart.modeling_bart.BartForCausalLM with Bart->MBart, facebook/bart-base->facebook/mbart-large-cc25
+class UnimerMBartForCausalLM(UnimerMBartPreTrainedModel, GenerationMixin):
+    _tied_weights_keys = ["lm_head.weight"]
+
+    def __init__(self, config):
+        config = copy.deepcopy(config)
+        config.is_decoder = True
+        config.is_encoder_decoder = False
+        super().__init__(config)
+        self.model = UnimerMBartDecoderWrapper(config)
+
+        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.model.decoder.embed_tokens
+
+    def set_input_embeddings(self, value):
+        self.model.decoder.embed_tokens = value
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def set_decoder(self, decoder):
+        self.model.decoder = decoder
+
+    def get_decoder(self):
+        return self.model.decoder
+
+    @replace_return_docstrings(output_type=CausalLMOutputWithCrossAttentionsAndCounting, config_class=_CONFIG_FOR_DOC)
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        encoder_hidden_states: Optional[torch.FloatTensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        cross_attn_head_mask: Optional[torch.Tensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        count_gt: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, CausalLMOutputWithCrossAttentions]:
+        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)
+            encoder_hidden_states  (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+                Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention
+                if the model is configured as a decoder.
+            encoder_attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used
+                in the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`:
+            head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+                Mask to nullify selected heads of the attention modules. Mask values selected in `[0, 1]`:
+
+                - 1 indicates the head is **not masked**,
+                - 0 indicates the head is **masked**.
+
+            cross_attn_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*):
+                Mask to nullify selected heads of the cross-attention modules. Mask values selected in `[0, 1]`:
+
+                - 1 indicates the head is **not masked**,
+                - 0 indicates the head is **masked**.
+
+            past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+                Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of
+                shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of
+                shape `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`. The two additional
+                tensors are only required when the model is used as a decoder in a Sequence to Sequence model.
+
+                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)`.
+            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]`.
+            use_cache (`bool`, *optional*):
+                If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding
+                (see `past_key_values`).
+
+                - 1 for tokens that are **not masked**,
+                - 0 for tokens that are **masked**.
+            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.
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import AutoTokenizer, MBartForCausalLM
+
+        >>> tokenizer = AutoTokenizer.from_pretrained("facebook/mbart-large-cc25")
+        >>> model = MBartForCausalLM.from_pretrained("facebook/mbart-large-cc25", add_cross_attention=False)
+        >>> assert model.config.is_decoder, f"{model.__class__} has to be configured as a decoder."
+        >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
+        >>> outputs = model(**inputs)
+
+        >>> logits = outputs.logits
+        >>> expected_shape = [1, inputs.input_ids.shape[-1], model.config.vocab_size]
+        >>> list(logits.shape) == expected_shape
+        True
+        ```"""
+
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        count_pred = None
+
+        # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+        outputs = self.model.decoder(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            count_pred=count_pred,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            head_mask=head_mask,
+            cross_attn_head_mask=cross_attn_head_mask,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        logits = self.lm_head(outputs[0])
+
+        loss = None
+        if labels is not None:
+            labels = labels.to(logits.device)
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(logits.view(-1, self.config.vocab_size), labels.view(-1))
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return CausalLMOutputWithCrossAttentionsAndCounting(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            cross_attentions=outputs.cross_attentions,
+            counting=count_pred,
+        )
+
+    def prepare_inputs_for_generation(
+        self, input_ids, past_key_values=None, attention_mask=None, use_cache=None, **kwargs
+    ):
+        # if model is used as a decoder in encoder-decoder model, the decoder attention mask is created on the fly
+        if attention_mask is None:
+            attention_mask = input_ids.new_ones(input_ids.shape)
+
+        if past_key_values:
+            past_length = past_key_values[0][0].shape[2]
+
+            # Some generation methods already pass only the last input ID
+            if input_ids.shape[1] > past_length:
+                remove_prefix_length = past_length
+            else:
+                # Default to old behavior: keep only final ID
+                remove_prefix_length = input_ids.shape[1] - 1
+
+            input_ids = input_ids[:, remove_prefix_length:]
+        # first step, decoder_cached_states are empty
+        return {
+            "input_ids": input_ids,  # encoder_outputs is defined. input_ids not needed
+            "attention_mask": attention_mask,
+            "past_key_values": past_key_values,
+            "use_cache": use_cache,
+        }
+
+    @staticmethod
+    def _reorder_cache(past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past),
+            )
+        return reordered_past

mineru/model/mfr/unimernet/unimernet_hf/unimer_swin/__init__.py

+from .configuration_unimer_swin import UnimerSwinConfig
+from .modeling_unimer_swin import UnimerSwinModel
+from .image_processing_unimer_swin import UnimerSwinImageProcessor
+
+__all__ = [
+    "UnimerSwinConfig",
+    "UnimerSwinModel",
+    "UnimerSwinImageProcessor",
+]

mineru/model/mfr/unimernet/unimernet_hf/unimer_swin/configuration_unimer_swin.py

+# coding=utf-8
+# Copyright 2022 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.
+"""Donut Swin Transformer model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+
+class UnimerSwinConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`UnimerSwinModel`]. It is used to instantiate a
+    Donut model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of the Donut
+    [naver-clova-ix/donut-base](https://huggingface.co/naver-clova-ix/donut-base) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        image_size (`int`, *optional*, defaults to 224):
+            The size (resolution) of each image.
+        patch_size (`int`, *optional*, defaults to 4):
+            The size (resolution) of each patch.
+        num_channels (`int`, *optional*, defaults to 3):
+            The number of input channels.
+        embed_dim (`int`, *optional*, defaults to 96):
+            Dimensionality of patch embedding.
+        depths (`list(int)`, *optional*, defaults to `[2, 2, 6, 2]`):
+            Depth of each layer in the Transformer encoder.
+        num_heads (`list(int)`, *optional*, defaults to `[3, 6, 12, 24]`):
+            Number of attention heads in each layer of the Transformer encoder.
+        window_size (`int`, *optional*, defaults to 7):
+            Size of windows.
+        mlp_ratio (`float`, *optional*, defaults to 4.0):
+            Ratio of MLP hidden dimensionality to embedding dimensionality.
+        qkv_bias (`bool`, *optional*, defaults to `True`):
+            Whether or not a learnable bias should be added to the queries, keys and values.
+        hidden_dropout_prob (`float`, *optional*, defaults to 0.0):
+            The dropout probability for all fully connected layers in the embeddings and encoder.
+        attention_probs_dropout_prob (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+        drop_path_rate (`float`, *optional*, defaults to 0.1):
+            Stochastic depth rate.
+        hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder. If string, `"gelu"`, `"relu"`,
+            `"selu"` and `"gelu_new"` are supported.
+        use_absolute_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether or not to add absolute position embeddings to the patch embeddings.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-05):
+            The epsilon used by the layer normalization layers.
+
+    Example:
+
+    ```python
+    >>> from transformers import UnimerSwinConfig, UnimerSwinModel
+
+    >>> # Initializing a Donut naver-clova-ix/donut-base style configuration
+    >>> configuration = UnimerSwinConfig()
+
+    >>> # Randomly initializing a model from the naver-clova-ix/donut-base style configuration
+    >>> model = UnimerSwinModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "unimer-swin"
+
+    attribute_map = {
+        "num_attention_heads": "num_heads",
+        "num_hidden_layers": "num_layers",
+    }
+
+    def __init__(
+        self,
+        image_size=224,
+        patch_size=4,
+        num_channels=3,
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=7,
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        hidden_dropout_prob=0.0,
+        attention_probs_dropout_prob=0.0,
+        drop_path_rate=0.1,
+        hidden_act="gelu",
+        use_absolute_embeddings=False,
+        initializer_range=0.02,
+        layer_norm_eps=1e-5,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self.image_size = image_size
+        self.patch_size = patch_size
+        self.num_channels = num_channels
+        self.embed_dim = embed_dim
+        self.depths = depths
+        self.num_layers = len(depths)
+        self.num_heads = num_heads
+        self.window_size = window_size
+        self.mlp_ratio = mlp_ratio
+        self.qkv_bias = qkv_bias
+        self.hidden_dropout_prob = hidden_dropout_prob
+        self.attention_probs_dropout_prob = attention_probs_dropout_prob
+        self.drop_path_rate = drop_path_rate
+        self.hidden_act = hidden_act
+        self.use_absolute_embeddings = use_absolute_embeddings
+        self.layer_norm_eps = layer_norm_eps
+        self.initializer_range = initializer_range
+        # we set the hidden_size attribute in order to make Swin work with VisionEncoderDecoderModel
+        # this indicates the channel dimension after the last stage of the model
+        self.hidden_size = int(embed_dim * 2 ** (len(depths) - 1))

mineru/model/mfr/unimernet/unimernet_hf/unimer_swin/image_processing_unimer_swin.py

+from transformers.image_processing_utils import BaseImageProcessor
+import numpy as np
+import cv2
+import albumentations as alb
+from albumentations.pytorch import ToTensorV2
+
+
+# TODO: dereference cv2 if possible
+class UnimerSwinImageProcessor(BaseImageProcessor):
+    def __init__(
+            self,
+            image_size = (192, 672),
+        ):
+        self.input_size = [int(_) for _ in image_size]
+        assert len(self.input_size) == 2
+    
+        self.transform = alb.Compose(
+            [
+                alb.ToGray(),
+                alb.Normalize((0.7931, 0.7931, 0.7931), (0.1738, 0.1738, 0.1738)),
+                # alb.Sharpen()
+                ToTensorV2(),
+            ]
+        )
+
+    def __call__(self, item):
+        image = self.prepare_input(item)
+        return self.transform(image=image)['image'][:1]
+
+    @staticmethod
+    def crop_margin_numpy(img: np.ndarray) -> np.ndarray:
+        """Crop margins of image using NumPy operations"""
+        # Convert to grayscale if it's a color image
+        if len(img.shape) == 3 and img.shape[2] == 3:
+            gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
+        else:
+            gray = img.copy()
+
+        # Normalize and threshold
+        if gray.max() == gray.min():
+            return img
+
+        normalized = (((gray - gray.min()) / (gray.max() - gray.min())) * 255).astype(np.uint8)
+        binary = 255 * (normalized < 200).astype(np.uint8)
+
+        # Find bounding box
+        coords = cv2.findNonZero(binary)  # Find all non-zero points (text)
+        x, y, w, h = cv2.boundingRect(coords)  # Find minimum spanning bounding box
+
+        # Return cropped image
+        return img[y:y + h, x:x + w]
+
+    def prepare_input(self, img, random_padding: bool = False):
+        """
+        Convert PIL Image or numpy array to properly sized and padded image after:
+            - crop margins
+            - resize while maintaining aspect ratio
+            - pad to target size
+        """
+        if img is None:
+            return None
+
+        # try:
+        #     img = self.crop_margin_numpy(img)
+        # except Exception:
+        #     # might throw an error for broken files
+        #     return None
+
+        if img.shape[0] == 0 or img.shape[1] == 0:
+            return None
+
+        # Get current dimensions
+        h, w = img.shape[:2]
+        target_h, target_w = self.input_size
+
+        # Calculate scale to preserve aspect ratio (equivalent to resize + thumbnail)
+        scale = min(target_h / h, target_w / w)
+
+        # Calculate new dimensions
+        new_h, new_w = int(h * scale), int(w * scale)
+
+        # Resize the image while preserving aspect ratio
+        resized_img = cv2.resize(img, (new_w, new_h))
+
+        # Calculate padding values using the existing method
+        delta_width = target_w - new_w
+        delta_height = target_h - new_h
+
+        pad_width, pad_height = self._get_padding_values(new_w, new_h, random_padding)
+
+        # Apply padding (convert PIL padding format to OpenCV format)
+        padding_color = [0, 0, 0] if len(img.shape) == 3 else [0]
+
+        padded_img = cv2.copyMakeBorder(
+            resized_img,
+            pad_height,  # top
+            delta_height - pad_height,  # bottom
+            pad_width,  # left
+            delta_width - pad_width,  # right
+            cv2.BORDER_CONSTANT,
+            value=padding_color
+        )
+
+        return padded_img
+
+    def _calculate_padding(self, new_w, new_h, random_padding):
+        """Calculate padding values for PIL images"""
+        delta_width = self.input_size[1] - new_w
+        delta_height = self.input_size[0] - new_h
+
+        pad_width, pad_height = self._get_padding_values(new_w, new_h, random_padding)
+
+        return (
+            pad_width,
+            pad_height,
+            delta_width - pad_width,
+            delta_height - pad_height,
+        )
+
+    def _get_padding_values(self, new_w, new_h, random_padding):
+        """Get padding values based on image dimensions and padding strategy"""
+        delta_width = self.input_size[1] - new_w
+        delta_height = self.input_size[0] - new_h
+
+        if random_padding:
+            pad_width = np.random.randint(low=0, high=delta_width + 1)
+            pad_height = np.random.randint(low=0, high=delta_height + 1)
+        else:
+            pad_width = delta_width // 2
+            pad_height = delta_height // 2
+
+        return pad_width, pad_height

mineru/model/mfr/unimernet/unimernet_hf/unimer_swin/modeling_unimer_swin.py

+# coding=utf-8
+# Copyright 2022 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 UnimerSwin Transformer model.
+
+This implementation is identical to a regular Swin Transformer, without final layer norm on top of the final hidden
+states."""
+
+import collections.abc
+import math
+from dataclasses import dataclass
+from typing import Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+from torch import nn
+
+from transformers.activations import ACT2FN
+from transformers.modeling_utils import PreTrainedModel
+from transformers.pytorch_utils import find_pruneable_heads_and_indices, meshgrid, prune_linear_layer
+from transformers.utils import (
+    ModelOutput,
+    add_code_sample_docstrings,
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    logging,
+    torch_int,
+)
+from .configuration_unimer_swin import UnimerSwinConfig
+
+
+logger = logging.get_logger(__name__)
+
+# General docstring
+_CONFIG_FOR_DOC = "UnimerSwinConfig"
+
+# Base docstring
+_CHECKPOINT_FOR_DOC = "https://huggingface.co/naver-clova-ix/donut-base"
+_EXPECTED_OUTPUT_SHAPE = [1, 49, 768]
+
+
+@dataclass
+# Copied from transformers.models.swin.modeling_swin.SwinEncoderOutput with Swin->UnimerSwin
+class UnimerSwinEncoderOutput(ModelOutput):
+    """
+    UnimerSwin encoder's outputs, with potential hidden states and attentions.
+
+    Args:
+        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 model.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each stage) of
+            shape `(batch_size, sequence_length, hidden_size)`.
+
+            Hidden-states of the model at the output of each layer plus the initial embedding outputs.
+        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 stage) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        reshaped_hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each stage) of
+            shape `(batch_size, hidden_size, height, width)`.
+
+            Hidden-states of the model at the output of each layer plus the initial embedding outputs reshaped to
+            include the spatial dimensions.
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    reshaped_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+
+
+@dataclass
+# Copied from transformers.models.swin.modeling_swin.SwinModelOutput with Swin->UnimerSwin
+class UnimerSwinModelOutput(ModelOutput):
+    """
+    UnimerSwin model's outputs that also contains a pooling of the last hidden states.
+
+    Args:
+        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 model.
+        pooler_output (`torch.FloatTensor` of shape `(batch_size, hidden_size)`, *optional*, returned when `add_pooling_layer=True` is passed):
+            Average pooling of the last layer hidden-state.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each stage) of
+            shape `(batch_size, sequence_length, hidden_size)`.
+
+            Hidden-states of the model at the output of each layer plus the initial embedding outputs.
+        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 stage) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        reshaped_hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each stage) of
+            shape `(batch_size, hidden_size, height, width)`.
+
+            Hidden-states of the model at the output of each layer plus the initial embedding outputs reshaped to
+            include the spatial dimensions.
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    pooler_output: Optional[torch.FloatTensor] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    reshaped_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+
+
+# Copied from transformers.models.swin.modeling_swin.window_partition
+def window_partition(input_feature, window_size):
+    """
+    Partitions the given input into windows.
+    """
+    batch_size, height, width, num_channels = input_feature.shape
+    input_feature = input_feature.view(
+        batch_size, height // window_size, window_size, width // window_size, window_size, num_channels
+    )
+    windows = input_feature.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, num_channels)
+    return windows
+
+
+# Copied from transformers.models.swin.modeling_swin.window_reverse
+def window_reverse(windows, window_size, height, width):
+    """
+    Merges windows to produce higher resolution features.
+    """
+    num_channels = windows.shape[-1]
+    windows = windows.view(-1, height // window_size, width // window_size, window_size, window_size, num_channels)
+    windows = windows.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, height, width, num_channels)
+    return windows
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinEmbeddings with Swin->UnimerSwin
+class UnimerSwinEmbeddings(nn.Module):
+    """
+    Construct the patch and position embeddings. Optionally, also the mask token.
+    """
+
+    def __init__(self, config, use_mask_token=False):
+        super().__init__()
+
+        self.patch_embeddings = UnimerSwinPatchEmbeddings(config)
+        num_patches = self.patch_embeddings.num_patches
+        self.patch_grid = self.patch_embeddings.grid_size
+        self.mask_token = nn.Parameter(torch.zeros(1, 1, config.embed_dim)) if use_mask_token else None
+
+        if config.use_absolute_embeddings:
+            self.position_embeddings = nn.Parameter(torch.zeros(1, num_patches + 1, config.embed_dim))
+        else:
+            self.position_embeddings = None
+
+        ### code added. ###
+        if config.use_2d_embeddings:
+            self.row_embeddings = nn.Parameter(torch.zeros(1, self.patch_grid[0] + 1, config.embed_dim))
+            self.column_embeddings = nn.Parameter(torch.zeros(1, self.patch_grid[1] + 1, config.embed_dim))
+        else:
+            self.row_embeddings = None
+            self.column_embeddings = None
+        ######
+
+        self.norm = nn.LayerNorm(config.embed_dim)
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+
+    def interpolate_pos_encoding(self, embeddings: torch.Tensor, height: int, width: int) -> torch.Tensor:
+        """
+        This method allows to interpolate the pre-trained position encodings, to be able to use the model on higher
+        resolution images.
+
+        Source:
+        https://github.com/facebookresearch/dino/blob/de9ee3df6cf39fac952ab558447af1fa1365362a/vision_transformer.py#L174
+        """
+
+        num_patches = embeddings.shape[1] - 1
+        num_positions = self.position_embeddings.shape[1] - 1
+        if num_patches == num_positions and height == width:
+            return self.position_embeddings
+        class_pos_embed = self.position_embeddings[:, 0]
+        patch_pos_embed = self.position_embeddings[:, 1:]
+        dim = embeddings.shape[-1]
+        h0 = height // self.config.patch_size
+        w0 = width // self.config.patch_size
+        # we add a small number to avoid floating point error in the interpolation
+        # see discussion at https://github.com/facebookresearch/dino/issues/8
+        h0, w0 = h0 + 0.1, w0 + 0.1
+        patch_pos_embed = patch_pos_embed.reshape(1, int(math.sqrt(num_positions)), int(math.sqrt(num_positions)), dim)
+        patch_pos_embed = patch_pos_embed.permute(0, 3, 1, 2)
+        patch_pos_embed = nn.functional.interpolate(
+            patch_pos_embed,
+            scale_factor=(h0 / math.sqrt(num_positions), w0 / math.sqrt(num_positions)),
+            mode="bicubic",
+            align_corners=False,
+        )
+        patch_pos_embed = patch_pos_embed.permute(0, 2, 3, 1).view(1, -1, dim)
+        return torch.cat((class_pos_embed.unsqueeze(0), patch_pos_embed), dim=1)
+
+    def forward(
+        self,
+        pixel_values: Optional[torch.FloatTensor],
+        bool_masked_pos: Optional[torch.BoolTensor] = None,
+        interpolate_pos_encoding: bool = False,
+    ) -> Tuple[torch.Tensor]:
+        _, num_channels, height, width = pixel_values.shape
+        embeddings, output_dimensions = self.patch_embeddings(pixel_values)
+        embeddings = self.norm(embeddings)
+        batch_size, seq_len, _ = embeddings.size()
+
+        if bool_masked_pos is not None:
+            mask_tokens = self.mask_token.expand(batch_size, seq_len, -1)
+            # replace the masked visual tokens by mask_tokens
+            mask = bool_masked_pos.unsqueeze(-1).type_as(mask_tokens)
+            embeddings = embeddings * (1.0 - mask) + mask_tokens * mask
+
+        if self.position_embeddings is not None:
+            # if interpolate_pos_encoding:
+            #     embeddings = embeddings + self.interpolate_pos_encoding(embeddings, height, width)
+            # else:
+            #     embeddings = embeddings + self.position_embeddings
+            embeddings = embeddings + self.position_embeddings[:, :seq_len, :] # code edited.
+
+        ### code added. ###
+        if self.row_embeddings is not None and self.column_embeddings is not None:
+            # Repeat the x position embeddings across the y axis like 0, 1, 2, 3, 0, 1, 2, 3, ...
+            row_embeddings = self.row_embeddings[:, :output_dimensions[0], :].repeat_interleave(output_dimensions[1], dim=1)
+            column_embeddings = self.column_embeddings[:, :output_dimensions[1], :].repeat(1, output_dimensions[0], 1)
+            embeddings = embeddings + row_embeddings + column_embeddings
+        ######
+
+        embeddings = self.dropout(embeddings)
+
+        return embeddings, output_dimensions
+
+class StemLayer(nn.Module):
+    r""" Stem layer of InternImage
+    Args:
+        in_chans (int): number of input channels
+        out_chans (int): number of output channels
+        act_layer (str): activation layer
+        norm_layer (str): normalization layer
+    """
+
+    def __init__(self, in_chans=3, out_chans=96, act_layer=nn.GELU, norm_layer='BN'):
+        super().__init__()
+        self.conv1 = nn.Conv2d(in_chans, out_chans // 2, kernel_size=3, stride=2, padding=1)
+        self.norm1 = self.build_norm_layer(out_chans // 2, norm_layer)
+        self.act = act_layer()
+        self.conv2 = nn.Conv2d(out_chans // 2, out_chans, kernel_size=3, stride=2, padding=1)
+
+    def build_norm_layer(self, dim, norm_layer):
+        layers = []
+        if norm_layer == 'BN':
+            layers.append(nn.BatchNorm2d(dim))
+        else:
+            raise NotImplementedError(f'build_norm_layer does not support {norm_layer}')
+        return nn.Sequential(*layers)
+
+    def forward(self, x):
+        x = self.conv1(x)
+        x = self.norm1(x)
+        x = self.act(x)
+        x = self.conv2(x)
+        return x
+
+# Copied from transformers.models.swin.modeling_swin.SwinPatchEmbeddings with Swin->UnimerSwin
+class UnimerSwinPatchEmbeddings(nn.Module):
+    """
+    This class turns `pixel_values` of shape `(batch_size, num_channels, height, width)` into the initial
+    `hidden_states` (patch embeddings) of shape `(batch_size, seq_length, hidden_size)` to be consumed by a
+    Transformer.
+    """
+
+    def __init__(self, config):
+        super().__init__()
+        image_size, patch_size = config.image_size, config.patch_size
+        num_channels, hidden_size = config.num_channels, config.embed_dim
+        image_size = image_size if isinstance(image_size, collections.abc.Iterable) else (image_size, image_size)
+        patch_size = patch_size if isinstance(patch_size, collections.abc.Iterable) else (patch_size, patch_size)
+        num_patches = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0])
+        self.image_size = image_size
+        self.patch_size = patch_size
+        self.num_channels = num_channels
+        self.num_patches = num_patches
+        self.grid_size = (image_size[0] // patch_size[0], image_size[1] // patch_size[1])
+
+        ### code edited. ###
+        # self.projection = nn.Conv2d(num_channels, hidden_size, kernel_size=patch_size, stride=patch_size)
+        self.projection = StemLayer(in_chans=num_channels, out_chans=hidden_size)
+        ###
+
+    def maybe_pad(self, pixel_values, height, width):
+        if width % self.patch_size[1] != 0:
+            pad_values = (0, self.patch_size[1] - width % self.patch_size[1])
+            pixel_values = nn.functional.pad(pixel_values, pad_values)
+        if height % self.patch_size[0] != 0:
+            pad_values = (0, 0, 0, self.patch_size[0] - height % self.patch_size[0])
+            pixel_values = nn.functional.pad(pixel_values, pad_values)
+        return pixel_values
+
+    def forward(self, pixel_values: Optional[torch.FloatTensor]) -> Tuple[torch.Tensor, Tuple[int]]:
+        _, num_channels, height, width = pixel_values.shape
+        # pad the input to be divisible by self.patch_size, if needed
+        pixel_values = self.maybe_pad(pixel_values, height, width)
+        embeddings = self.projection(pixel_values)
+        _, _, height, width = embeddings.shape
+        output_dimensions = (height, width)
+        embeddings = embeddings.flatten(2).transpose(1, 2)
+
+        return embeddings, output_dimensions
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinPatchMerging
+class UnimerSwinPatchMerging(nn.Module):
+    """
+    Patch Merging Layer.
+
+    Args:
+        input_resolution (`Tuple[int]`):
+            Resolution of input feature.
+        dim (`int`):
+            Number of input channels.
+        norm_layer (`nn.Module`, *optional*, defaults to `nn.LayerNorm`):
+            Normalization layer class.
+    """
+
+    def __init__(self, input_resolution: Tuple[int], dim: int, norm_layer: nn.Module = nn.LayerNorm) -> None:
+        super().__init__()
+        self.input_resolution = input_resolution
+        self.dim = dim
+        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
+        self.norm = norm_layer(4 * dim)
+
+    def maybe_pad(self, input_feature, height, width):
+        should_pad = (height % 2 == 1) or (width % 2 == 1)
+        if should_pad:
+            pad_values = (0, 0, 0, width % 2, 0, height % 2)
+            input_feature = nn.functional.pad(input_feature, pad_values)
+
+        return input_feature
+
+    def forward(self, input_feature: torch.Tensor, input_dimensions: Tuple[int, int]) -> torch.Tensor:
+        height, width = input_dimensions
+        # `dim` is height * width
+        batch_size, dim, num_channels = input_feature.shape
+
+        input_feature = input_feature.view(batch_size, height, width, num_channels)
+        # pad input to be disible by width and height, if needed
+        input_feature = self.maybe_pad(input_feature, height, width)
+        # [batch_size, height/2, width/2, num_channels]
+        input_feature_0 = input_feature[:, 0::2, 0::2, :]
+        # [batch_size, height/2, width/2, num_channels]
+        input_feature_1 = input_feature[:, 1::2, 0::2, :]
+        # [batch_size, height/2, width/2, num_channels]
+        input_feature_2 = input_feature[:, 0::2, 1::2, :]
+        # [batch_size, height/2, width/2, num_channels]
+        input_feature_3 = input_feature[:, 1::2, 1::2, :]
+        # batch_size height/2 width/2 4*num_channels
+        input_feature = torch.cat([input_feature_0, input_feature_1, input_feature_2, input_feature_3], -1)
+        input_feature = input_feature.view(batch_size, -1, 4 * num_channels)  # batch_size height/2*width/2 4*C
+
+        input_feature = self.norm(input_feature)
+        input_feature = self.reduction(input_feature)
+
+        return input_feature
+
+
+# Copied from transformers.models.beit.modeling_beit.drop_path
+def drop_path(input: torch.Tensor, drop_prob: float = 0.0, training: bool = False) -> torch.Tensor:
+    """
+    Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
+
+    Comment by Ross Wightman: This is the same as the DropConnect impl I created for EfficientNet, etc networks,
+    however, the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
+    See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for changing the
+    layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use 'survival rate' as the
+    argument.
+    """
+    if drop_prob == 0.0 or not training:
+        return input
+    keep_prob = 1 - drop_prob
+    shape = (input.shape[0],) + (1,) * (input.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = keep_prob + torch.rand(shape, dtype=input.dtype, device=input.device)
+    random_tensor.floor_()  # binarize
+    output = input.div(keep_prob) * random_tensor
+    return output
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinDropPath
+class UnimerSwinDropPath(nn.Module):
+    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks)."""
+
+    def __init__(self, drop_prob: Optional[float] = None) -> None:
+        super().__init__()
+        self.drop_prob = drop_prob
+
+    def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
+        return drop_path(hidden_states, self.drop_prob, self.training)
+
+    def extra_repr(self) -> str:
+        return "p={}".format(self.drop_prob)
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinSelfAttention with Swin->UnimerSwin
+class UnimerSwinSelfAttention(nn.Module):
+    def __init__(self, config, dim, num_heads, window_size):
+        super().__init__()
+        if dim % num_heads != 0:
+            raise ValueError(
+                f"The hidden size ({dim}) is not a multiple of the number of attention heads ({num_heads})"
+            )
+
+        self.num_attention_heads = num_heads
+        self.attention_head_size = int(dim / num_heads)
+        self.all_head_size = self.num_attention_heads * self.attention_head_size
+        self.window_size = (
+            window_size if isinstance(window_size, collections.abc.Iterable) else (window_size, window_size)
+        )
+
+        self.relative_position_bias_table = nn.Parameter(
+            torch.zeros((2 * self.window_size[0] - 1) * (2 * self.window_size[1] - 1), num_heads)
+        )
+
+        # get pair-wise relative position index for each token inside the window
+        coords_h = torch.arange(self.window_size[0])
+        coords_w = torch.arange(self.window_size[1])
+        coords = torch.stack(meshgrid([coords_h, coords_w], indexing="ij"))
+        coords_flatten = torch.flatten(coords, 1)
+        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]
+        relative_coords = relative_coords.permute(1, 2, 0).contiguous()
+        relative_coords[:, :, 0] += self.window_size[0] - 1
+        relative_coords[:, :, 1] += self.window_size[1] - 1
+        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1
+        relative_position_index = relative_coords.sum(-1)
+        self.register_buffer("relative_position_index", relative_position_index)
+
+        self.query = nn.Linear(self.all_head_size, self.all_head_size, bias=config.qkv_bias)
+        self.key = nn.Linear(self.all_head_size, self.all_head_size, bias=config.qkv_bias)
+        self.value = nn.Linear(self.all_head_size, self.all_head_size, bias=config.qkv_bias)
+
+        self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
+
+    def transpose_for_scores(self, x):
+        new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.attention_head_size)
+        x = x.view(new_x_shape)
+        return x.permute(0, 2, 1, 3)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+    ) -> Tuple[torch.Tensor]:
+        batch_size, dim, num_channels = hidden_states.shape
+        mixed_query_layer = self.query(hidden_states)
+
+        key_layer = self.transpose_for_scores(self.key(hidden_states))
+        value_layer = self.transpose_for_scores(self.value(hidden_states))
+        query_layer = self.transpose_for_scores(mixed_query_layer)
+
+        # Take the dot product between "query" and "key" to get the raw attention scores.
+        attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))
+
+        attention_scores = attention_scores / math.sqrt(self.attention_head_size)
+
+        relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)]
+        relative_position_bias = relative_position_bias.view(
+            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1
+        )
+
+        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()
+        attention_scores = attention_scores + relative_position_bias.unsqueeze(0)
+
+        if attention_mask is not None:
+            # Apply the attention mask is (precomputed for all layers in UnimerSwinModel forward() function)
+            mask_shape = attention_mask.shape[0]
+            attention_scores = attention_scores.view(
+                batch_size // mask_shape, mask_shape, self.num_attention_heads, dim, dim
+            )
+            attention_scores = attention_scores + attention_mask.unsqueeze(1).unsqueeze(0)
+            attention_scores = attention_scores.view(-1, self.num_attention_heads, dim, dim)
+
+        # Normalize the attention scores to probabilities.
+        attention_probs = nn.functional.softmax(attention_scores, dim=-1)
+
+        # This is actually dropping out entire tokens to attend to, which might
+        # seem a bit unusual, but is taken from the original Transformer paper.
+        attention_probs = self.dropout(attention_probs)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attention_probs = attention_probs * head_mask
+
+        context_layer = torch.matmul(attention_probs, value_layer)
+        context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
+        new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,)
+        context_layer = context_layer.view(new_context_layer_shape)
+
+        outputs = (context_layer, attention_probs) if output_attentions else (context_layer,)
+
+        return outputs
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinSelfOutput
+class UnimerSwinSelfOutput(nn.Module):
+    def __init__(self, config, dim):
+        super().__init__()
+        self.dense = nn.Linear(dim, dim)
+        self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
+
+    def forward(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor) -> torch.Tensor:
+        hidden_states = self.dense(hidden_states)
+        hidden_states = self.dropout(hidden_states)
+
+        return hidden_states
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinAttention with Swin->UnimerSwin
+class UnimerSwinAttention(nn.Module):
+    def __init__(self, config, dim, num_heads, window_size):
+        super().__init__()
+        self.self = UnimerSwinSelfAttention(config, dim, num_heads, window_size)
+        self.output = UnimerSwinSelfOutput(config, dim)
+        self.pruned_heads = set()
+
+    def prune_heads(self, heads):
+        if len(heads) == 0:
+            return
+        heads, index = find_pruneable_heads_and_indices(
+            heads, self.self.num_attention_heads, self.self.attention_head_size, self.pruned_heads
+        )
+
+        # Prune linear layers
+        self.self.query = prune_linear_layer(self.self.query, index)
+        self.self.key = prune_linear_layer(self.self.key, index)
+        self.self.value = prune_linear_layer(self.self.value, index)
+        self.output.dense = prune_linear_layer(self.output.dense, index, dim=1)
+
+        # Update hyper params and store pruned heads
+        self.self.num_attention_heads = self.self.num_attention_heads - len(heads)
+        self.self.all_head_size = self.self.attention_head_size * self.self.num_attention_heads
+        self.pruned_heads = self.pruned_heads.union(heads)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+    ) -> Tuple[torch.Tensor]:
+        self_outputs = self.self(hidden_states, attention_mask, head_mask, output_attentions)
+        attention_output = self.output(self_outputs[0], hidden_states)
+        outputs = (attention_output,) + self_outputs[1:]  # add attentions if we output them
+        return outputs
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinIntermediate
+class UnimerSwinIntermediate(nn.Module):
+    def __init__(self, config, dim):
+        super().__init__()
+        self.dense = nn.Linear(dim, int(config.mlp_ratio * dim))
+        if isinstance(config.hidden_act, str):
+            self.intermediate_act_fn = ACT2FN[config.hidden_act]
+        else:
+            self.intermediate_act_fn = config.hidden_act
+
+    def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
+        hidden_states = self.dense(hidden_states)
+        hidden_states = self.intermediate_act_fn(hidden_states)
+        return hidden_states
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinOutput
+class UnimerSwinOutput(nn.Module):
+    def __init__(self, config, dim):
+        super().__init__()
+        self.dense = nn.Linear(int(config.mlp_ratio * dim), dim)
+        self.dropout = nn.Dropout(config.hidden_dropout_prob)
+
+    def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
+        hidden_states = self.dense(hidden_states)
+        hidden_states = self.dropout(hidden_states)
+        return hidden_states
+
+
+class ConvEnhance(nn.Module):
+    """Depth-wise convolution to get the positional information.
+    """
+    def __init__(self, config, dim, k=3):
+        super(ConvEnhance, self).__init__()
+        self.proj = nn.Conv2d(dim,
+                              dim,
+                              (k,k),
+                              (1,1),
+                              (k // 2,k // 2),
+                              groups=dim)
+        self.act_fn = ACT2FN[config.hidden_act]
+
+    def forward(self, x, size: Tuple[int, int]):
+        B, N, C = x.shape
+        H, W = size
+        assert N == H * W
+
+        feat = x.transpose(1, 2).view(B, C, H, W)
+        feat = self.proj(feat)
+        feat = self.act_fn(feat)
+        feat = feat.flatten(2).transpose(1, 2)
+
+        x = x + feat
+        return x
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinLayer with Swin->UnimerSwin
+class UnimerSwinLayer(nn.Module):
+    def __init__(self, config, dim, input_resolution, num_heads, shift_size=0):
+        super().__init__()
+        self.chunk_size_feed_forward = config.chunk_size_feed_forward
+        self.shift_size = shift_size
+        self.window_size = config.window_size
+        self.input_resolution = input_resolution
+        self.layernorm_before = nn.LayerNorm(dim, eps=config.layer_norm_eps)
+
+        self.ce = nn.ModuleList([ConvEnhance(config, dim=dim, k=3),
+                                  ConvEnhance(config, dim=dim, k=3)])
+
+        self.attention = UnimerSwinAttention(config, dim, num_heads, window_size=self.window_size)
+        self.drop_path = UnimerSwinDropPath(config.drop_path_rate) if config.drop_path_rate > 0.0 else nn.Identity()
+        self.layernorm_after = nn.LayerNorm(dim, eps=config.layer_norm_eps)
+        self.intermediate = UnimerSwinIntermediate(config, dim)
+        self.output = UnimerSwinOutput(config, dim)
+
+    def set_shift_and_window_size(self, input_resolution):
+        if min(input_resolution) <= self.window_size:
+            # if window size is larger than input resolution, we don't partition windows
+            self.shift_size = torch_int(0)
+            self.window_size = (
+                torch.min(torch.tensor(input_resolution)) if torch.jit.is_tracing() else min(input_resolution)
+            )
+
+    def get_attn_mask(self, height, width, dtype, device):
+        if self.shift_size > 0:
+            # calculate attention mask for SW-MSA
+            img_mask = torch.zeros((1, height, width, 1), dtype=dtype, device=device)
+            height_slices = (
+                slice(0, -self.window_size),
+                slice(-self.window_size, -self.shift_size),
+                slice(-self.shift_size, None),
+            )
+            width_slices = (
+                slice(0, -self.window_size),
+                slice(-self.window_size, -self.shift_size),
+                slice(-self.shift_size, None),
+            )
+            count = 0
+            for height_slice in height_slices:
+                for width_slice in width_slices:
+                    img_mask[:, height_slice, width_slice, :] = count
+                    count += 1
+
+            mask_windows = window_partition(img_mask, self.window_size)
+            mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
+            attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
+            attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
+        else:
+            attn_mask = None
+        return attn_mask
+
+    def maybe_pad(self, hidden_states, height, width):
+        pad_right = (self.window_size - width % self.window_size) % self.window_size
+        pad_bottom = (self.window_size - height % self.window_size) % self.window_size
+        pad_values = (0, 0, 0, pad_right, 0, pad_bottom)
+        hidden_states = nn.functional.pad(hidden_states, pad_values)
+        return hidden_states, pad_values
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        input_dimensions: Tuple[int, int],
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+        always_partition: Optional[bool] = False,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        if not always_partition:
+            self.set_shift_and_window_size(input_dimensions)
+        else:
+            pass
+        height, width = input_dimensions
+        batch_size, _, channels = hidden_states.size()
+        
+
+
+        hidden_states = self.ce[0](hidden_states, input_dimensions)
+        shortcut = hidden_states
+
+
+        hidden_states = self.layernorm_before(hidden_states)
+        hidden_states = hidden_states.view(batch_size, height, width, channels)
+
+        # pad hidden_states to multiples of window size
+        hidden_states, pad_values = self.maybe_pad(hidden_states, height, width)
+
+        _, height_pad, width_pad, _ = hidden_states.shape
+        # cyclic shift
+        if self.shift_size > 0:
+            shifted_hidden_states = torch.roll(hidden_states, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
+        else:
+            shifted_hidden_states = hidden_states
+
+        # partition windows
+        hidden_states_windows = window_partition(shifted_hidden_states, self.window_size)
+        hidden_states_windows = hidden_states_windows.view(-1, self.window_size * self.window_size, channels)
+        attn_mask = self.get_attn_mask(
+            height_pad, width_pad, dtype=hidden_states.dtype, device=hidden_states_windows.device
+        )
+
+        attention_outputs = self.attention(
+            hidden_states_windows, attn_mask, head_mask, output_attentions=output_attentions
+        )
+
+        attention_output = attention_outputs[0]
+
+        attention_windows = attention_output.view(-1, self.window_size, self.window_size, channels)
+        shifted_windows = window_reverse(attention_windows, self.window_size, height_pad, width_pad)
+
+        # reverse cyclic shift
+        if self.shift_size > 0:
+            attention_windows = torch.roll(shifted_windows, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
+        else:
+            attention_windows = shifted_windows
+
+        was_padded = pad_values[3] > 0 or pad_values[5] > 0
+        if was_padded:
+            attention_windows = attention_windows[:, :height, :width, :].contiguous()
+
+        attention_windows = attention_windows.view(batch_size, height * width, channels)
+
+        hidden_states = shortcut + self.drop_path(attention_windows)
+
+
+
+        hidden_states = self.ce[1](hidden_states, input_dimensions)
+        layer_output = self.layernorm_after(hidden_states)
+        layer_output = self.intermediate(layer_output)
+        layer_output = hidden_states + self.output(layer_output)
+
+        layer_outputs = (layer_output, attention_outputs[1]) if output_attentions else (layer_output,)
+        return layer_outputs
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinStage with Swin->UnimerSwin
+class UnimerSwinStage(nn.Module):
+    def __init__(self, config, dim, input_resolution, depth, num_heads, drop_path, downsample):
+        super().__init__()
+        self.config = config
+        self.dim = dim
+        self.blocks = nn.ModuleList(
+            [
+                UnimerSwinLayer(
+                    config=config,
+                    dim=dim,
+                    input_resolution=input_resolution,
+                    num_heads=num_heads,
+                    shift_size=0,
+                )
+                for i in range(depth)
+            ]
+        )
+
+        # patch merging layer
+        if downsample is not None:
+            self.downsample = downsample(input_resolution, dim=dim, norm_layer=nn.LayerNorm)
+        else:
+            self.downsample = None
+
+        self.pointing = False
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        input_dimensions: Tuple[int, int],
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+        always_partition: Optional[bool] = False,
+    ) -> Tuple[torch.Tensor]:
+        height, width = input_dimensions
+        for i, layer_module in enumerate(self.blocks):
+            layer_head_mask = head_mask[i] if head_mask is not None else None
+
+            layer_outputs = layer_module(
+                hidden_states, input_dimensions, layer_head_mask, output_attentions, always_partition
+            )
+
+            hidden_states = layer_outputs[0]
+
+        hidden_states_before_downsampling = hidden_states
+        if self.downsample is not None:
+            height_downsampled, width_downsampled = (height + 1) // 2, (width + 1) // 2
+            output_dimensions = (height, width, height_downsampled, width_downsampled)
+            hidden_states = self.downsample(hidden_states_before_downsampling, input_dimensions)
+        else:
+            output_dimensions = (height, width, height, width)
+
+        stage_outputs = (hidden_states, hidden_states_before_downsampling, output_dimensions)
+
+        if output_attentions:
+            stage_outputs += layer_outputs[1:]
+        return stage_outputs
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinEncoder with Swin->UnimerSwin
+class UnimerSwinEncoder(nn.Module):
+    def __init__(self, config, grid_size):
+        super().__init__()
+        self.num_layers = len(config.depths)
+        self.config = config
+        dpr = [x.item() for x in torch.linspace(0, config.drop_path_rate, sum(config.depths))]
+        self.layers = nn.ModuleList(
+            [
+                UnimerSwinStage(
+                    config=config,
+                    dim=int(config.embed_dim * 2**i_layer),
+                    input_resolution=(grid_size[0] // (2**i_layer), grid_size[1] // (2**i_layer)),
+                    depth=config.depths[i_layer],
+                    num_heads=config.num_heads[i_layer],
+                    drop_path=dpr[sum(config.depths[:i_layer]) : sum(config.depths[: i_layer + 1])],
+                    downsample=UnimerSwinPatchMerging if (i_layer < self.num_layers - 1) else None,
+                )
+                for i_layer in range(self.num_layers)
+            ]
+        )
+
+        self.gradient_checkpointing = False
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        input_dimensions: Tuple[int, int],
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+        output_hidden_states: Optional[bool] = False,
+        output_hidden_states_before_downsampling: Optional[bool] = False,
+        always_partition: Optional[bool] = False,
+        return_dict: Optional[bool] = True,
+    ) -> Union[Tuple, UnimerSwinEncoderOutput]:
+        all_hidden_states = () if output_hidden_states else None
+        all_reshaped_hidden_states = () if output_hidden_states else None
+        all_self_attentions = () if output_attentions else None
+
+        if output_hidden_states:
+            batch_size, _, hidden_size = hidden_states.shape
+            # rearrange b (h w) c -> b c h w
+            reshaped_hidden_state = hidden_states.view(batch_size, *input_dimensions, hidden_size)
+            reshaped_hidden_state = reshaped_hidden_state.permute(0, 3, 1, 2)
+            all_hidden_states += (hidden_states,)
+            all_reshaped_hidden_states += (reshaped_hidden_state,)
+
+        for i, layer_module in enumerate(self.layers):
+            layer_head_mask = head_mask[i] if head_mask is not None else None
+
+            if self.gradient_checkpointing and self.training:
+                layer_outputs = self._gradient_checkpointing_func(
+                    layer_module.__call__,
+                    hidden_states,
+                    input_dimensions,
+                    layer_head_mask,
+                    output_attentions,
+                    always_partition,
+                )
+            else:
+                layer_outputs = layer_module(
+                    hidden_states, input_dimensions, layer_head_mask, output_attentions, always_partition
+                )
+
+            hidden_states = layer_outputs[0]
+            hidden_states_before_downsampling = layer_outputs[1]
+            output_dimensions = layer_outputs[2]
+
+            input_dimensions = (output_dimensions[-2], output_dimensions[-1])
+
+            if output_hidden_states and output_hidden_states_before_downsampling:
+                batch_size, _, hidden_size = hidden_states_before_downsampling.shape
+                # rearrange b (h w) c -> b c h w
+                # here we use the original (not downsampled) height and width
+                reshaped_hidden_state = hidden_states_before_downsampling.view(
+                    batch_size, *(output_dimensions[0], output_dimensions[1]), hidden_size
+                )
+                reshaped_hidden_state = reshaped_hidden_state.permute(0, 3, 1, 2)
+                all_hidden_states += (hidden_states_before_downsampling,)
+                all_reshaped_hidden_states += (reshaped_hidden_state,)
+            elif output_hidden_states and not output_hidden_states_before_downsampling:
+                batch_size, _, hidden_size = hidden_states.shape
+                # rearrange b (h w) c -> b c h w
+                reshaped_hidden_state = hidden_states.view(batch_size, *input_dimensions, hidden_size)
+                reshaped_hidden_state = reshaped_hidden_state.permute(0, 3, 1, 2)
+                all_hidden_states += (hidden_states,)
+                all_reshaped_hidden_states += (reshaped_hidden_state,)
+
+            if output_attentions:
+                all_self_attentions += layer_outputs[3:]
+
+        if not return_dict:
+            return tuple(v for v in [hidden_states, all_hidden_states, all_self_attentions] if v is not None)
+
+        return UnimerSwinEncoderOutput(
+            last_hidden_state=hidden_states,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
+            reshaped_hidden_states=all_reshaped_hidden_states,
+        )
+
+
+# Copied from transformers.models.swin.modeling_swin.SwinPreTrainedModel with Swin->UnimerSwin
+class UnimerSwinPreTrainedModel(PreTrainedModel):
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = UnimerSwinConfig
+    base_model_prefix = "unimer-swin"
+    main_input_name = "pixel_values"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["UnimerSwinStage"]
+
+    def _init_weights(self, module):
+        """Initialize the weights"""
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            # Slightly different from the TF version which uses truncated_normal for initialization
+            # cf https://github.com/pytorch/pytorch/pull/5617
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+
+SWIN_START_DOCSTRING = r"""
+    This model is a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) sub-class. Use
+    it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and
+    behavior.
+
+    Parameters:
+        config ([`UnimerSwinConfig`]): Model configuration class with all the parameters of the model.
+            Initializing with a config file does not load the weights associated with the model, only the
+            configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+SWIN_INPUTS_DOCSTRING = r"""
+    Args:
+        pixel_values (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)`):
+            Pixel values. Pixel values can be obtained using [`AutoImageProcessor`]. See
+            [`DonutImageProcessor.__call__`] for details.
+        head_mask (`torch.FloatTensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*):
+            Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+
+        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.
+        interpolate_pos_encoding (`bool`, *optional*, defaults to `False`):
+            Whether to interpolate the pre-trained position encodings.
+        return_dict (`bool`, *optional*):
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
+"""
+
+
+@add_start_docstrings(
+    "The bare UnimerSwin Model transformer outputting raw hidden-states without any specific head on top.",
+    SWIN_START_DOCSTRING,
+)
+class UnimerSwinModel(UnimerSwinPreTrainedModel):
+    def __init__(self, config, add_pooling_layer=True, use_mask_token=False):
+        super().__init__(config)
+        self.config = config
+        self.num_layers = len(config.depths)
+        self.num_features = int(config.embed_dim * 2 ** (self.num_layers - 1))
+
+        self.embeddings = UnimerSwinEmbeddings(config, use_mask_token=use_mask_token)
+        self.encoder = UnimerSwinEncoder(config, self.embeddings.patch_grid)
+        self.pooler = nn.AdaptiveAvgPool1d(1) if add_pooling_layer else None
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.embeddings.patch_embeddings
+
+    def _prune_heads(self, heads_to_prune):
+        """
+        Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base
+        class PreTrainedModel
+        """
+        for layer, heads in heads_to_prune.items():
+            self.encoder.layer[layer].attention.prune_heads(heads)
+
+    @add_start_docstrings_to_model_forward(SWIN_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=UnimerSwinModelOutput,
+        config_class=_CONFIG_FOR_DOC,
+        modality="vision",
+        expected_output=_EXPECTED_OUTPUT_SHAPE,
+    )
+    def forward(
+        self,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        bool_masked_pos: Optional[torch.BoolTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        interpolate_pos_encoding: bool = False,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, UnimerSwinModelOutput]:
+        r"""
+        bool_masked_pos (`torch.BoolTensor` of shape `(batch_size, num_patches)`):
+            Boolean masked positions. Indicates which patches are masked (1) and which aren't (0).
+        """
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if pixel_values is None:
+            raise ValueError("You have to specify pixel_values")
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_heads x N x N
+        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
+        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
+        head_mask = self.get_head_mask(head_mask, len(self.config.depths))
+
+        embedding_output, input_dimensions = self.embeddings(
+            pixel_values, bool_masked_pos=bool_masked_pos, interpolate_pos_encoding=interpolate_pos_encoding
+        )
+
+        encoder_outputs = self.encoder(
+            embedding_output,
+            input_dimensions,
+            head_mask=head_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+
+        sequence_output = encoder_outputs[0]
+
+        pooled_output = None
+        if self.pooler is not None:
+            pooled_output = self.pooler(sequence_output.transpose(1, 2))
+            pooled_output = torch.flatten(pooled_output, 1)
+
+        if not return_dict:
+            output = (sequence_output, pooled_output) + encoder_outputs[1:]
+
+            return output
+
+        return UnimerSwinModelOutput(
+            last_hidden_state=sequence_output,
+            pooler_output=pooled_output,
+            hidden_states=encoder_outputs.hidden_states,
+            attentions=encoder_outputs.attentions,
+            reshaped_hidden_states=encoder_outputs.reshaped_hidden_states,
+        )

mineru/model/ocr/__init__.py

+# Copyright (c) Opendatalab. All rights reserved.

mineru/model/ocr/paddleocr2pytorch/__init__.py

+# Copyright (c) Opendatalab. All rights reserved.

mineru/model/ocr/paddleocr2pytorch/pytorch_paddle.py

+# Copyright (c) Opendatalab. All rights reserved.
+import copy
+import os.path
+import warnings
+from pathlib import Path
+
+import cv2
+import numpy as np
+import yaml
+from loguru import logger
+
+from magic_pdf.libs.config_reader import get_device, get_local_models_dir
+from ....utils.ocr_utils import check_img, preprocess_image, sorted_boxes, merge_det_boxes, update_det_boxes, get_rotate_crop_image
+from .tools.infer.predict_system import TextSystem
+from .tools.infer import pytorchocr_utility as utility
+import argparse
+
+
+latin_lang = [
+        'af', 'az', 'bs', 'cs', 'cy', 'da', 'de', 'es', 'et', 'fr', 'ga', 'hr',  # noqa: E126
+        'hu', 'id', 'is', 'it', 'ku', 'la', 'lt', 'lv', 'mi', 'ms', 'mt', 'nl',
+        'no', 'oc', 'pi', 'pl', 'pt', 'ro', 'rs_latin', 'sk', 'sl', 'sq', 'sv',
+        'sw', 'tl', 'tr', 'uz', 'vi', 'french', 'german'
+]
+arabic_lang = ['ar', 'fa', 'ug', 'ur']
+cyrillic_lang = [
+        'ru', 'rs_cyrillic', 'be', 'bg', 'uk', 'mn', 'abq', 'ady', 'kbd', 'ava',  # noqa: E126
+        'dar', 'inh', 'che', 'lbe', 'lez', 'tab'
+]
+devanagari_lang = [
+        'hi', 'mr', 'ne', 'bh', 'mai', 'ang', 'bho', 'mah', 'sck', 'new', 'gom',  # noqa: E126
+        'sa', 'bgc'
+]
+
+
+def get_model_params(lang, config):
+    if lang in config['lang']:
+        params = config['lang'][lang]
+        det = params.get('det')
+        rec = params.get('rec')
+        dict_file = params.get('dict')
+        return det, rec, dict_file
+    else:
+        raise Exception (f'Language {lang} not supported')
+
+
+root_dir = Path(__file__).resolve().parent
+
+
+class PytorchPaddleOCR(TextSystem):
+    def __init__(self, *args, **kwargs):
+        parser = utility.init_args()
+        args = parser.parse_args(args)
+
+        self.lang = kwargs.get('lang', 'ch')
+
+        device = get_device()
+        if device == 'cpu' and self.lang in ['ch', 'ch_server']:
+            logger.warning("The current device in use is CPU. To ensure the speed of parsing, the language is automatically switched to ch_lite.")
+            self.lang = 'ch_lite'
+
+        if self.lang in latin_lang:
+            self.lang = 'latin'
+        elif self.lang in arabic_lang:
+            self.lang = 'arabic'
+        elif self.lang in cyrillic_lang:
+            self.lang = 'cyrillic'
+        elif self.lang in devanagari_lang:
+            self.lang = 'devanagari'
+        else:
+            pass
+
+        models_config_path = os.path.join(root_dir, 'pytorchocr', 'utils', 'resources', 'models_config.yml')
+        with open(models_config_path) as file:
+            config = yaml.safe_load(file)
+            det, rec, dict_file = get_model_params(self.lang, config)
+        ocr_models_dir = os.path.join(get_local_models_dir(), 'OCR', 'paddleocr_torch')
+        kwargs['det_model_path'] = os.path.join(ocr_models_dir, det)
+        kwargs['rec_model_path'] = os.path.join(ocr_models_dir, rec)
+        kwargs['rec_char_dict_path'] = os.path.join(root_dir, 'pytorchocr', 'utils', 'resources', 'dict', dict_file)
+        # kwargs['rec_batch_num'] = 8
+
+        kwargs['device'] = device
+
+        default_args = vars(args)
+        default_args.update(kwargs)
+        args = argparse.Namespace(**default_args)
+
+        super().__init__(args)
+
+    def ocr(self,
+            img,
+            det=True,
+            rec=True,
+            mfd_res=None,
+            tqdm_enable=False,
+            ):
+        assert isinstance(img, (np.ndarray, list, str, bytes))
+        if isinstance(img, list) and det == True:
+            logger.error('When input a list of images, det must be false')
+            exit(0)
+        img = check_img(img)
+        imgs = [img]
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore", category=RuntimeWarning)
+            if det and rec:
+                ocr_res = []
+                for img in imgs:
+                    img = preprocess_image(img)
+                    dt_boxes, rec_res = self.__call__(img, mfd_res=mfd_res)
+                    if not dt_boxes and not rec_res:
+                        ocr_res.append(None)
+                        continue
+                    tmp_res = [[box.tolist(), res] for box, res in zip(dt_boxes, rec_res)]
+                    ocr_res.append(tmp_res)
+                return ocr_res
+            elif det and not rec:
+                ocr_res = []
+                for img in imgs:
+                    img = preprocess_image(img)
+                    dt_boxes, elapse = self.text_detector(img)
+                    # logger.debug("dt_boxes num : {}, elapsed : {}".format(len(dt_boxes), elapse))
+                    if dt_boxes is None:
+                        ocr_res.append(None)
+                        continue
+                    dt_boxes = sorted_boxes(dt_boxes)
+                    # merge_det_boxes 和 update_det_boxes 都会把poly转成bbox再转回poly，因此需要过滤所有倾斜程度较大的文本框
+                    dt_boxes = merge_det_boxes(dt_boxes)
+                    if mfd_res:
+                        dt_boxes = update_det_boxes(dt_boxes, mfd_res)
+                    tmp_res = [box.tolist() for box in dt_boxes]
+                    ocr_res.append(tmp_res)
+                return ocr_res
+            elif not det and rec:
+                ocr_res = []
+                for img in imgs:
+                    if not isinstance(img, list):
+                        img = preprocess_image(img)
+                        img = [img]
+                    rec_res, elapse = self.text_recognizer(img, tqdm_enable=tqdm_enable)
+                    # logger.debug("rec_res num  : {}, elapsed : {}".format(len(rec_res), elapse))
+                    ocr_res.append(rec_res)
+                return ocr_res
+
+    def __call__(self, img, mfd_res=None):
+
+        if img is None:
+            logger.debug("no valid image provided")
+            return None, None
+
+        ori_im = img.copy()
+        dt_boxes, elapse = self.text_detector(img)
+
+        if dt_boxes is None:
+            logger.debug("no dt_boxes found, elapsed : {}".format(elapse))
+            return None, None
+        else:
+            pass
+            # logger.debug("dt_boxes num : {}, elapsed : {}".format(len(dt_boxes), elapse))
+        img_crop_list = []
+
+        dt_boxes = sorted_boxes(dt_boxes)
+
+        # merge_det_boxes 和 update_det_boxes 都会把poly转成bbox再转回poly，因此需要过滤所有倾斜程度较大的文本框
+        dt_boxes = merge_det_boxes(dt_boxes)
+
+        if mfd_res:
+            dt_boxes = update_det_boxes(dt_boxes, mfd_res)
+
+        for bno in range(len(dt_boxes)):
+            tmp_box = copy.deepcopy(dt_boxes[bno])
+            img_crop = get_rotate_crop_image(ori_im, tmp_box)
+            img_crop_list.append(img_crop)
+
+        rec_res, elapse = self.text_recognizer(img_crop_list)
+        # logger.debug("rec_res num  : {}, elapsed : {}".format(len(rec_res), elapse))
+
+        filter_boxes, filter_rec_res = [], []
+        for box, rec_result in zip(dt_boxes, rec_res):
+            text, score = rec_result
+            if score >= self.drop_score:
+                filter_boxes.append(box)
+                filter_rec_res.append(rec_result)
+
+        return filter_boxes, filter_rec_res
+
+if __name__ == '__main__':
+    pytorch_paddle_ocr = PytorchPaddleOCR()
+    img = cv2.imread("/Users/myhloli/Downloads/screenshot-20250326-194348.png")
+    dt_boxes, rec_res = pytorch_paddle_ocr(img)
+    ocr_res = []
+    if not dt_boxes and not rec_res:
+        ocr_res.append(None)
+    else:
+        tmp_res = [[box.tolist(), res] for box, res in zip(dt_boxes, rec_res)]
+        ocr_res.append(tmp_res)
+    print(ocr_res)
+
+

mineru/model/ocr/paddleocr2pytorch/pytorchocr/base_ocr_v20.py

+import os
+import torch
+from .modeling.architectures.base_model import BaseModel
+
+class BaseOCRV20:
+    def __init__(self, config, **kwargs):
+        self.config = config
+        self.build_net(**kwargs)
+        self.net.eval()
+
+
+    def build_net(self, **kwargs):
+        self.net = BaseModel(self.config, **kwargs)
+
+    def read_pytorch_weights(self, weights_path):
+        if not os.path.exists(weights_path):
+            raise FileNotFoundError('{} is not existed.'.format(weights_path))
+        weights = torch.load(weights_path)
+        return weights
+
+    def get_out_channels(self, weights):
+        if list(weights.keys())[-1].endswith('.weight') and len(list(weights.values())[-1].shape) == 2:
+            out_channels = list(weights.values())[-1].numpy().shape[1]
+        else:
+            out_channels = list(weights.values())[-1].numpy().shape[0]
+        return out_channels
+
+    def load_state_dict(self, weights):
+        self.net.load_state_dict(weights)
+        # print('weights is loaded.')
+
+    def load_pytorch_weights(self, weights_path):
+        self.net.load_state_dict(torch.load(weights_path, weights_only=True))
+        # print('model is loaded: {}'.format(weights_path))
+
+    def inference(self, inputs):
+        with torch.no_grad():
+            infer = self.net(inputs)
+        return infer

mineru/model/ocr/paddleocr2pytorch/pytorchocr/data/__init__.py

+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+from .imaug import transform, create_operators
+
+

mineru/model/ocr/paddleocr2pytorch/pytorchocr/data/imaug/__init__.py

+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+# from .iaa_augment import IaaAugment
+# from .make_border_map import MakeBorderMap
+# from .make_shrink_map import MakeShrinkMap
+# from .random_crop_data import EastRandomCropData, PSERandomCrop
+
+# from .rec_img_aug import RecAug, RecResizeImg, ClsResizeImg
+# from .randaugment import RandAugment
+from .operators import *
+# from .label_ops import *
+
+# from .east_process import *
+# from .sast_process import *
+# from .gen_table_mask import *
+
+def transform(data, ops=None):
+    """ transform """
+    if ops is None:
+        ops = []
+    for op in ops:
+        data = op(data)
+        if data is None:
+            return None
+    return data
+
+
+def create_operators(op_param_list, global_config=None):
+    """
+    create operators based on the config
+    Args:
+        params(list): a dict list, used to create some operators
+    """
+    assert isinstance(op_param_list, list), ('operator config should be a list')
+    ops = []
+    for operator in op_param_list:
+        assert isinstance(operator,
+                          dict) and len(operator) == 1, "yaml format error"
+        op_name = list(operator)[0]
+        param = {} if operator[op_name] is None else operator[op_name]
+        if global_config is not None:
+            param.update(global_config)
+        op = eval(op_name)(**param)
+        ops.append(op)
+    return ops
\ No newline at end of file

mineru/model/ocr/paddleocr2pytorch/pytorchocr/data/imaug/operators.py

+"""
+# Copyright (c) 2020 PaddlePaddle Authors. 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.
+"""
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+import sys
+import six
+import cv2
+import numpy as np
+
+
+class DecodeImage(object):
+    """ decode image """
+
+    def __init__(self, img_mode='RGB', channel_first=False, **kwargs):
+        self.img_mode = img_mode
+        self.channel_first = channel_first
+
+    def __call__(self, data):
+        img = data['image']
+        if six.PY2:
+            assert type(img) is str and len(
+                img) > 0, "invalid input 'img' in DecodeImage"
+        else:
+            assert type(img) is bytes and len(
+                img) > 0, "invalid input 'img' in DecodeImage"
+        img = np.frombuffer(img, dtype='uint8')
+        img = cv2.imdecode(img, 1)
+        if img is None:
+            return None
+        if self.img_mode == 'GRAY':
+            img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR)
+        elif self.img_mode == 'RGB':
+            assert img.shape[2] == 3, 'invalid shape of image[%s]' % (img.shape)
+            img = img[:, :, ::-1]
+
+        if self.channel_first:
+            img = img.transpose((2, 0, 1))
+
+        data['image'] = img
+        return data
+
+
+class NRTRDecodeImage(object):
+    """ decode image """
+
+    def __init__(self, img_mode='RGB', channel_first=False, **kwargs):
+        self.img_mode = img_mode
+        self.channel_first = channel_first
+
+    def __call__(self, data):
+        img = data['image']
+        if six.PY2:
+            assert type(img) is str and len(
+                img) > 0, "invalid input 'img' in DecodeImage"
+        else:
+            assert type(img) is bytes and len(
+                img) > 0, "invalid input 'img' in DecodeImage"
+        img = np.frombuffer(img, dtype='uint8')
+
+        img = cv2.imdecode(img, 1)
+
+        if img is None:
+            return None
+        if self.img_mode == 'GRAY':
+            img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR)
+        elif self.img_mode == 'RGB':
+            assert img.shape[2] == 3, 'invalid shape of image[%s]' % (img.shape)
+            img = img[:, :, ::-1]
+        img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+        if self.channel_first:
+            img = img.transpose((2, 0, 1))
+        data['image'] = img
+        return data
+
+
+class NormalizeImage(object):
+    """ normalize image such as substract mean, divide std
+    """
+
+    def __init__(self, scale=None, mean=None, std=None, order='chw', **kwargs):
+        if isinstance(scale, str):
+            scale = eval(scale)
+        self.scale = np.float32(scale if scale is not None else 1.0 / 255.0)
+        mean = mean if mean is not None else [0.485, 0.456, 0.406]
+        std = std if std is not None else [0.229, 0.224, 0.225]
+
+        shape = (3, 1, 1) if order == 'chw' else (1, 1, 3)
+        self.mean = np.array(mean).reshape(shape).astype('float32')
+        self.std = np.array(std).reshape(shape).astype('float32')
+
+    def __call__(self, data):
+        img = data['image']
+        from PIL import Image
+        if isinstance(img, Image.Image):
+            img = np.array(img)
+        assert isinstance(img,
+                          np.ndarray), "invalid input 'img' in NormalizeImage"
+        data['image'] = (
+            img.astype('float32') * self.scale - self.mean) / self.std
+        return data
+
+
+class ToCHWImage(object):
+    """ convert hwc image to chw image
+    """
+
+    def __init__(self, **kwargs):
+        pass
+
+    def __call__(self, data):
+        img = data['image']
+        from PIL import Image
+        if isinstance(img, Image.Image):
+            img = np.array(img)
+        data['image'] = img.transpose((2, 0, 1))
+        return data
+
+
+class Fasttext(object):
+    def __init__(self, path="None", **kwargs):
+        import fasttext
+        self.fast_model = fasttext.load_model(path)
+
+    def __call__(self, data):
+        label = data['label']
+        fast_label = self.fast_model[label]
+        data['fast_label'] = fast_label
+        return data
+
+
+class KeepKeys(object):
+    def __init__(self, keep_keys, **kwargs):
+        self.keep_keys = keep_keys
+
+    def __call__(self, data):
+        data_list = []
+        for key in self.keep_keys:
+            data_list.append(data[key])
+        return data_list
+
+
+class Resize(object):
+    def __init__(self, size=(640, 640), **kwargs):
+        self.size = size
+
+    def resize_image(self, img):
+        resize_h, resize_w = self.size
+        ori_h, ori_w = img.shape[:2]  # (h, w, c)
+        ratio_h = float(resize_h) / ori_h
+        ratio_w = float(resize_w) / ori_w
+        img = cv2.resize(img, (int(resize_w), int(resize_h)))
+        return img, [ratio_h, ratio_w]
+
+    def __call__(self, data):
+        img = data['image']
+        text_polys = data['polys']
+
+        img_resize, [ratio_h, ratio_w] = self.resize_image(img)
+        new_boxes = []
+        for box in text_polys:
+            new_box = []
+            for cord in box:
+                new_box.append([cord[0] * ratio_w, cord[1] * ratio_h])
+            new_boxes.append(new_box)
+        data['image'] = img_resize
+        data['polys'] = np.array(new_boxes, dtype=np.float32)
+        return data
+
+
+class DetResizeForTest(object):
+    def __init__(self, **kwargs):
+        super(DetResizeForTest, self).__init__()
+        self.resize_type = 0
+        if 'image_shape' in kwargs:
+            self.image_shape = kwargs['image_shape']
+            self.resize_type = 1
+        elif 'limit_side_len' in kwargs:
+            self.limit_side_len = kwargs['limit_side_len']
+            self.limit_type = kwargs.get('limit_type', 'min')
+        elif 'resize_long' in kwargs:
+            self.resize_type = 2
+            self.resize_long = kwargs.get('resize_long', 960)
+        else:
+            self.limit_side_len = 736
+            self.limit_type = 'min'
+
+    def __call__(self, data):
+        img = data['image']
+        src_h, src_w, _ = img.shape
+
+        if self.resize_type == 0:
+            # img, shape = self.resize_image_type0(img)
+            img, [ratio_h, ratio_w] = self.resize_image_type0(img)
+        elif self.resize_type == 2:
+            img, [ratio_h, ratio_w] = self.resize_image_type2(img)
+        else:
+            # img, shape = self.resize_image_type1(img)
+            img, [ratio_h, ratio_w] = self.resize_image_type1(img)
+        data['image'] = img
+        data['shape'] = np.array([src_h, src_w, ratio_h, ratio_w])
+        return data
+
+    def resize_image_type1(self, img):
+        resize_h, resize_w = self.image_shape
+        ori_h, ori_w = img.shape[:2]  # (h, w, c)
+        ratio_h = float(resize_h) / ori_h
+        ratio_w = float(resize_w) / ori_w
+        img = cv2.resize(img, (int(resize_w), int(resize_h)))
+        # return img, np.array([ori_h, ori_w])
+        return img, [ratio_h, ratio_w]
+
+    def resize_image_type0(self, img):
+        """
+        resize image to a size multiple of 32 which is required by the network
+        args:
+            img(array): array with shape [h, w, c]
+        return(tuple):
+            img, (ratio_h, ratio_w)
+        """
+        limit_side_len = self.limit_side_len
+        h, w, c = img.shape
+
+        # limit the max side
+        if self.limit_type == 'max':
+            if max(h, w) > limit_side_len:
+                if h > w:
+                    ratio = float(limit_side_len) / h
+                else:
+                    ratio = float(limit_side_len) / w
+            else:
+                ratio = 1.
+        elif self.limit_type == 'min':
+            if min(h, w) < limit_side_len:
+                if h < w:
+                    ratio = float(limit_side_len) / h
+                else:
+                    ratio = float(limit_side_len) / w
+            else:
+                ratio = 1.
+        elif self.limit_type == 'resize_long':
+            ratio = float(limit_side_len) / max(h, w)
+        else:
+            raise Exception('not support limit type, image ')
+        resize_h = int(h * ratio)
+        resize_w = int(w * ratio)
+
+        resize_h = max(int(round(resize_h / 32) * 32), 32)
+        resize_w = max(int(round(resize_w / 32) * 32), 32)
+
+        try:
+            if int(resize_w) <= 0 or int(resize_h) <= 0:
+                return None, (None, None)
+            img = cv2.resize(img, (int(resize_w), int(resize_h)))
+        except:
+            print(img.shape, resize_w, resize_h)
+            sys.exit(0)
+        ratio_h = resize_h / float(h)
+        ratio_w = resize_w / float(w)
+        return img, [ratio_h, ratio_w]
+
+    def resize_image_type2(self, img):
+        h, w, _ = img.shape
+
+        resize_w = w
+        resize_h = h
+
+        if resize_h > resize_w:
+            ratio = float(self.resize_long) / resize_h
+        else:
+            ratio = float(self.resize_long) / resize_w
+
+        resize_h = int(resize_h * ratio)
+        resize_w = int(resize_w * ratio)
+
+        max_stride = 128
+        resize_h = (resize_h + max_stride - 1) // max_stride * max_stride
+        resize_w = (resize_w + max_stride - 1) // max_stride * max_stride
+        img = cv2.resize(img, (int(resize_w), int(resize_h)))
+        ratio_h = resize_h / float(h)
+        ratio_w = resize_w / float(w)
+
+        return img, [ratio_h, ratio_w]
+
+
+class E2EResizeForTest(object):
+    def __init__(self, **kwargs):
+        super(E2EResizeForTest, self).__init__()
+        self.max_side_len = kwargs['max_side_len']
+        self.valid_set = kwargs['valid_set']
+
+    def __call__(self, data):
+        img = data['image']
+        src_h, src_w, _ = img.shape
+        if self.valid_set == 'totaltext':
+            im_resized, [ratio_h, ratio_w] = self.resize_image_for_totaltext(
+                img, max_side_len=self.max_side_len)
+        else:
+            im_resized, (ratio_h, ratio_w) = self.resize_image(
+                img, max_side_len=self.max_side_len)
+        data['image'] = im_resized
+        data['shape'] = np.array([src_h, src_w, ratio_h, ratio_w])
+        return data
+
+    def resize_image_for_totaltext(self, im, max_side_len=512):
+
+        h, w, _ = im.shape
+        resize_w = w
+        resize_h = h
+        ratio = 1.25
+        if h * ratio > max_side_len:
+            ratio = float(max_side_len) / resize_h
+        resize_h = int(resize_h * ratio)
+        resize_w = int(resize_w * ratio)
+
+        max_stride = 128
+        resize_h = (resize_h + max_stride - 1) // max_stride * max_stride
+        resize_w = (resize_w + max_stride - 1) // max_stride * max_stride
+        im = cv2.resize(im, (int(resize_w), int(resize_h)))
+        ratio_h = resize_h / float(h)
+        ratio_w = resize_w / float(w)
+        return im, (ratio_h, ratio_w)
+
+    def resize_image(self, im, max_side_len=512):
+        """
+        resize image to a size multiple of max_stride which is required by the network
+        :param im: the resized image
+        :param max_side_len: limit of max image size to avoid out of memory in gpu
+        :return: the resized image and the resize ratio
+        """
+        h, w, _ = im.shape
+
+        resize_w = w
+        resize_h = h
+
+        # Fix the longer side
+        if resize_h > resize_w:
+            ratio = float(max_side_len) / resize_h
+        else:
+            ratio = float(max_side_len) / resize_w
+
+        resize_h = int(resize_h * ratio)
+        resize_w = int(resize_w * ratio)
+
+        max_stride = 128
+        resize_h = (resize_h + max_stride - 1) // max_stride * max_stride
+        resize_w = (resize_w + max_stride - 1) // max_stride * max_stride
+        im = cv2.resize(im, (int(resize_w), int(resize_h)))
+        ratio_h = resize_h / float(h)
+        ratio_w = resize_w / float(w)
+
+        return im, (ratio_h, ratio_w)
+
+
+class KieResize(object):
+    def __init__(self, **kwargs):
+        super(KieResize, self).__init__()
+        self.max_side, self.min_side = kwargs['img_scale'][0], kwargs[
+            'img_scale'][1]
+
+    def __call__(self, data):
+        img = data['image']
+        points = data['points']
+        src_h, src_w, _ = img.shape
+        im_resized, scale_factor, [ratio_h, ratio_w
+                                   ], [new_h, new_w] = self.resize_image(img)
+        resize_points = self.resize_boxes(img, points, scale_factor)
+        data['ori_image'] = img
+        data['ori_boxes'] = points
+        data['points'] = resize_points
+        data['image'] = im_resized
+        data['shape'] = np.array([new_h, new_w])
+        return data
+
+    def resize_image(self, img):
+        norm_img = np.zeros([1024, 1024, 3], dtype='float32')
+        scale = [512, 1024]
+        h, w = img.shape[:2]
+        max_long_edge = max(scale)
+        max_short_edge = min(scale)
+        scale_factor = min(max_long_edge / max(h, w),
+                           max_short_edge / min(h, w))
+        resize_w, resize_h = int(w * float(scale_factor) + 0.5), int(h * float(
+            scale_factor) + 0.5)
+        max_stride = 32
+        resize_h = (resize_h + max_stride - 1) // max_stride * max_stride
+        resize_w = (resize_w + max_stride - 1) // max_stride * max_stride
+        im = cv2.resize(img, (resize_w, resize_h))
+        new_h, new_w = im.shape[:2]
+        w_scale = new_w / w
+        h_scale = new_h / h
+        scale_factor = np.array(
+            [w_scale, h_scale, w_scale, h_scale], dtype=np.float32)
+        norm_img[:new_h, :new_w, :] = im
+        return norm_img, scale_factor, [h_scale, w_scale], [new_h, new_w]
+
+    def resize_boxes(self, im, points, scale_factor):
+        points = points * scale_factor
+        img_shape = im.shape[:2]
+        points[:, 0::2] = np.clip(points[:, 0::2], 0, img_shape[1])
+        points[:, 1::2] = np.clip(points[:, 1::2], 0, img_shape[0])
+        return points

mineru/model/ocr/paddleocr2pytorch/pytorchocr/modeling/architectures/__init__.py

+# Copyright (c) 2020 PaddlePaddle Authors. 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.
+
+import copy
+
+__all__ = ["build_model"]
+
+
+def build_model(config, **kwargs):
+    from .base_model import BaseModel
+
+    config = copy.deepcopy(config)
+    module_class = BaseModel(config, **kwargs)
+    return module_class

mineru/model/ocr/paddleocr2pytorch/pytorchocr/modeling/architectures/base_model.py

+from torch import nn
+
+from ..backbones import build_backbone
+from ..heads import build_head
+from ..necks import build_neck
+
+
+class BaseModel(nn.Module):
+    def __init__(self, config, **kwargs):
+        """
+        the module for OCR.
+        args:
+            config (dict): the super parameters for module.
+        """
+        super(BaseModel, self).__init__()
+
+        in_channels = config.get("in_channels", 3)
+        model_type = config["model_type"]
+        # build backbone, backbone is need for del, rec and cls
+        if "Backbone" not in config or config["Backbone"] is None:
+            self.use_backbone = False
+        else:
+            self.use_backbone = True
+            config["Backbone"]["in_channels"] = in_channels
+            self.backbone = build_backbone(config["Backbone"], model_type)
+            in_channels = self.backbone.out_channels
+
+        # build neck
+        # for rec, neck can be cnn,rnn or reshape(None)
+        # for det, neck can be FPN, BIFPN and so on.
+        # for cls, neck should be none
+        if "Neck" not in config or config["Neck"] is None:
+            self.use_neck = False
+        else:
+            self.use_neck = True
+            config["Neck"]["in_channels"] = in_channels
+            self.neck = build_neck(config["Neck"])
+            in_channels = self.neck.out_channels
+
+        # # build head, head is need for det, rec and cls
+        if "Head" not in config or config["Head"] is None:
+            self.use_head = False
+        else:
+            self.use_head = True
+            config["Head"]["in_channels"] = in_channels
+            self.head = build_head(config["Head"], **kwargs)
+
+        self.return_all_feats = config.get("return_all_feats", False)
+
+        self._initialize_weights()
+
+    def _initialize_weights(self):
+        # weight initialization
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                nn.init.kaiming_normal_(m.weight, mode="fan_out")
+                if m.bias is not None:
+                    nn.init.zeros_(m.bias)
+            elif isinstance(m, nn.BatchNorm2d):
+                nn.init.ones_(m.weight)
+                nn.init.zeros_(m.bias)
+            elif isinstance(m, nn.Linear):
+                nn.init.normal_(m.weight, 0, 0.01)
+                if m.bias is not None:
+                    nn.init.zeros_(m.bias)
+            elif isinstance(m, nn.ConvTranspose2d):
+                nn.init.kaiming_normal_(m.weight, mode="fan_out")
+                if m.bias is not None:
+                    nn.init.zeros_(m.bias)
+
+    def forward(self, x):
+        y = dict()
+        if self.use_backbone:
+            x = self.backbone(x)
+        if isinstance(x, dict):
+            y.update(x)
+        else:
+            y["backbone_out"] = x
+        final_name = "backbone_out"
+        if self.use_neck:
+            x = self.neck(x)
+            if isinstance(x, dict):
+                y.update(x)
+            else:
+                y["neck_out"] = x
+            final_name = "neck_out"
+        if self.use_head:
+            x = self.head(x)
+        # for multi head, save ctc neck out for udml
+        if isinstance(x, dict) and "ctc_nect" in x.keys():
+            y["neck_out"] = x["ctc_neck"]
+            y["head_out"] = x
+        elif isinstance(x, dict):
+            y.update(x)
+        else:
+            y["head_out"] = x
+        if self.return_all_feats:
+            if self.training:
+                return y
+            elif isinstance(x, dict):
+                return x
+            else:
+                return {final_name: x}
+        else:
+            return x

mineru/model/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/__init__.py

+# Copyright (c) 2020 PaddlePaddle Authors. 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.
+
+__all__ = ["build_backbone"]
+
+
+def build_backbone(config, model_type):
+    if model_type == "det":
+        from .det_mobilenet_v3 import MobileNetV3
+        from .rec_hgnet import PPHGNet_small
+        from .rec_lcnetv3 import PPLCNetV3
+
+        support_dict = [
+            "MobileNetV3",
+            "ResNet",
+            "ResNet_vd",
+            "ResNet_SAST",
+            "PPLCNetV3",
+            "PPHGNet_small",
+        ]
+    elif model_type == "rec" or model_type == "cls":
+        from .rec_hgnet import PPHGNet_small
+        from .rec_lcnetv3 import PPLCNetV3
+        from .rec_mobilenet_v3 import MobileNetV3
+        from .rec_svtrnet import SVTRNet
+        from .rec_mv1_enhance import MobileNetV1Enhance
+        from .rec_pphgnetv2 import PPHGNetV2_B4
+        support_dict = [
+            "MobileNetV1Enhance",
+            "MobileNetV3",
+            "ResNet",
+            "ResNetFPN",
+            "MTB",
+            "ResNet31",
+            "SVTRNet",
+            "ViTSTR",
+            "DenseNet",
+            "PPLCNetV3",
+            "PPHGNet_small",
+            "PPHGNetV2_B4",
+        ]
+    else:
+        raise NotImplementedError
+
+    module_name = config.pop("name")
+    assert module_name in support_dict, Exception(
+        "when model typs is {}, backbone only support {}".format(
+            model_type, support_dict
+        )
+    )
+    module_class = eval(module_name)(**config)
+    return module_class

mineru/model/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/det_mobilenet_v3.py

+from torch import nn
+
+from ..common import Activation
+
+
+def make_divisible(v, divisor=8, min_value=None):
+    if min_value is None:
+        min_value = divisor
+    new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
+    if new_v < 0.9 * v:
+        new_v += divisor
+    return new_v
+
+
+class ConvBNLayer(nn.Module):
+    def __init__(
+        self,
+        in_channels,
+        out_channels,
+        kernel_size,
+        stride,
+        padding,
+        groups=1,
+        if_act=True,
+        act=None,
+        name=None,
+    ):
+        super(ConvBNLayer, self).__init__()
+        self.if_act = if_act
+        self.conv = nn.Conv2d(
+            in_channels=in_channels,
+            out_channels=out_channels,
+            kernel_size=kernel_size,
+            stride=stride,
+            padding=padding,
+            groups=groups,
+            bias=False,
+        )
+
+        self.bn = nn.BatchNorm2d(
+            out_channels,
+        )
+        if self.if_act:
+            self.act = Activation(act_type=act, inplace=True)
+
+    def forward(self, x):
+        x = self.conv(x)
+        x = self.bn(x)
+        if self.if_act:
+            x = self.act(x)
+        return x
+
+
+class SEModule(nn.Module):
+    def __init__(self, in_channels, reduction=4, name=""):
+        super(SEModule, self).__init__()
+        self.avg_pool = nn.AdaptiveAvgPool2d(1)
+        self.conv1 = nn.Conv2d(
+            in_channels=in_channels,
+            out_channels=in_channels // reduction,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+            bias=True,
+        )
+        self.relu1 = Activation(act_type="relu", inplace=True)
+        self.conv2 = nn.Conv2d(
+            in_channels=in_channels // reduction,
+            out_channels=in_channels,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+            bias=True,
+        )
+        self.hard_sigmoid = Activation(act_type="hard_sigmoid", inplace=True)
+
+    def forward(self, inputs):
+        outputs = self.avg_pool(inputs)
+        outputs = self.conv1(outputs)
+        outputs = self.relu1(outputs)
+        outputs = self.conv2(outputs)
+        outputs = self.hard_sigmoid(outputs)
+        outputs = inputs * outputs
+        return outputs
+
+
+class ResidualUnit(nn.Module):
+    def __init__(
+        self,
+        in_channels,
+        mid_channels,
+        out_channels,
+        kernel_size,
+        stride,
+        use_se,
+        act=None,
+        name="",
+    ):
+        super(ResidualUnit, self).__init__()
+        self.if_shortcut = stride == 1 and in_channels == out_channels
+        self.if_se = use_se
+
+        self.expand_conv = ConvBNLayer(
+            in_channels=in_channels,
+            out_channels=mid_channels,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+            if_act=True,
+            act=act,
+            name=name + "_expand",
+        )
+        self.bottleneck_conv = ConvBNLayer(
+            in_channels=mid_channels,
+            out_channels=mid_channels,
+            kernel_size=kernel_size,
+            stride=stride,
+            padding=int((kernel_size - 1) // 2),
+            groups=mid_channels,
+            if_act=True,
+            act=act,
+            name=name + "_depthwise",
+        )
+        if self.if_se:
+            self.mid_se = SEModule(mid_channels, name=name + "_se")
+        self.linear_conv = ConvBNLayer(
+            in_channels=mid_channels,
+            out_channels=out_channels,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+            if_act=False,
+            act=None,
+            name=name + "_linear",
+        )
+
+    def forward(self, inputs):
+        x = self.expand_conv(inputs)
+        x = self.bottleneck_conv(x)
+        if self.if_se:
+            x = self.mid_se(x)
+        x = self.linear_conv(x)
+        if self.if_shortcut:
+            x = inputs + x
+        return x
+
+
+class MobileNetV3(nn.Module):
+    def __init__(
+        self, in_channels=3, model_name="large", scale=0.5, disable_se=False, **kwargs
+    ):
+        """
+        the MobilenetV3 backbone network for detection module.
+        Args:
+            params(dict): the super parameters for build network
+        """
+        super(MobileNetV3, self).__init__()
+
+        self.disable_se = disable_se
+
+        if model_name == "large":
+            cfg = [
+                # k, exp, c,  se,     nl,  s,
+                [3, 16, 16, False, "relu", 1],
+                [3, 64, 24, False, "relu", 2],
+                [3, 72, 24, False, "relu", 1],
+                [5, 72, 40, True, "relu", 2],
+                [5, 120, 40, True, "relu", 1],
+                [5, 120, 40, True, "relu", 1],
+                [3, 240, 80, False, "hard_swish", 2],
+                [3, 200, 80, False, "hard_swish", 1],
+                [3, 184, 80, False, "hard_swish", 1],
+                [3, 184, 80, False, "hard_swish", 1],
+                [3, 480, 112, True, "hard_swish", 1],
+                [3, 672, 112, True, "hard_swish", 1],
+                [5, 672, 160, True, "hard_swish", 2],
+                [5, 960, 160, True, "hard_swish", 1],
+                [5, 960, 160, True, "hard_swish", 1],
+            ]
+            cls_ch_squeeze = 960
+        elif model_name == "small":
+            cfg = [
+                # k, exp, c,  se,     nl,  s,
+                [3, 16, 16, True, "relu", 2],
+                [3, 72, 24, False, "relu", 2],
+                [3, 88, 24, False, "relu", 1],
+                [5, 96, 40, True, "hard_swish", 2],
+                [5, 240, 40, True, "hard_swish", 1],
+                [5, 240, 40, True, "hard_swish", 1],
+                [5, 120, 48, True, "hard_swish", 1],
+                [5, 144, 48, True, "hard_swish", 1],
+                [5, 288, 96, True, "hard_swish", 2],
+                [5, 576, 96, True, "hard_swish", 1],
+                [5, 576, 96, True, "hard_swish", 1],
+            ]
+            cls_ch_squeeze = 576
+        else:
+            raise NotImplementedError(
+                "mode[" + model_name + "_model] is not implemented!"
+            )
+
+        supported_scale = [0.35, 0.5, 0.75, 1.0, 1.25]
+        assert (
+            scale in supported_scale
+        ), "supported scale are {} but input scale is {}".format(supported_scale, scale)
+        inplanes = 16
+        # conv1
+        self.conv = ConvBNLayer(
+            in_channels=in_channels,
+            out_channels=make_divisible(inplanes * scale),
+            kernel_size=3,
+            stride=2,
+            padding=1,
+            groups=1,
+            if_act=True,
+            act="hard_swish",
+            name="conv1",
+        )
+
+        self.stages = nn.ModuleList()
+        self.out_channels = []
+        block_list = []
+        i = 0
+        inplanes = make_divisible(inplanes * scale)
+        for k, exp, c, se, nl, s in cfg:
+            se = se and not self.disable_se
+            if s == 2 and i > 2:
+                self.out_channels.append(inplanes)
+                self.stages.append(nn.Sequential(*block_list))
+                block_list = []
+            block_list.append(
+                ResidualUnit(
+                    in_channels=inplanes,
+                    mid_channels=make_divisible(scale * exp),
+                    out_channels=make_divisible(scale * c),
+                    kernel_size=k,
+                    stride=s,
+                    use_se=se,
+                    act=nl,
+                    name="conv" + str(i + 2),
+                )
+            )
+            inplanes = make_divisible(scale * c)
+            i += 1
+        block_list.append(
+            ConvBNLayer(
+                in_channels=inplanes,
+                out_channels=make_divisible(scale * cls_ch_squeeze),
+                kernel_size=1,
+                stride=1,
+                padding=0,
+                groups=1,
+                if_act=True,
+                act="hard_swish",
+                name="conv_last",
+            )
+        )
+        self.stages.append(nn.Sequential(*block_list))
+        self.out_channels.append(make_divisible(scale * cls_ch_squeeze))
+        # for i, stage in enumerate(self.stages):
+        #     self.add_sublayer(sublayer=stage, name="stage{}".format(i))
+
+    def forward(self, x):
+        x = self.conv(x)
+        out_list = []
+        for stage in self.stages:
+            x = stage(x)
+            out_list.append(x)
+        return out_list

mineru/model/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_hgnet.py

+import torch
+import torch.nn.functional as F
+from torch import nn
+
+
+class ConvBNAct(nn.Module):
+    def __init__(
+        self, in_channels, out_channels, kernel_size, stride, groups=1, use_act=True
+    ):
+        super().__init__()
+        self.use_act = use_act
+        self.conv = nn.Conv2d(
+            in_channels,
+            out_channels,
+            kernel_size,
+            stride,
+            padding=(kernel_size - 1) // 2,
+            groups=groups,
+            bias=False,
+        )
+        self.bn = nn.BatchNorm2d(out_channels)
+        if self.use_act:
+            self.act = nn.ReLU()
+
+    def forward(self, x):
+        x = self.conv(x)
+        x = self.bn(x)
+        if self.use_act:
+            x = self.act(x)
+        return x
+
+
+class ESEModule(nn.Module):
+    def __init__(self, channels):
+        super().__init__()
+        self.avg_pool = nn.AdaptiveAvgPool2d(1)
+        self.conv = nn.Conv2d(
+            in_channels=channels,
+            out_channels=channels,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+        )
+        self.sigmoid = nn.Sigmoid()
+
+    def forward(self, x):
+        identity = x
+        x = self.avg_pool(x)
+        x = self.conv(x)
+        x = self.sigmoid(x)
+        return x * identity
+
+
+class HG_Block(nn.Module):
+    def __init__(
+        self,
+        in_channels,
+        mid_channels,
+        out_channels,
+        layer_num,
+        identity=False,
+    ):
+        super().__init__()
+        self.identity = identity
+
+        self.layers = nn.ModuleList()
+        self.layers.append(
+            ConvBNAct(
+                in_channels=in_channels,
+                out_channels=mid_channels,
+                kernel_size=3,
+                stride=1,
+            )
+        )
+        for _ in range(layer_num - 1):
+            self.layers.append(
+                ConvBNAct(
+                    in_channels=mid_channels,
+                    out_channels=mid_channels,
+                    kernel_size=3,
+                    stride=1,
+                )
+            )
+
+        # feature aggregation
+        total_channels = in_channels + layer_num * mid_channels
+        self.aggregation_conv = ConvBNAct(
+            in_channels=total_channels,
+            out_channels=out_channels,
+            kernel_size=1,
+            stride=1,
+        )
+        self.att = ESEModule(out_channels)
+
+    def forward(self, x):
+        identity = x
+        output = []
+        output.append(x)
+        for layer in self.layers:
+            x = layer(x)
+            output.append(x)
+        x = torch.cat(output, dim=1)
+        x = self.aggregation_conv(x)
+        x = self.att(x)
+        if self.identity:
+            x += identity
+        return x
+
+
+class HG_Stage(nn.Module):
+    def __init__(
+        self,
+        in_channels,
+        mid_channels,
+        out_channels,
+        block_num,
+        layer_num,
+        downsample=True,
+        stride=[2, 1],
+    ):
+        super().__init__()
+        self.downsample = downsample
+        if downsample:
+            self.downsample = ConvBNAct(
+                in_channels=in_channels,
+                out_channels=in_channels,
+                kernel_size=3,
+                stride=stride,
+                groups=in_channels,
+                use_act=False,
+            )
+
+        blocks_list = []
+        blocks_list.append(
+            HG_Block(in_channels, mid_channels, out_channels, layer_num, identity=False)
+        )
+        for _ in range(block_num - 1):
+            blocks_list.append(
+                HG_Block(
+                    out_channels, mid_channels, out_channels, layer_num, identity=True
+                )
+            )
+        self.blocks = nn.Sequential(*blocks_list)
+
+    def forward(self, x):
+        if self.downsample:
+            x = self.downsample(x)
+        x = self.blocks(x)
+        return x
+
+
+class PPHGNet(nn.Module):
+    """
+    PPHGNet
+    Args:
+        stem_channels: list. Stem channel list of PPHGNet.
+        stage_config: dict. The configuration of each stage of PPHGNet. such as the number of channels, stride, etc.
+        layer_num: int. Number of layers of HG_Block.
+        use_last_conv: boolean. Whether to use a 1x1 convolutional layer before the classification layer.
+        class_expand: int=2048. Number of channels for the last 1x1 convolutional layer.
+        dropout_prob: float. Parameters of dropout, 0.0 means dropout is not used.
+        class_num: int=1000. The number of classes.
+    Returns:
+        model: nn.Layer. Specific PPHGNet model depends on args.
+    """
+
+    def __init__(
+        self,
+        stem_channels,
+        stage_config,
+        layer_num,
+        in_channels=3,
+        det=False,
+        out_indices=None,
+    ):
+        super().__init__()
+        self.det = det
+        self.out_indices = out_indices if out_indices is not None else [0, 1, 2, 3]
+
+        # stem
+        stem_channels.insert(0, in_channels)
+        self.stem = nn.Sequential(
+            *[
+                ConvBNAct(
+                    in_channels=stem_channels[i],
+                    out_channels=stem_channels[i + 1],
+                    kernel_size=3,
+                    stride=2 if i == 0 else 1,
+                )
+                for i in range(len(stem_channels) - 1)
+            ]
+        )
+
+        if self.det:
+            self.pool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
+        # stages
+        self.stages = nn.ModuleList()
+        self.out_channels = []
+        for block_id, k in enumerate(stage_config):
+            (
+                in_channels,
+                mid_channels,
+                out_channels,
+                block_num,
+                downsample,
+                stride,
+            ) = stage_config[k]
+            self.stages.append(
+                HG_Stage(
+                    in_channels,
+                    mid_channels,
+                    out_channels,
+                    block_num,
+                    layer_num,
+                    downsample,
+                    stride,
+                )
+            )
+            if block_id in self.out_indices:
+                self.out_channels.append(out_channels)
+
+        if not self.det:
+            self.out_channels = stage_config["stage4"][2]
+
+        self._init_weights()
+
+    def _init_weights(self):
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                nn.init.kaiming_normal_(m.weight)
+            elif isinstance(m, nn.BatchNorm2d):
+                nn.init.ones_(m.weight)
+                nn.init.zeros_(m.bias)
+            elif isinstance(m, nn.Linear):
+                nn.init.zeros_(m.bias)
+
+    def forward(self, x):
+        x = self.stem(x)
+        if self.det:
+            x = self.pool(x)
+
+        out = []
+        for i, stage in enumerate(self.stages):
+            x = stage(x)
+            if self.det and i in self.out_indices:
+                out.append(x)
+        if self.det:
+            return out
+
+        if self.training:
+            x = F.adaptive_avg_pool2d(x, [1, 40])
+        else:
+            x = F.avg_pool2d(x, [3, 2])
+        return x
+
+
+def PPHGNet_small(pretrained=False, use_ssld=False, det=False, **kwargs):
+    """
+    PPHGNet_small
+    Args:
+        pretrained: bool=False or str. If `True` load pretrained parameters, `False` otherwise.
+                    If str, means the path of the pretrained model.
+        use_ssld: bool=False. Whether using distillation pretrained model when pretrained=True.
+    Returns:
+        model: nn.Layer. Specific `PPHGNet_small` model depends on args.
+    """
+    stage_config_det = {
+        # in_channels, mid_channels, out_channels, blocks, downsample
+        "stage1": [128, 128, 256, 1, False, 2],
+        "stage2": [256, 160, 512, 1, True, 2],
+        "stage3": [512, 192, 768, 2, True, 2],
+        "stage4": [768, 224, 1024, 1, True, 2],
+    }
+
+    stage_config_rec = {
+        # in_channels, mid_channels, out_channels, blocks, downsample
+        "stage1": [128, 128, 256, 1, True, [2, 1]],
+        "stage2": [256, 160, 512, 1, True, [1, 2]],
+        "stage3": [512, 192, 768, 2, True, [2, 1]],
+        "stage4": [768, 224, 1024, 1, True, [2, 1]],
+    }
+
+    model = PPHGNet(
+        stem_channels=[64, 64, 128],
+        stage_config=stage_config_det if det else stage_config_rec,
+        layer_num=6,
+        det=det,
+        **kwargs
+    )
+    return model