# Dilated Neighborhood Attention Transformer

## Overview

DiNAT は [Dilated Neighborhood Attender Transformer](https://huggingface.co/papers/2209.15001) で提案されました。
Ali Hassani and Humphrey Shi.

[NAT](nat) を拡張するために、拡張近隣アテンション パターンを追加してグローバル コンテキストをキャプチャします。
そしてそれと比較して大幅なパフォーマンスの向上が見られます。

論文の要約は次のとおりです。

*トランスフォーマーは急速に、さまざまなモダリティにわたって最も頻繁に適用される深層学習アーキテクチャの 1 つになりつつあります。
ドメインとタスク。ビジョンでは、単純なトランスフォーマーへの継続的な取り組みに加えて、階層型トランスフォーマーが
また、そのパフォーマンスと既存のフレームワークへの簡単な統合のおかげで、大きな注目を集めました。
これらのモデルは通常、スライディング ウィンドウの近隣アテンション (NA) などの局所的な注意メカニズムを採用しています。
または Swin Transformer のシフト ウィンドウ セルフ アテンション。自己注意の二次複雑さを軽減するのに効果的ですが、
局所的な注意は、自己注意の最も望ましい 2 つの特性を弱めます。それは、長距離の相互依存性モデリングです。
そして全体的な受容野。このペーパーでは、自然で柔軟で、
NA への効率的な拡張により、よりグローバルなコンテキストを捕捉し、受容野をゼロから指数関数的に拡張することができます。
追加費用。 NA のローカルな注目と DiNA のまばらなグローバルな注目は相互に補完し合うため、私たちは
両方に基づいて構築された新しい階層型ビジョン トランスフォーマーである Dilated Neighborhood Attendant Transformer (DiNAT) を導入します。
DiNAT のバリアントは、NAT、Swin、ConvNeXt などの強力なベースラインに比べて大幅に改善されています。
私たちの大規模モデルは、COCO オブジェクト検出において Swin モデルよりも高速で、ボックス AP が 1.5% 優れています。
COCO インスタンス セグメンテーションでは 1.3% のマスク AP、ADE20K セマンティック セグメンテーションでは 1.1% の mIoU。
新しいフレームワークと組み合わせた当社の大規模バリアントは、COCO (58.2 PQ) 上の新しい最先端のパノプティック セグメンテーション モデルです。
および ADE20K (48.5 PQ)、および Cityscapes (44.5 AP) および ADE20K (35.4 AP) のインスタンス セグメンテーション モデル (追加データなし)。
また、ADE20K (58.2 mIoU) 上の最先端の特殊なセマンティック セグメンテーション モデルとも一致します。
都市景観 (84.5 mIoU) では 2 位にランクされています (追加データなし)。 *

<img
src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/dilated-neighborhood-attention-pattern.jpg"
alt="drawing" width="600"/>

 異なる拡張値を使用した近隣アテンション。
元の論文から抜粋。

このモデルは [Ali Hassani](https://huggingface.co/alihassanijr) によって提供されました。
元のコードは [ここ](https://github.com/SHI-Labs/Neighborhood-Attendance-Transformer) にあります。

## Usage tips

DiNAT は *バックボーン* として使用できます。 「output_hidden_​​states = True」の場合、
`hidden_​​states` と `reshaped_hidden_​​states` の両方を出力します。 `reshape_hidden_​​states` は、`(batch_size, height, width, num_channels)` ではなく、`(batch, num_channels, height, width)` の形状を持っています。

ノート：
- DiNAT は、[NATTEN](https://github.com/SHI-Labs/NATTEN/) による近隣アテンションと拡張近隣アテンションの実装に依存しています。
[shi-labs.com/natten](https://shi-labs.com/natten) を参照して、Linux 用のビルド済みホイールを使用してインストールするか、`pip install natten` を実行してシステム上に構築できます。
後者はコンパイルに時間がかかる可能性があることに注意してください。 NATTEN はまだ Windows デバイスをサポートしていません。
- 現時点ではパッチ サイズ 4 のみがサポートされています。

## Resources

DiNAT の使用を開始するのに役立つ公式 Hugging Face およびコミュニティ (🌎 で示されている) リソースのリスト。

- [DinatForImageClassification](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatForImageClassification) は、この [サンプル スクリプト](https://github.com/huggingface/transformers/tree/main/examples/pytorch/image-classification) および [ノートブック](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/image_classification.ipynb)。
- 参照: [画像分類タスク ガイド](../tasks/image_classification)

ここに含めるリソースの送信に興味がある場合は、お気軽にプル リクエストを開いてください。審査させていただきます。リソースは、既存のリソースを複製するのではなく、何か新しいものを示すことが理想的です。

## DinatConfig[[transformers.DinatConfig]]

#### transformers.DinatConfig[[transformers.DinatConfig]]

[Source](https://github.com/huggingface/transformers/blob/v5.7.0/src/transformers/models/dinat/configuration_dinat.py#L25)

This is the configuration class to store the configuration of a DinatModel. It is used to instantiate a Dinat
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 [shi-labs/dinat-mini-in1k-224](https://huggingface.co/shi-labs/dinat-mini-in1k-224)

Configuration objects inherit from [PreTrainedConfig](/docs/transformers/v5.7.0/ja/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the
documentation from [PreTrainedConfig](/docs/transformers/v5.7.0/ja/main_classes/configuration#transformers.PreTrainedConfig) for more information.

Example:

```python
>>> from transformers import DinatConfig, DinatModel

>>> # Initializing a Dinat shi-labs/dinat-mini-in1k-224 style configuration
>>> configuration = DinatConfig()

>>> # Initializing a model (with random weights) from the shi-labs/dinat-mini-in1k-224 style configuration
>>> model = DinatModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config
```

**Parameters:**

patch_size (`Union[int, list[int], tuple[int, 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 `64`) : Dimensionality of the embeddings and hidden states.

depths (`Union[list[int], tuple[int, ...]]`, *optional*, defaults to `(3, 4, 6, 5)`) : Depth of each layer in the Transformer.

num_heads (`Union[list[int], tuple[int, ...]]`, *optional*, defaults to `(2, 4, 8, 16)`) : Number of attention heads for each attention layer in the Transformer decoder.

kernel_size (`int`, *optional*, defaults to `7`) : The size of the convolutional kernel.

dilations (`list[list[int]]`, *optional*, defaults to `[[1, 8, 1], [1, 4, 1, 4], [1, 2, 1, 2, 1, 2], [1, 1, 1, 1, 1]]`) : Dilation value of each NA layer in the Transformer encoder.

mlp_ratio (`float`, *optional*, defaults to `3.0`) : Ratio of the MLP hidden dim to the embedding dim.

qkv_bias (`bool`, *optional*, defaults to `True`) : Whether to add a bias to the queries, keys and values.

hidden_dropout_prob (`Union[float, int]`, *optional*, defaults to `0.0`) : The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.

attention_probs_dropout_prob (`Union[float, int]`, *optional*, defaults to `0.0`) : The dropout ratio for the attention probabilities.

drop_path_rate (`Union[float, int]`, *optional*, defaults to `0.1`) : Drop path rate for the patch fusion.

hidden_act (`str`, *optional*, defaults to `gelu`) : The non-linear activation function (function or string) in the decoder. For example, `"gelu"`, `"relu"`, `"silu"`, etc.

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.

layer_scale_init_value (`float`, *optional*, defaults to `0.0`) : Scale to use in the self-attention layers. 0.1 for base, 1e-6 for large. Set 0 to disable layer scale.

## DinatModel[[transformers.DinatModel]]

#### transformers.DinatModel[[transformers.DinatModel]]

[Source](https://github.com/huggingface/transformers/blob/v5.7.0/src/transformers/models/dinat/modeling_dinat.py#L554)

The bare Dinat Model outputting raw hidden-states without any specific head on top.

This model inherits from [PreTrainedModel](/docs/transformers/v5.7.0/ja/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

forwardtransformers.DinatModel.forwardhttps://github.com/huggingface/transformers/blob/v5.7.0/src/transformers/models/dinat/modeling_dinat.py#L580[{"name": "pixel_values", "val": ": torch.FloatTensor | None = None"}, {"name": "output_attentions", "val": ": bool | None = None"}, {"name": "output_hidden_states", "val": ": bool | None = None"}, {"name": "return_dict", "val": ": bool | None = None"}, {"name": "**kwargs", "val": ""}]- **pixel_values** (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`, *optional*) --
  The tensors corresponding to the input images. Pixel values can be obtained using
  `ViTImageProcessor`. See `ViTImageProcessor.__call__()` for details (`processor_class` uses
  `ViTImageProcessor` for processing images).
- **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 [ModelOutput](/docs/transformers/v5.7.0/ja/main_classes/output#transformers.utils.ModelOutput) instead of a plain tuple.0`DinatModelOutput` or `tuple(torch.FloatTensor)`A `DinatModelOutput` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([DinatConfig](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatConfig)) and inputs.
The [DinatModel](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatModel) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **last_hidden_state** (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*, defaults to `None`) -- 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, 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.
- **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.

Example:

```python
```

**Parameters:**

config ([DinatModel](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatModel)) : 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 [from_pretrained()](/docs/transformers/v5.7.0/ja/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

add_pooling_layer (`bool`, *optional*, defaults to `True`) : Whether to add a pooling layer

**Returns:**

``DinatModelOutput` or `tuple(torch.FloatTensor)``

A `DinatModelOutput` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([DinatConfig](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatConfig)) and inputs.

## DinatForImageClassification[[transformers.DinatForImageClassification]]

#### transformers.DinatForImageClassification[[transformers.DinatForImageClassification]]

[Source](https://github.com/huggingface/transformers/blob/v5.7.0/src/transformers/models/dinat/modeling_dinat.py#L635)

Dinat Model transformer with an image classification head on top (a linear layer on top of the final hidden state
of the [CLS] token) e.g. for ImageNet.

This model inherits from [PreTrainedModel](/docs/transformers/v5.7.0/ja/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

forwardtransformers.DinatForImageClassification.forwardhttps://github.com/huggingface/transformers/blob/v5.7.0/src/transformers/models/dinat/modeling_dinat.py#L652[{"name": "pixel_values", "val": ": torch.FloatTensor | None = None"}, {"name": "labels", "val": ": torch.LongTensor | None = None"}, {"name": "output_attentions", "val": ": bool | None = None"}, {"name": "output_hidden_states", "val": ": bool | None = None"}, {"name": "return_dict", "val": ": bool | None = None"}, {"name": "**kwargs", "val": ""}]- **pixel_values** (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`, *optional*) --
  The tensors corresponding to the input images. Pixel values can be obtained using
  `ViTImageProcessor`. See `ViTImageProcessor.__call__()` for details (`processor_class` uses
  `ViTImageProcessor` for processing images).
- **labels** (`torch.LongTensor` of shape `(batch_size,)`, *optional*) --
  Labels for computing the image classification/regression loss. Indices should be in `[0, ...,
  config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
  `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
- **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 [ModelOutput](/docs/transformers/v5.7.0/ja/main_classes/output#transformers.utils.ModelOutput) instead of a plain tuple.0`DinatImageClassifierOutput` or `tuple(torch.FloatTensor)`A `DinatImageClassifierOutput` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([DinatConfig](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatConfig)) and inputs.
The [DinatForImageClassification](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatForImageClassification) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Classification (or regression if config.num_labels==1) loss.
- **logits** (`torch.FloatTensor` of shape `(batch_size, config.num_labels)`) -- Classification (or regression if config.num_labels==1) scores (before SoftMax).
- **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.
- **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.

Example:

```python
>>> from transformers import AutoImageProcessor, DinatForImageClassification
>>> import torch
>>> from datasets import load_dataset

>>> dataset = load_dataset("huggingface/cats-image")
>>> image = dataset["test"]["image"][0]

>>> image_processor = AutoImageProcessor.from_pretrained("shi-labs/dinat-mini-in1k-224")
>>> model = DinatForImageClassification.from_pretrained("shi-labs/dinat-mini-in1k-224")

>>> inputs = image_processor(image, return_tensors="pt")

>>> with torch.no_grad():
...     logits = model(**inputs).logits

>>> # model predicts one of the 1000 ImageNet classes
>>> predicted_label = logits.argmax(-1).item()
>>> print(model.config.id2label[predicted_label])
...
```

**Parameters:**

config ([DinatForImageClassification](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatForImageClassification)) : 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 [from_pretrained()](/docs/transformers/v5.7.0/ja/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

**Returns:**

``DinatImageClassifierOutput` or `tuple(torch.FloatTensor)``

A `DinatImageClassifierOutput` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([DinatConfig](/docs/transformers/v5.7.0/ja/model_doc/dinat#transformers.DinatConfig)) and inputs.