DiNAT
Last updated
Last updated
DiNAT was proposed in Dilated Neighborhood Attention Transformer by Ali Hassani and Humphrey Shi.
It extends NAT by adding a Dilated Neighborhood Attention pattern to capture global context, and shows significant performance improvements over it.
The abstract from the paper is the following:
Transformers are quickly becoming one of the most heavily applied deep learning architectures across modalities, domains, and tasks. In vision, on top of ongoing efforts into plain transformers, hierarchical transformers have also gained significant attention, thanks to their performance and easy integration into existing frameworks. These models typically employ localized attention mechanisms, such as the sliding-window Neighborhood Attention (NA) or Swin Transformer’s Shifted Window Self Attention. While effective at reducing self attention’s quadratic complexity, local attention weakens two of the most desirable properties of self attention: long range inter-dependency modeling, and global receptive field. In this paper, we introduce Dilated Neighborhood Attention (DiNA), a natural, flexible and efficient extension to NA that can capture more global context and expand receptive fields exponentially at no additional cost. NA’s local attention and DiNA’s sparse global attention complement each other, and therefore we introduce Dilated Neighborhood Attention Transformer (DiNAT), a new hierarchical vision transformer built upon both. DiNAT variants enjoy significant improvements over strong baselines such as NAT, Swin, and ConvNeXt. Our large model is faster and ahead of its Swin counterpart by 1.5% box AP in COCO object detection, 1.3% mask AP in COCO instance segmentation, and 1.1% mIoU in ADE20K semantic segmentation. Paired with new frameworks, our large variant is the new state of the art panoptic segmentation model on COCO (58.2 PQ) and ADE20K (48.5 PQ), and instance segmentation model on Cityscapes (44.5 AP) and ADE20K (35.4 AP) (no extra data). It also matches the state of the art specialized semantic segmentation models on ADE20K (58.2 mIoU), and ranks second on Cityscapes (84.5 mIoU) (no extra data).
Tips:
One can use the AutoImageProcessor API to prepare images for the model.
DiNAT can be used as a backbone. When output_hidden_states = True, it will output both hidden_states and reshaped_hidden_states. The reshaped_hidden_states have a shape of (batch, num_channels, height, width) rather than (batch_size, height, width, num_channels).
Notes:
DiNAT depends on NATTEN’s implementation of Neighborhood Attention and Dilated Neighborhood Attention. You can install it with pre-built wheels for Linux by referring to shi-labs.com/natten, or build on your system by running pip install natten. Note that the latter will likely take time to compile. NATTEN does not support Windows devices yet.
Patch size of 4 is only supported at the moment.
Neighborhood Attention with different dilation values. Taken from the original paper.
This model was contributed by Ali Hassani. The original code can be found here.
A list of official BOINC AI and community (indicated by 🌎) resources to help you get started with DiNAT.
Image Classification
DinatForImageClassification is supported by this example script and notebook.
See also: Image classification task guide
If you’re interested in submitting a resource to be included here, please feel free to open a Pull Request and we’ll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
( patch_size = 4num_channels = 3embed_dim = 64depths = [3, 4, 6, 5]num_heads = [2, 4, 8, 16]kernel_size = 7dilations = [[1, 8, 1], [1, 4, 1, 4], [1, 2, 1, 2, 1, 2], [1, 1, 1, 1, 1]]mlp_ratio = 3.0qkv_bias = Truehidden_dropout_prob = 0.0attention_probs_dropout_prob = 0.0drop_path_rate = 0.1hidden_act = 'gelu'initializer_range = 0.02layer_norm_eps = 1e-05layer_scale_init_value = 0.0out_features = Noneout_indices = None**kwargs )
Parameters
patch_size (int, optional, defaults to 4) — The size (resolution) of each patch. NOTE: Only patch size of 4 is supported at the moment.
num_channels (int, optional, defaults to 3) — The number of input channels.
embed_dim (int, optional, defaults to 64) — Dimensionality of patch embedding.
depths (List[int], optional, defaults to [2, 2, 6, 2]) — Number of layers in each level of the encoder.
num_heads (List[int], optional, defaults to [3, 6, 12, 24]) — Number of attention heads in each layer of the Transformer encoder.
kernel_size (int, optional, defaults to 7) — Neighborhood Attention kernel size.
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 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.
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-12) — The epsilon used by the layer normalization layers.
layer_scale_init_value (float, optional, defaults to 0.0) — The initial value for the layer scale. Disabled if <=0.
out_features (List[str], optional) — If used as backbone, list of features to output. Can be any of "stem", "stage1", "stage2", etc. (depending on how many stages the model has). If unset and out_indices is set, will default to the corresponding stages. If unset and out_indices is unset, will default to the last stage.
out_indices (List[int], optional) — If used as backbone, list of indices of features to output. Can be any of 0, 1, 2, etc. (depending on how many stages the model has). If unset and out_features is set, will default to the corresponding stages. If unset and out_features is unset, will default to the last stage.
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 Dinat shi-labs/dinat-mini-in1k-224 architecture.
Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.
Example:
Copied
( configadd_pooling_layer = True )
Parameters
config (DinatConfig) — 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() method to load the model weights.
The bare Dinat Model transformer outputting raw hidden-states without any specific head on top. This model is a PyTorch 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.
forward
( pixel_values: typing.Optional[torch.FloatTensor] = Noneoutput_attentions: typing.Optional[bool] = Noneoutput_hidden_states: typing.Optional[bool] = Nonereturn_dict: typing.Optional[bool] = None ) → transformers.models.dinat.modeling_dinat.DinatModelOutput or tuple(torch.FloatTensor)
Parameters
pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using AutoImageProcessor. See ViTImageProcessor.call() for details.
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 instead of a plain tuple.
Returns
transformers.models.dinat.modeling_dinat.DinatModelOutput or tuple(torch.FloatTensor)
A transformers.models.dinat.modeling_dinat.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) and inputs.
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.
The 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.
Example:
Copied
( config )
Parameters
config (DinatConfig) — 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() method to load the model weights.
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 is a PyTorch 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.
forward
( pixel_values: typing.Optional[torch.FloatTensor] = Nonelabels: typing.Optional[torch.LongTensor] = Noneoutput_attentions: typing.Optional[bool] = Noneoutput_hidden_states: typing.Optional[bool] = Nonereturn_dict: typing.Optional[bool] = None ) → transformers.models.dinat.modeling_dinat.DinatImageClassifierOutput or tuple(torch.FloatTensor)
Parameters
pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using AutoImageProcessor. See ViTImageProcessor.call() for details.
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 instead of a plain tuple.
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).
Returns
transformers.models.dinat.modeling_dinat.DinatImageClassifierOutput or tuple(torch.FloatTensor)
A transformers.models.dinat.modeling_dinat.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) and inputs.
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 + 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.
The 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.
Example:
Copied
