LeViT
Last updated
Last updated
The LeViT model was proposed in by Ben Graham, Alaaeldin El-Nouby, Hugo Touvron, Pierre Stock, Armand Joulin, Hervé Jégou, Matthijs Douze. LeViT improves the in performance and efficiency by a few architectural differences such as activation maps with decreasing resolutions in Transformers and the introduction of an attention bias to integrate positional information.
The abstract from the paper is the following:
We design a family of image classification architectures that optimize the trade-off between accuracy and efficiency in a high-speed regime. Our work exploits recent findings in attention-based architectures, which are competitive on highly parallel processing hardware. We revisit principles from the extensive literature on convolutional neural networks to apply them to transformers, in particular activation maps with decreasing resolutions. We also introduce the attention bias, a new way to integrate positional information in vision transformers. As a result, we propose LeVIT: a hybrid neural network for fast inference image classification. We consider different measures of efficiency on different hardware platforms, so as to best reflect a wide range of application scenarios. Our extensive experiments empirically validate our technical choices and show they are suitable to most architectures. Overall, LeViT significantly outperforms existing convnets and vision transformers with respect to the speed/accuracy tradeoff. For example, at 80% ImageNet top-1 accuracy, LeViT is 5 times faster than EfficientNet on CPU.
LeViT Architecture. Taken from the .
Tips:
Compared to ViT, LeViT models use an additional distillation head to effectively learn from a teacher (which, in the LeViT paper, is a ResNet like-model). The distillation head is learned through backpropagation under supervision of a ResNet like-model. They also draw inspiration from convolution neural networks to use activation maps with decreasing resolutions to increase the efficiency.
A list of official BOINC AI and community (indicated by 🌎) resources to help you get started with LeViT.
Image Classification
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.
( image_size = 224num_channels = 3kernel_size = 3stride = 2padding = 1patch_size = 16hidden_sizes = [128, 256, 384]num_attention_heads = [4, 8, 12]depths = [4, 4, 4]key_dim = [16, 16, 16]drop_path_rate = 0mlp_ratio = [2, 2, 2]attention_ratio = [2, 2, 2]initializer_range = 0.02**kwargs )
Parameters
image_size (int, optional, defaults to 224) — The size of the input image.
num_channels (int, optional, defaults to 3) — Number of channels in the input image.
kernel_size (int, optional, defaults to 3) — The kernel size for the initial convolution layers of patch embedding.
stride (int, optional, defaults to 2) — The stride size for the initial convolution layers of patch embedding.
padding (int, optional, defaults to 1) — The padding size for the initial convolution layers of patch embedding.
patch_size (int, optional, defaults to 16) — The patch size for embeddings.
hidden_sizes (List[int], optional, defaults to [128, 256, 384]) — Dimension of each of the encoder blocks.
num_attention_heads (List[int], optional, defaults to [4, 8, 12]) — Number of attention heads for each attention layer in each block of the Transformer encoder.
depths (List[int], optional, defaults to [4, 4, 4]) — The number of layers in each encoder block.
key_dim (List[int], optional, defaults to [16, 16, 16]) — The size of key in each of the encoder blocks.
drop_path_rate (int, optional, defaults to 0) — The dropout probability for stochastic depths, used in the blocks of the Transformer encoder.
mlp_ratios (List[int], optional, defaults to [2, 2, 2]) — Ratio of the size of the hidden layer compared to the size of the input layer of the Mix FFNs in the encoder blocks.
attention_ratios (List[int], optional, defaults to [2, 2, 2]) — Ratio of the size of the output dimension compared to input dimension of attention layers.
initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
Example:
Copied
( *args**kwargs )
__call__
( images**kwargs )
Preprocess an image or a batch of images.
( do_resize: bool = Truesize: typing.Dict[str, int] = Noneresample: Resampling = <Resampling.BICUBIC: 3>do_center_crop: bool = Truecrop_size: typing.Dict[str, int] = Nonedo_rescale: bool = Truerescale_factor: typing.Union[int, float] = 0.00392156862745098do_normalize: bool = Trueimage_mean: typing.Union[float, typing.Iterable[float], NoneType] = [0.485, 0.456, 0.406]image_std: typing.Union[float, typing.Iterable[float], NoneType] = [0.229, 0.224, 0.225]**kwargs )
Parameters
do_resize (bool, optional, defaults to True) — Wwhether to resize the shortest edge of the input to int(256/224 *size). Can be overridden by the do_resize parameter in the preprocess method.
size (Dict[str, int], optional, defaults to {"shortest_edge" -- 224}): Size of the output image after resizing. If size is a dict with keys “width” and “height”, the image will be resized to (size["height"], size["width"]). If size is a dict with key “shortest_edge”, the shortest edge value c is rescaled to int(c * (256/224)). The smaller edge of the image will be matched to this value i.e, if height > width, then image will be rescaled to (size["shortest_egde"] * height / width, size["shortest_egde"]). Can be overridden by the size parameter in the preprocess method.
resample (PILImageResampling, optional, defaults to PILImageResampling.BICUBIC) — Resampling filter to use if resizing the image. Can be overridden by the resample parameter in the preprocess method.
do_center_crop (bool, optional, defaults to True) — Whether or not to center crop the input to (crop_size["height"], crop_size["width"]). Can be overridden by the do_center_crop parameter in the preprocess method.
crop_size (Dict, optional, defaults to {"height" -- 224, "width": 224}): Desired image size after center_crop. Can be overridden by the crop_size parameter in the preprocess method.
do_rescale (bool, optional, defaults to True) — Controls whether to rescale the image by the specified scale rescale_factor. Can be overridden by the do_rescale parameter in the preprocess method.
rescale_factor (int or float, optional, defaults to 1/255) — Scale factor to use if rescaling the image. Can be overridden by the rescale_factor parameter in the preprocess method.
do_normalize (bool, optional, defaults to True) — Controls whether to normalize the image. Can be overridden by the do_normalize parameter in the preprocess method.
image_mean (List[int], defaults to [0.229, 0.224, 0.225]) — Mean to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the image_mean parameter in the preprocess method.
image_std (List[int], defaults to [0.485, 0.456, 0.406]) — Standard deviation to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the image_std parameter in the preprocess method.
Constructs a LeViT image processor.
preprocess
( images: typing.Union[ForwardRef('PIL.Image.Image'), numpy.ndarray, ForwardRef('torch.Tensor'), typing.List[ForwardRef('PIL.Image.Image')], typing.List[numpy.ndarray], typing.List[ForwardRef('torch.Tensor')]]do_resize: typing.Optional[bool] = Nonesize: typing.Union[typing.Dict[str, int], NoneType] = Noneresample: Resampling = Nonedo_center_crop: typing.Optional[bool] = Nonecrop_size: typing.Union[typing.Dict[str, int], NoneType] = Nonedo_rescale: typing.Optional[bool] = Nonerescale_factor: typing.Optional[float] = Nonedo_normalize: typing.Optional[bool] = Noneimage_mean: typing.Union[float, typing.Iterable[float], NoneType] = Noneimage_std: typing.Union[float, typing.Iterable[float], NoneType] = Nonereturn_tensors: typing.Optional[transformers.utils.generic.TensorType] = Nonedata_format: ChannelDimension = <ChannelDimension.FIRST: 'channels_first'>input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None**kwargs )
Parameters
images (ImageInput) — Image or batch of images to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If passing in images with pixel values between 0 and 1, set do_rescale=False.
do_resize (bool, optional, defaults to self.do_resize) — Whether to resize the image.
size (Dict[str, int], optional, defaults to self.size) — Size of the output image after resizing. If size is a dict with keys “width” and “height”, the image will be resized to (height, width). If size is a dict with key “shortest_edge”, the shortest edge value c is rescaled to int(c (256/224)). The smaller edge of the image will be matched to this value i.e, if height > width, then image will be rescaled to (size height / width, size).
resample (PILImageResampling, optional, defaults to PILImageResampling.BICUBIC) — Resampling filter to use when resiizing the image.
do_center_crop (bool, optional, defaults to self.do_center_crop) — Whether to center crop the image.
crop_size (Dict[str, int], optional, defaults to self.crop_size) — Size of the output image after center cropping. Crops images to (crop_size[“height”], crop_size[“width”]).
do_rescale (bool, optional, defaults to self.do_rescale) — Whether to rescale the image pixel values by rescaling_factor - typical to values between 0 and 1.
rescale_factor (float, optional, defaults to self.rescale_factor) — Factor to rescale the image pixel values by.
do_normalize (bool, optional, defaults to self.do_normalize) — Whether to normalize the image pixel values by image_mean and image_std.
image_mean (float or List[float], optional, defaults to self.image_mean) — Mean to normalize the image pixel values by.
image_std (float or List[float], optional, defaults to self.image_std) — Standard deviation to normalize the image pixel values by.
return_tensors (str or TensorType, optional) — The type of tensors to return. Can be one of:
Unset: Return a list of np.ndarray.
TensorType.TENSORFLOW or 'tf': Return a batch of type tf.Tensor.
TensorType.PYTORCH or 'pt': Return a batch of type torch.Tensor.
TensorType.NUMPY or 'np': Return a batch of type np.ndarray.
TensorType.JAX or 'jax': Return a batch of type jax.numpy.ndarray.
data_format (str or ChannelDimension, optional, defaults to ChannelDimension.FIRST) — The channel dimension format for the output image. If unset, the channel dimension format of the input image is used. Can be one of:
"channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
"channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format.
input_data_format (ChannelDimension or str, optional) — The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of:
"channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
"channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format.
"none" or ChannelDimension.NONE: image in (height, width) format.
Preprocess an image or batch of images to be used as input to a LeViT model.
( config )
Parameters
forward
( pixel_values: FloatTensor = Noneoutput_hidden_states: typing.Optional[bool] = Nonereturn_dict: typing.Optional[bool] = None ) → transformers.modeling_outputs.BaseModelOutputWithPoolingAndNoAttention or tuple(torch.FloatTensor)
Parameters
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.
Returns
transformers.modeling_outputs.BaseModelOutputWithPoolingAndNoAttention or tuple(torch.FloatTensor)
last_hidden_state (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Sequence of hidden-states at the output of the last layer of the model.
pooler_output (torch.FloatTensor of shape (batch_size, hidden_size)) — Last layer hidden-state after a pooling operation on the spatial dimensions.
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, num_channels, height, width).
Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
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
Levit Model with an image classification head on top (a linear layer on top of the pooled features), e.g. for ImageNet.
forward
Parameters
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.
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
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 stage) of shape (batch_size, num_channels, height, width). Hidden-states (also called feature maps) of the model at the output of each stage.
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
LeViT Model transformer with image classification heads on top (a linear layer on top of the final hidden state and a linear layer on top of the final hidden state of the distillation token) e.g. for ImageNet. .. warning:: This model supports inference-only. Fine-tuning with distillation (i.e. with a teacher) is not yet supported.
forward
( pixel_values: FloatTensor = Noneoutput_hidden_states: typing.Optional[bool] = Nonereturn_dict: typing.Optional[bool] = None ) → transformers.models.levit.modeling_levit.LevitForImageClassificationWithTeacherOutput or tuple(torch.FloatTensor)
Parameters
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.
Returns
transformers.models.levit.modeling_levit.LevitForImageClassificationWithTeacherOutput or tuple(torch.FloatTensor)
logits (torch.FloatTensor of shape (batch_size, config.num_labels)) — Prediction scores as the average of the cls_logits and distillation_logits.
cls_logits (torch.FloatTensor of shape (batch_size, config.num_labels)) — Prediction scores of the classification head (i.e. the linear layer on top of the final hidden state of the class token).
distillation_logits (torch.FloatTensor of shape (batch_size, config.num_labels)) — Prediction scores of the distillation head (i.e. the linear layer on top of the final hidden state of the distillation token).
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 layer) of shape (batch_size, sequence_length, hidden_size). Hidden-states of the model at the output of each layer plus the initial embedding outputs.
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
There are 2 ways to fine-tune distilled models, either (1) in a classic way, by only placing a prediction head on top of the final hidden state and not using the distillation head, or (2) by placing both a prediction head and distillation head on top of the final hidden state. In that case, the prediction head is trained using regular cross-entropy between the prediction of the head and the ground-truth label, while the distillation prediction head is trained using hard distillation (cross-entropy between the prediction of the distillation head and the label predicted by the teacher). At inference time, one takes the average prediction between both heads as final prediction. (2) is also called “fine-tuning with distillation”, because one relies on a teacher that has already been fine-tuned on the downstream dataset. In terms of models, (1) corresponds to and (2) corresponds to .
All released checkpoints were pre-trained and fine-tuned on (also referred to as ILSVRC 2012, a collection of 1.3 million images and 1,000 classes). only. No external data was used. This is in contrast with the original ViT model, which used external data like the JFT-300M dataset/Imagenet-21k for pre-training.
The authors of LeViT released 5 trained LeViT models, which you can directly plug into or . Techniques like data augmentation, optimization, and regularization were used in order to simulate training on a much larger dataset (while only using ImageNet-1k for pre-training). The 5 variants available are (all trained on images of size 224x224): facebook/levit-128S, facebook/levit-128, facebook/levit-192, facebook/levit-256 and facebook/levit-384. Note that one should use in order to prepare images for the model.
currently supports only inference and not training or fine-tuning.
You can check out demo notebooks regarding inference as well as fine-tuning on custom data (you can just replace by and by or ).
This model was contributed by . The original code can be found .
is supported by this and .
See also:
This is the configuration class to store the configuration of a . It is used to instantiate a LeViT 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 LeViT architecture.
Configuration objects inherit from and can be used to control the model outputs. Read the documentation from for more information.
config () — 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 method to load the model weights.
The bare Levit model outputting raw features without any specific head on top. This model is a PyTorch subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using . See for details.
return_dict (bool, optional) — Whether or not to return a instead of a plain tuple.
A transformers.modeling_outputs.BaseModelOutputWithPoolingAndNoAttention 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 () and inputs.
The forward method, overrides the __call__ special method.
config () — 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 method to load the model weights.
This model is a PyTorch subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
( pixel_values: FloatTensor = Nonelabels: typing.Optional[torch.LongTensor] = Noneoutput_hidden_states: typing.Optional[bool] = Nonereturn_dict: typing.Optional[bool] = None ) → or tuple(torch.FloatTensor)
pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using . See for details.
return_dict (bool, optional) — Whether or not to return a instead of a plain tuple.
or tuple(torch.FloatTensor)
A 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 () and inputs.
The forward method, overrides the __call__ special method.
config () — 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 method to load the model weights.
This model is a PyTorch subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using . See for details.
return_dict (bool, optional) — Whether or not to return a instead of a plain tuple.
A transformers.models.levit.modeling_levit.LevitForImageClassificationWithTeacherOutput 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 () and inputs.
The forward method, overrides the __call__ special method.
