Hugging Face's logo

Transformers documentation

ViTMAE

You are viewing v4.16.2 version. A newer version v4.46.3 is available.
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

ViTMAE

Overview

The ViTMAE model was proposed in Masked Autoencoders Are Scalable Vision Learners by Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Dollár, Ross Girshick. The paper shows that, by pre-training a Vision Transformer (ViT) to reconstruct pixel values for masked patches, one can get results after fine-tuning that outperform supervised pre-training.

The abstract from the paper is the following:

This paper shows that masked autoencoders (MAE) are scalable self-supervised learners for computer vision. Our MAE approach is simple: we mask random patches of the input image and reconstruct the missing pixels. It is based on two core designs. First, we develop an asymmetric encoder-decoder architecture, with an encoder that operates only on the visible subset of patches (without mask tokens), along with a lightweight decoder that reconstructs the original image from the latent representation and mask tokens. Second, we find that masking a high proportion of the input image, e.g., 75%, yields a nontrivial and meaningful self-supervisory task. Coupling these two designs enables us to train large models efficiently and effectively: we accelerate training (by 3x or more) and improve accuracy. Our scalable approach allows for learning high-capacity models that generalize well: e.g., a vanilla ViT-Huge model achieves the best accuracy (87.8%) among methods that use only ImageNet-1K data. Transfer performance in downstream tasks outperforms supervised pre-training and shows promising scaling behavior.

Tips:

  • MAE (masked auto encoding) is a method for self-supervised pre-training of Vision Transformers (ViTs). The pre-training objective is relatively simple: by masking a large portion (75%) of the image patches, the model must reconstruct raw pixel values. One can use ViTMAEForPreTraining for this purpose.
  • A notebook that illustrates how to visualize reconstructed pixel values with ViTMAEForPreTraining can be found here.
  • After pre-training, one “throws away” the decoder used to reconstruct pixels, and one uses the encoder for fine-tuning/linear probing. This means that after fine-tuning, one can directly plug in the weights into a ViTForImageClassification.
  • One can use ViTFeatureExtractor to prepare images for the model. See the code examples for more info.
  • Note that the encoder of MAE is only used to encode the visual patches. The encoded patches are then concatenated with mask tokens, which the decoder (which also consists of Transformer blocks) takes as input. Each mask token is a shared, learned vector that indicates the presence of a missing patch to be predicted. Fixed sin/cos position embeddings are added both to the input of the encoder and the decoder.
drawing MAE architecture. Taken from the original paper.

This model was contributed by nielsr. The original code can be found here.

ViTMAEConfig

class transformers.ViTMAEConfig < >

( hidden_size = 768 num_hidden_layers = 12 num_attention_heads = 12 intermediate_size = 3072 hidden_act = 'gelu' hidden_dropout_prob = 0.0 attention_probs_dropout_prob = 0.0 initializer_range = 0.02 layer_norm_eps = 1e-12 is_encoder_decoder = False image_size = 224 patch_size = 16 num_channels = 3 qkv_bias = True decoder_num_attention_heads = 16 decoder_hidden_size = 512 decoder_num_hidden_layers = 8 decoder_intermediate_size = 2048 mask_ratio = 0.75 norm_pix_loss = False **kwargs )

Parameters

  • hidden_size (int, optional, defaults to 768) — Dimensionality of the encoder layers and the pooler layer.
  • num_hidden_layers (int, optional, defaults to 12) — Number of hidden layers in the Transformer encoder.
  • num_attention_heads (int, optional, defaults to 12) — Number of attention heads for each attention layer in the Transformer encoder.
  • intermediate_size (int, optional, defaults to 3072) — Dimensionality of the “intermediate” (i.e., feed-forward) layer in the Transformer encoder.
  • hidden_act (str or function, optional, defaults to "gelu") — The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "selu" and "gelu_new" are supported.
  • hidden_dropout_prob (float, optional, defaults to 0.0) — The dropout probabilitiy for all fully connected layers in the embeddings, encoder, and pooler.
  • attention_probs_dropout_prob (float, optional, defaults to 0.0) — The dropout ratio for the attention probabilities.
  • 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.
  • image_size (int, optional, defaults to 224) — The size (resolution) of each image.
  • patch_size (int, optional, defaults to 16) — The size (resolution) of each patch.
  • num_channels (int, optional, defaults to 3) — The number of input channels.
  • qkv_bias (bool, optional, defaults to True) — Whether to add a bias to the queries, keys and values.
  • decoder_num_attention_heads (int, optional, defaults to 12) — Number of attention heads for each attention layer in the decoder.
  • decoder_hidden_size (int, optional, defaults to 512) — Dimensionality of the decoder.
  • decoder_num_hidden_layers (int, optional, defaults to 8) — Number of hidden layers in the decoder.
  • decoder_intermediate_size (int, optional, defaults to 2048) — Dimensionality of the “intermediate” (i.e., feed-forward) layer in the decoder.
  • mask_ratio (float, optional, defaults to 0.75) — The ratio of the number of masked tokens in the input sequence.
  • norm_pix_loss (bool, optional, defaults to False) — Whether or not to train with normalized pixels (see Table 3 in the paper). Using normalized pixels improved representation quality in the experiments of the authors.

This is the configuration class to store the configuration of a ViTMAEModel. It is used to instantiate an ViT MAE 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 ViT facebook/vit-mae-base architecture.

Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.

Example:

>>> from transformers import ViTMAEModel, ViTMAEConfig

>>> # Initializing a ViT MAE vit-mae-base style configuration
>>> configuration = ViTMAEConfig()

>>> # Initializing a model from the vit-mae-base style configuration
>>> model = ViTMAEModel(configuration)

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

ViTMAEModel

class transformers.ViTMAEModel < >

( config )

Parameters

  • config (ViTMAEConfig) — 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 ViTMAE Model transformer outputting raw hidden-states without any specific head on top. This model is a PyTorch 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.

forward < >

( pixel_values = None head_mask = None output_attentions = None output_hidden_states = None return_dict = None ) → transformers.models.vit_mae.modeling_vit_mae.ViTMAEModelOutputor tuple(torch.FloatTensor)

Parameters

  • pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using AutoFeatureExtractor. See AutoFeatureExtractor.__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.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.

Returns

transformers.models.vit_mae.modeling_vit_mae.ViTMAEModelOutputor tuple(torch.FloatTensor)

A transformers.models.vit_mae.modeling_vit_mae.ViTMAEModelOutputor a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (ViTMAEConfig) 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.
  • mask (torch.FloatTensor of shape (batch_size, sequence_length)) — Tensor indicating which patches are masked (1) and which are not (0).
  • ids_restore (torch.LongTensor of shape (batch_size, sequence_length)) — Tensor containing the original index of the (shuffled) masked patches.
  • 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.
  • 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.

The ViTMAEModel 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.

Examples:

>>> from transformers import AutoFeatureExtractor, ViTMAEModel
>>> from PIL import Image
>>> import requests

>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
>>> image = Image.open(requests.get(url, stream=True).raw)

>>> feature_extractor = AutoFeatureExtractor.from_pretrained("facebook/vit-mae-base")
>>> model = ViTMAEModel.from_pretrained("facebook/vit-mae-base")

>>> inputs = feature_extractor(images=image, return_tensors="pt")
>>> outputs = model(**inputs)
>>> last_hidden_states = outputs.last_hidden_state

ViTMAEForPreTraining

class transformers.ViTMAEForPreTraining < >

( config )

Parameters

  • config (ViTMAEConfig) — 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 ViTMAE Model transformer with the decoder on top for self-supervised pre-training. This model is a PyTorch 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.

forward < >

( pixel_values = None head_mask = None output_attentions = None output_hidden_states = None return_dict = None ) → transformers.models.vit_mae.modeling_vit_mae.ViTMAEForPreTrainingOutputor tuple(torch.FloatTensor)

Parameters

  • pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using AutoFeatureExtractor. See AutoFeatureExtractor.__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.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.

Returns

transformers.models.vit_mae.modeling_vit_mae.ViTMAEForPreTrainingOutputor tuple(torch.FloatTensor)

A transformers.models.vit_mae.modeling_vit_mae.ViTMAEForPreTrainingOutputor a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (ViTMAEConfig) and inputs.

  • loss (torch.FloatTensor of shape (1,)) — Pixel reconstruction loss.
  • logits (torch.FloatTensor of shape (batch_size, patch_size ** 2 * num_channels)) — Pixel reconstruction logits.
  • mask (torch.FloatTensor of shape (batch_size, sequence_length)) — Tensor indicating which patches are masked (1) and which are not (0).
  • ids_restore (torch.LongTensor of shape (batch_size, sequence_length)) — Tensor containing the original index of the (shuffled) masked patches.
  • 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.
  • 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.

The ViTMAEForPreTraining 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.

Examples:

>>> from transformers import AutoFeatureExtractor, ViTMAEForPreTraining
>>> from PIL import Image
>>> import requests

>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
>>> image = Image.open(requests.get(url, stream=True).raw)

>>> feature_extractor = AutoFeatureExtractor.from_pretrained("facebook/vit-mae-base")
>>> model = ViTMAEForPreTraining.from_pretrained("facebook/vit-mae-base")

>>> inputs = feature_extractor(images=image, return_tensors="pt")
>>> outputs = model(**inputs)
>>> loss = outputs.loss
>>> mask = outputs.mask
>>> ids_restore = outputs.ids_restore
←Vision Transformer (ViT) VisualBERT→