Encoders
- class multimodal_compare.models.encoders.Enc_CNN(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_FNN(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_MNIST(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_MNIST2(latent_dim, data_dim, latent_private, num_hidden_layers=1)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_PolyMNIST(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_SVHN(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_SVHN2(latent_dim, data_dim, latent_private)
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_Transformer(latent_dim, data_dim, latent_private, ff_size=1024, num_layers=8, num_heads=2, dropout=0.1, activation='gelu')
Bases:
VaeEncoderTransformer VAE as implemented in https://github.com/Mathux/ACTOR
- _is_full_backward_hook: bool | None
- forward(batch)
Forward pass
- Parameters:
batch (list, torch.tensor) – list of a data batch and boolean masks for the sequences
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_TransformerIMG(latent_dim, data_dim, latent_private, ff_size=1024, num_layers=8, num_heads=4, dropout=0.1, activation='gelu')
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(batch)
Forward pass
- Parameters:
batch (list, torch.tensor) – list of a data batch and boolean masks for the sequences
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_TxtTransformer(latent_dim, data_dim, latent_private, ff_size=1024, num_layers=8, num_heads=2, dropout=0.1, activation='gelu')
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(batch)
Forward pass
- Parameters:
batch (list, torch.tensor) – list of a data batch and boolean masks for the sequences
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.Enc_VideoGPT(latent_dim, data_dim, latent_private, n_res_layers=4, downsample=(2, 4, 4))
Bases:
VaeEncoder- _is_full_backward_hook: bool | None
- forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.VaeComponent(latent_dim: int, data_dim: tuple, latent_private=None, net_type=NetworkTypes.UNSPECIFIED, net_role=NetworkRoles.UNSPECIFIED)
Bases:
Module- _is_full_backward_hook: bool | None
- abstract forward(x)
Forward pass
- Parameters:
x (list, torch.tensor) – data batch
- Returns:
tensor of means, tensor of log variances
- Return type:
tuple(torch.tensor, torch.tensor)
- training: bool
- class multimodal_compare.models.encoders.VaeEncoder(latent_dim, data_dim, latent_private, net_type: NetworkTypes)
Bases:
VaeComponent- _backward_hooks: Dict[int, Callable]
- _buffers: Dict[str, Tensor | None]
- _forward_hooks: Dict[int, Callable]
- _forward_pre_hooks: Dict[int, Callable]
- _is_full_backward_hook: bool | None
- _load_state_dict_post_hooks: Dict[int, Callable]
- _load_state_dict_pre_hooks: Dict[int, Callable]
- _modules: Dict[str, 'Module' | None]
- _non_persistent_buffers_set: Set[str]
- _parameters: Dict[str, Parameter | None]
- _state_dict_hooks: Dict[int, Callable]
- training: bool