gt4sd.frameworks.granular.ml.models.vae_mlp.core module¶
VaeMlp implementation.
Summary¶
Classes:
VaeMlp - variational encoder using MLP with Gaussian prior and approximate posterior. |
Reference¶
- class VaeMlp(name, position, data, input_size_enc=256, hidden_size_enc=256, n_layers_enc=2, activation_enc='linear', dropout_enc=0.0, hidden_size_dec=256, n_layers_dec=2, activation_dec='linear', dropout_dec=0.0, output_size_dec=256, latent_size=196, loss_function='mse', kl_low=0.0, kl_high=0.1, kl_n_epochs=100, kl_start_epoch=0, **kwargs)[source]¶
Bases:
GranularEncoderDecoderModelVaeMlp - variational encoder using MLP with Gaussian prior and approximate posterior.
- __init__(name, position, data, input_size_enc=256, hidden_size_enc=256, n_layers_enc=2, activation_enc='linear', dropout_enc=0.0, hidden_size_dec=256, n_layers_dec=2, activation_dec='linear', dropout_dec=0.0, output_size_dec=256, latent_size=196, loss_function='mse', kl_low=0.0, kl_high=0.1, kl_n_epochs=100, kl_start_epoch=0, **kwargs)[source]¶
Construct VaeMlp.
- Parameters
name (
str) – model name.position (
int) – position of the model.data (
Dict[str,str]) – data name mappings.input_size_enc (
int) – encoder input size. Defaults to 256.hidden_size_enc (
int) – encoder hidden size. Defaults to 256.n_layers_enc (
int) – number of layers for the encoder. Defaults to 2.activation_enc (
str) – activation function for the encoder. Defaults to “linear”.dropout_enc (
float) – encoder dropout rate. Defaults to 0.0.hidden_size_dec (
int) – decoder hidden size. Defaults to 256.n_layers_dec (
int) – number of layers for the decoder. Defaults to 2.activation_dec (
str) – activation function for the decoder. Defaults to “linear”.dropout_dec (
float) – decoder dropout rate. Defaults to 0.0.output_size_dec (
int) – decoder output size. Defaults to 256.latent_size (
int) – size of the latent space. Defaults to 196.loss_function (
str) – loss function. Defaults to “mse”.kl_low (
float) – low KL weight.kl_high (
float) – high KL weight.kl_n_epochs (
int) – KL number of epochs.kl_start_epoch (
int) – KL starting epoch.
- Raises
ValueError – in case the provided loss function is not supported.
- decode(z, *args, **kwargs)[source]¶
Decode a latent space point.
- Parameters
z (
Any) – latent point.- Return type
Any- Returns
decoded sample.
- _sampling_step(x, *args, **kwargs)[source]¶
Run a sampling step in the model.
- Parameters
x (
Any) – model input.- Return type
Any- Returns
model sampling step output.
- encode(x, *args, **kwargs)[source]¶
Encode a sample.
- Parameters
x (
Any) – input sample.- Return type
Any- Returns
latent encoding.
- _run_step(x, *args, **kwargs)[source]¶
Run a step in the model.
- Parameters
x (
Any) – model input.- Return type
Any- Returns
model step output.
- step(input_data, target_data, device='cpu', current_epoch=0, *args, **kwargs)[source]¶
Training step for the model.
- Parameters
input_data (
Any) – input for the step.target_data (
Any) – target for the step.device (
str) – string representing the device to use. Defaults to “cpu”.current_epoch (
int) – current epoch. Defaults to 0.
- Return type
Tuple[Any,Any,Any]- Returns
a tuple containing the step output, the loss and the logs for the module.
- __annotations__ = {}¶
- __doc__ = 'VaeMlp - variational encoder using MLP with Gaussian prior and approximate posterior.'¶
- __module__ = 'gt4sd.frameworks.granular.ml.models.vae_mlp.core'¶
- val_step(input_data, target_data, device='cpu', current_epoch=0, *args, **kwargs)[source]¶
Validation step for the model.
- Parameters
input_data (
Any) – input for the step.target_data (
Any) – target for the step.device (
str) – string representing the device to use. Defaults to “cpu”.current_epoch (
int) – current epoch. Defaults to 0.
- Return type
Any- Returns
a tuple containing the step output, the loss and the logs for the module.