PyTorchDiscriminator
class PyTorchDiscriminator(n_features=1, n_out=1)
Discriminator based on PyTorch
Parameters
- n_features (
int) – Dimension of input data vector. - n_out (
int) – Dimension of the discriminator’s output vector.
Raises
NameError – Pytorch not installed
Attributes
discriminator_net
Get discriminator
Returns
discriminator object
Return type
object
Methods
get_label
PyTorchDiscriminator.get_label(x, detach=False)
Get data sample labels, i.e. true or fake.
Parameters
- x (Union(numpy.ndarray, torch.Tensor)) – Discriminator input, i.e. data sample.
- detach (bool) – if None detach from torch tensor variable (optional)
Returns
Discriminator output, i.e. data label
Return type
torch.Tensor
gradient_penalty
PyTorchDiscriminator.gradient_penalty(x, lambda_=5.0, k=0.01, c=1.0)
Compute gradient penalty for discriminator optimization
Parameters
- x (numpy.ndarray) – Generated data sample.
- lambda (float) – Gradient penalty coefficient 1.
- k (float) – Gradient penalty coefficient 2.
- c (float) – Gradient penalty coefficient 3.
Returns
Gradient penalty.
Return type
torch.Tensor
load_model
PyTorchDiscriminator.load_model(load_dir)
Load discriminator model
Parameters
load_dir (str) – file with stored pytorch discriminator model to be loaded
loss
PyTorchDiscriminator.loss(x, y, weights=None)
Loss function
Parameters
- x (torch.Tensor) – Discriminator output.
- y (torch.Tensor) – Label of the data point
- weights (torch.Tensor) – Data weights.
Returns
Loss w.r.t to the generated data points.
Return type
torch.Tensor
save_model
PyTorchDiscriminator.save_model(snapshot_dir)
Save discriminator model
Parameters
snapshot_dir (str) – directory path for saving the model
set_seed
PyTorchDiscriminator.set_seed(seed)
Set seed.
Parameters
seed (int) – seed
train
PyTorchDiscriminator.train(data, weights, penalty=True, quantum_instance=None, shots=None)
Perform one training step w.r.t. to the discriminator’s parameters
Parameters
- data (tuple) – real_batch: torch.Tensor, Training data batch. generated_batch: numpy array, Generated data batch.
- weights (tuple) – real problem, generated problem
- penalty (bool) – Indicate whether or not penalty function is applied to the loss function.
- quantum_instance (QuantumInstance) – Quantum Instance (depreciated)
- shots (int) – Number of shots for hardware or qasm execution. Not used for classical network (only quantum ones)
Returns
with Discriminator loss (torch.Tensor) and updated parameters (array).
Return type
dict