Type losses
Namespace tensorflow.contrib.losses
Methods
- absolute_difference
- absolute_difference
- absolute_difference_dyn
- add_loss
- add_loss
- add_loss
- add_loss_dyn
- compute_weighted_loss
- compute_weighted_loss
- compute_weighted_loss
- compute_weighted_loss
- compute_weighted_loss_dyn
- cosine_distance
- cosine_distance
- cosine_distance_dyn
- get_losses
- get_losses_dyn
- get_regularization_losses
- get_regularization_losses_dyn
- get_total_loss
- get_total_loss_dyn
- hinge_loss
- hinge_loss_dyn
- log_loss
- log_loss
- log_loss_dyn
- mean_pairwise_squared_error
- mean_pairwise_squared_error
- mean_pairwise_squared_error
- mean_pairwise_squared_error_dyn
- mean_squared_error
- mean_squared_error
- mean_squared_error_dyn
- sigmoid_cross_entropy
- sigmoid_cross_entropy
- sigmoid_cross_entropy
- sigmoid_cross_entropy
- sigmoid_cross_entropy_dyn
- softmax_cross_entropy
- softmax_cross_entropy
- softmax_cross_entropy
- softmax_cross_entropy
- softmax_cross_entropy_dyn
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy
- sparse_softmax_cross_entropy_dyn
Properties
- absolute_difference_fn
- add_loss_fn
- compute_weighted_loss_fn
- cosine_distance_fn
- get_losses_fn
- get_regularization_losses_fn
- get_total_loss_fn
- hinge_loss_fn
- log_loss_fn
- mean_pairwise_squared_error_fn
- mean_squared_error_fn
- sigmoid_cross_entropy_fn
- softmax_cross_entropy_fn
- sparse_softmax_cross_entropy_fn
Public static methods
object absolute_difference(IGraphNodeBase predictions, IGraphNodeBase labels, IGraphNodeBase weights, object scope)
object absolute_difference(IGraphNodeBase predictions, IGraphNodeBase labels, double weights, object scope)
object absolute_difference_dyn(object predictions, object labels, ImplicitContainer<T> weights, object scope)
void add_loss(ValueTuple<PythonClassContainer, PythonClassContainer> loss, ImplicitContainer<T> loss_collection)
Adds a externally defined loss to the collection of losses.
Parameters
-
ValueTuple<PythonClassContainer, PythonClassContainer>
loss - A loss `Tensor`.
-
ImplicitContainer<T>
loss_collection - Optional collection to add the loss to.
void add_loss(IndexedSlices loss, ImplicitContainer<T> loss_collection)
Adds a externally defined loss to the collection of losses.
Parameters
-
IndexedSlices
loss - A loss `Tensor`.
-
ImplicitContainer<T>
loss_collection - Optional collection to add the loss to.
void add_loss(IGraphNodeBase loss, ImplicitContainer<T> loss_collection)
Adds a externally defined loss to the collection of losses.
Parameters
-
IGraphNodeBase
loss - A loss `Tensor`.
-
ImplicitContainer<T>
loss_collection - Optional collection to add the loss to.
object add_loss_dyn(object loss, ImplicitContainer<T> loss_collection)
Adds a externally defined loss to the collection of losses.
Parameters
-
object
loss - A loss `Tensor`.
-
ImplicitContainer<T>
loss_collection - Optional collection to add the loss to.
object compute_weighted_loss(IGraphNodeBase losses, double weights, object scope)
object compute_weighted_loss(IGraphNodeBase losses, double weights, string scope)
object compute_weighted_loss(IGraphNodeBase losses, IGraphNodeBase weights, object scope)
object compute_weighted_loss(IGraphNodeBase losses, IGraphNodeBase weights, string scope)
object compute_weighted_loss_dyn(object losses, ImplicitContainer<T> weights, object scope)
object cosine_distance(IGraphNodeBase predictions, IGraphNodeBase labels, object axis, double weights, object scope, Nullable<int> dim)
object cosine_distance(IGraphNodeBase predictions, IGraphNodeBase labels, object axis, IGraphNodeBase weights, object scope, Nullable<int> dim)
object cosine_distance_dyn(object predictions, object labels, object axis, ImplicitContainer<T> weights, object scope, object dim)
IList<object> get_losses(object scope, ImplicitContainer<T> loss_collection)
Gets the list of losses from the loss_collection.
Parameters
-
object
scope - An optional scope name for filtering the losses to return.
-
ImplicitContainer<T>
loss_collection - Optional losses collection.
Returns
-
IList<object>
- a list of loss tensors.
object get_losses_dyn(object scope, ImplicitContainer<T> loss_collection)
Gets the list of losses from the loss_collection.
Parameters
-
object
scope - An optional scope name for filtering the losses to return.
-
ImplicitContainer<T>
loss_collection - Optional losses collection.
Returns
-
object
- a list of loss tensors.
IList<object> get_regularization_losses(object scope)
object get_regularization_losses_dyn(object scope)
Gets the list of regularization losses.
Parameters
-
object
scope - An optional scope name for filtering the losses to return.
Returns
-
object
- A list of regularization losses as Tensors.