neuralmonkey.trainers.generic_trainer module¶
-
class
neuralmonkey.trainers.generic_trainer.
GenericTrainer
(objectives: List[neuralmonkey.trainers.generic_trainer.Objective], l1_weight: float = 0.0, l2_weight: float = 0.0, clip_norm: float = None, optimizer: tensorflow.python.training.optimizer.Optimizer = None, var_scopes: List[str] = None, var_collection: str = None) → None¶ Bases:
neuralmonkey.runners.base_runner.GraphExecutor
-
class
Executable
(executor: neuralmonkey.trainers.generic_trainer.GenericTrainer, compute_losses: bool, summaries: bool, num_sessions: int) → None¶ Bases:
neuralmonkey.runners.base_runner.Executable
-
__init__
(executor: neuralmonkey.trainers.generic_trainer.GenericTrainer, compute_losses: bool, summaries: bool, num_sessions: int) → None¶ Initialize self. See help(type(self)) for accurate signature.
-
collect_results
(results: List[Dict]) → None¶
-
next_to_execute
() → Tuple[Union[Dict, List], List[Dict[tensorflow.python.framework.ops.Tensor, Union[int, float, numpy.ndarray]]]]¶ Get the tensors and additional feed dicts for execution.
-
-
__init__
(objectives: List[neuralmonkey.trainers.generic_trainer.Objective], l1_weight: float = 0.0, l2_weight: float = 0.0, clip_norm: float = None, optimizer: tensorflow.python.training.optimizer.Optimizer = None, var_scopes: List[str] = None, var_collection: str = None) → None¶ Initialize self. See help(type(self)) for accurate signature.
-
static
default_optimizer
()¶
-
differentiable_loss_sum
¶ Compute the differentiable loss (including regularization).
-
fetches
¶
-
gradients
¶
-
objective_values
¶ Compute unweighted losses for fetching.
-
raw_gradients
¶ Compute the gradients.
-
regularization_losses
¶ Compute the regularization losses, e.g. L1 and L2.
-
summaries
¶
-
train_op
¶ Construct the training op.
-
var_list
¶
-
class
-
class
neuralmonkey.trainers.generic_trainer.
Objective
¶ Bases:
neuralmonkey.trainers.generic_trainer.Objective
The training objective.
-
name
¶ The name for the objective. Used in TensorBoard.
-
decoder
¶ The decoder which generates the value to optimize.
-
loss
¶ The loss tensor fetched by the trainer.
-
gradients
¶ Manually specified gradients. Useful for reinforcement learning.
-
weight
¶ The weight of this objective. The loss will be multiplied by this so the gradients can be controled in case of multiple objectives.
-