def _tree_train_op_fn(loss):
"""Returns the op to optimize the loss."""
if dnn_to_tree_distillation_param:
loss_weight, loss_fn = dnn_to_tree_distillation_param
weight_tensor = head_lib._weight_tensor( # pylint: disable=protected-access
features, head.weight_column_name)
dnn_logits_fixed = array_ops.stop_gradient(dnn_logits)
if loss_fn is None:
# we create the loss_fn similar to the head loss_fn for
# multi_class_head used previously as the default one.
n_classes = 2 if head.logits_dimension == 1 else head.logits_dimension
loss_fn = distillation_loss.create_dnn_to_tree_cross_entropy_loss_fn(
n_classes)
dnn_to_tree_distillation_loss = loss_weight * loss_fn(
dnn_logits_fixed, tree_logits, weight_tensor)
summary.scalar("dnn_to_tree_distillation_loss",
dnn_to_tree_distillation_loss)
loss += dnn_to_tree_distillation_loss
update_op = gbdt_model.train(loss, predictions_dict, labels)
with ops.control_dependencies(
[update_op]), (ops.colocate_with(global_step)):
update_op = state_ops.assign_add(global_step, 1).op
return update_op
def _tree_train_op_fn(loss):
"""Returns the op to optimize the loss."""
if dnn_to_tree_distillation_param:
loss_weight, loss_fn = dnn_to_tree_distillation_param
weight_tensor = head_lib._weight_tensor( # pylint: disable=protected-access
features, head.weight_column_name)
dnn_logits_fixed = array_ops.stop_gradient(dnn_logits)
if loss_fn is None:
# we create the loss_fn similar to the head loss_fn for
# multi_class_head used previously as the default one.
n_classes = 2 if head.logits_dimension == 1 else head.logits_dimension
loss_fn = distillation_loss.create_dnn_to_tree_cross_entropy_loss_fn(
n_classes)
dnn_to_tree_distillation_loss = loss_weight * loss_fn(
dnn_logits_fixed, tree_logits, weight_tensor)
summary.scalar("dnn_to_tree_distillation_loss",
dnn_to_tree_distillation_loss)
loss += dnn_to_tree_distillation_loss
update_op = gbdt_model.train(loss, predictions_dict, labels)
with ops.control_dependencies(
[update_op]), (ops.colocate_with(global_step)):
update_op = state_ops.assign_add(global_step, 1).op
return update_op