Python tensorflow.Optimizer() Examples
The following are 30
code examples of tensorflow.Optimizer().
You can vote up the ones you like or vote down the ones you don't like,
and go to the original project or source file by following the links above each example.
You may also want to check out all available functions/classes of the module
tensorflow
, or try the search function
.
.
Example #1
| Source File: model_deploy.py From uai-sdk with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #2
| Source File: model_deploy.py From yolo_v2 with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #3
| Source File: __init__.py From training_results_v0.6 with Apache License 2.0 | 6 votes |
|
def compute_gradients(self, *args, **kwargs):
"""Compute gradients of all trainable variables.
See Optimizer.compute_gradients() for more info.
In DistributedOptimizer, compute_gradients() is overriden to also
allreduce the gradients before returning them.
"""
gradients = self._optimizer.compute_gradients(*args, **kwargs)
if size() > 1:
averaged_gradients = []
with tf.name_scope(self._name + "_Allreduce"):
for grad, var in gradients:
if grad is not None:
avg_grad = allreduce(grad,
device_dense=self._device_dense,
device_sparse=self._device_sparse,
compression=self._compression)
averaged_gradients.append((avg_grad, var))
else:
averaged_gradients.append((None, var))
return averaged_gradients
else:
return gradients
Example #4
| Source File: model_deploy.py From CVTron with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #5
| Source File: model_deploy.py From ctw-baseline with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #6
| Source File: model_deploy.py From morph-net with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss+clone.reg_loss, **kwargs)
return sum_loss, clone_grad
Example #7
| Source File: model_deploy.py From Gun-Detector with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #8
| Source File: model_deploy.py From Creative-Adversarial-Networks with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #9
| Source File: model_deploy.py From BMW-TensorFlow-Training-GUI with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #10
| Source File: model_deploy.py From terngrad with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #11
| Source File: model_deploy.py From hops-tensorflow with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #12
| Source File: model_deploy.py From MobileNet with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #13
| Source File: model_deploy.py From shuttleNet with GNU General Public License v3.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #14
| Source File: model_deploy.py From DOTA_models with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #15
| Source File: model_deploy.py From tensorflow_yolo2 with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #16
| Source File: model_deploy.py From MAX-Image-Segmenter with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #17
| Source File: model_deploy.py From tumblr-emotions with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #18
| Source File: model_deploy.py From YOLO2TensorFlow with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #19
| Source File: model_deploy.py From Cross-Modal-Projection-Learning with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #20
| Source File: model_deploy.py From tensorflow-densenet with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #21
| Source File: model_deploy.py From Action_Recognition_Zoo with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #22
| Source File: model_deploy.py From aster with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #23
| Source File: model_deploy.py From ECO-pytorch with BSD 2-Clause "Simplified" License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #24
| Source File: model_deploy.py From Machine-Learning-with-TensorFlow-1.x with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #25
| Source File: model_deploy.py From hands-detection with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #26
| Source File: model_deploy.py From object_detection_kitti with Apache License 2.0 | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #27
| Source File: model_deploy.py From MBMD with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #28
| Source File: model_deploy.py From Optical-Flow-Guided-Feature with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #29
| Source File: model_deploy.py From object_detection_with_tensorflow with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
Example #30
| Source File: model_deploy.py From SENet-tensorflow-slim with MIT License | 6 votes |
|
def _optimize_clone(optimizer, clone, num_clones, regularization_losses,
**kwargs):
"""Compute losses and gradients for a single clone.
Args:
optimizer: A tf.Optimizer object.
clone: A Clone namedtuple.
num_clones: The number of clones being deployed.
regularization_losses: Possibly empty list of regularization_losses
to add to the clone losses.
**kwargs: Dict of kwarg to pass to compute_gradients().
Returns:
A tuple (clone_loss, clone_grads_and_vars).
- clone_loss: A tensor for the total loss for the clone. Can be None.
- clone_grads_and_vars: List of (gradient, variable) for the clone.
Can be empty.
"""
sum_loss = _gather_clone_loss(clone, num_clones, regularization_losses)
clone_grad = None
if sum_loss is not None:
with tf.device(clone.device):
clone_grad = optimizer.compute_gradients(sum_loss, **kwargs)
return sum_loss, clone_grad
