| View source on GitHub |
Wrapper for Graph.control_dependencies() using the default graph.
tf.control_dependencies(
control_inputs
)
See tf.Graph.control_dependencies for more details.
When eager execution is enabled, any callable object in the control_inputs list will be called.
| Args | |
|---|---|
control_inputs | A list of Operation or Tensor objects which must be executed or computed before running the operations defined in the context. Can also be None to clear the control dependencies. If eager execution is enabled, any callable object in the control_inputs list will be called. |
| Returns | |
|---|---|
| A context manager that specifies control dependencies for all operations constructed within the context. |
© 2020 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/control_dependencies