|---| | [ View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.21.0/tensorflow/python/eager/polymorphic_function/eager_function_run.py#L31-L76) | Enables / disables eager execution of [`tf.function`](https://www.tensorflow.org/api_docs/python/tf/function)s. #### View aliases **Compat aliases for migration** See [Migration guide](https://www.tensorflow.org/guide/migrate) for more details. [`tf.compat.v1.config.run_functions_eagerly`](https://www.tensorflow.org/api_docs/python/tf/config/run_functions_eagerly)
tf.config.run_functions_eagerly( run_eagerly ) ### Used in the notebooks | Used in the guide | |---| | - [Introduction to graphs and tf.function](https://www.tensorflow.org/guide/intro_to_graphs) | Calling [`tf.config.run_functions_eagerly(True)`](https://www.tensorflow.org/api_docs/python/tf/config/run_functions_eagerly) will make all invocations of [`tf.function`](https://www.tensorflow.org/api_docs/python/tf/function) run eagerly instead of running as a traced graph function. This can be useful for debugging. As the code now runs line-by-line, you can add arbitrary `print` messages or pdb breakpoints to monitor the inputs/outputs of each Tensorflow operation. However, you should avoid using this for actual production because it significantly slows down execution. def my_func(a): print(f'a: {a}') return a + a a_fn = tf.function(my_func) # A side effect the first time the function is traced # In tracing time, `a` is printed with shape and dtype only a_fn(tf.constant(1)) a: Tensor("a:0", shape=(), dtype=int32) # `print` is a python side effect, it won't execute as the traced function # is called a_fn(tf.constant(2)) # Now, switch to eager running tf.config.run_functions_eagerly(True) # The code now runs eagerly and the actual value of `a` is printed a_fn(tf.constant(2)) a: 2 # Turn this back off tf.config.run_functions_eagerly(False) > [!NOTE] > **Note:** This flag has no effect on functions passed into tf.data transformations as arguments. tf.data functions are never executed eagerly and are always executed as a compiled Tensorflow Graph.


| ## Args || |---|---| | `run_eagerly` | Boolean. Whether to run functions eagerly. |