tf.py_function  |  TensorFlow v2.0.0 (original) (raw)

tf.py_function

Stay organized with collections Save and categorize content based on your preferences.

Wraps a python function into a TensorFlow op that executes it eagerly.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.py_function

tf.py_function(
    func, inp, Tout, name=None
)

This function allows expressing computations in a TensorFlow graph as Python functions. In particular, it wraps a Python function funcin a once-differentiable TensorFlow operation that executes it with eager execution enabled. As a consequence, tf.py_function makes it possible to express control flow using Python constructs (if, while,for, etc.), instead of TensorFlow control flow constructs (tf.cond,tf.while_loop). For example, you might use tf.py_function to implement the log huber function:

def log_huber(x, m):
  if tf.abs(x) <= m:
    return x**2
  else:
    return m**2 * (1 - 2 * tf.math.log(m) + tf.math.log(x**2))

x = tf.compat.v1.placeholder(tf.float32)
m = tf.compat.v1.placeholder(tf.float32)

y = tf.py_function(func=log_huber, inp=[x, m], Tout=tf.float32)
dy_dx = tf.gradients(y, x)[0]

with tf.compat.v1.Session() as sess:
  # The session executes `log_huber` eagerly. Given the feed values below,
  # it will take the first branch, so `y` evaluates to 1.0 and
  # `dy_dx` evaluates to 2.0.
  y, dy_dx = sess.run([y, dy_dx], feed_dict={x: 1.0, m: 2.0})

You can also use tf.py_function to debug your models at runtime using Python tools, i.e., you can isolate portions of your code that you want to debug, wrap them in Python functions and insert pdb tracepoints or print statements as desired, and wrap those functions intf.py_function.

For more information on eager execution, see theEager guide.

tf.py_function is similar in spirit to tf.compat.v1.py_func, but unlike the latter, the former lets you use TensorFlow operations in the wrapped Python function. In particular, while tf.compat.v1.py_func only runs on CPUs and wraps functions that take NumPy arrays as inputs and return NumPy arrays as outputs, tf.py_function can be placed on GPUs and wraps functions that take Tensors as inputs, execute TensorFlow operations in their bodies, and return Tensors as outputs.

Like tf.compat.v1.py_func, tf.py_function has the following limitations with respect to serialization and distribution:

• The body of the function (i.e. func) will not be serialized in aGraphDef. Therefore, you should not use this function if you need to serialize your model and restore it in a different environment.
• The operation must run in the same address space as the Python program that calls tf.py_function(). If you are using distributed TensorFlow, you must run a tf.distribute.Server in the same process as the program that calls tf.py_function() and you must pin the created operation to a device in that server (e.g. using with tf.device():).