tfdeploy 0.4.2

Deploy tensorflow graphs for fast evaluation and export to tensorflow-less environments running numpy.

|Build Status| |Documentation Status| |Package Status|

Deploy `tensorflow <https://www.tensorflow.org>`__ graphs for *fast*
evaluation and export to *tensorflow-less* environments running
`numpy <http://www.numpy.org>`__.

**Now with tensorflow 1.0 support.**

Evaluation usage
''''''''''''''''

.. code:: python

import tfdeploy as td
import numpy as np

model = td.Model("/path/to/model.pkl")
inp, outp = model.get("input", "output")

batch = np.random.rand(10000, 784)
result = outp.eval({inp: batch})

Installation and dependencies
'''''''''''''''''''''''''''''

Via `pip <https://pypi.python.org/pypi/tfdeploy>`__

.. code:: bash

pip install tfdeploy

or by simply copying the file into your project.

Numpy ≥ 1.10 should be installed on your system.
`Scipy <http://www.scipy.org/>`__ is optional. See
`optimization <#optimization>`__ for more info on optional packages.

By design, tensorflow is required when creating a model. All versions ≥
1.0.1 are supported.

Content
~~~~~~~

- `Why? <#why>`__
- `How? <#how>`__

- `Convert your graph <#convert-your-graph>`__
- `Load the model and evaluate <#load-the-model-and-evaluate>`__
- `Write your own operation <#write-your-own-operation>`__

- `Ensembles <#ensembles>`__
- `Optimization <#optimization>`__
- `Performance <#performance>`__
- `Contributing <#contributing>`__
- `Development <#development>`__
- `Authors <#authors>`__
- `License <#license>`__

Why?
----

Working with tensorflow is awesome. Model definition and training is
simple yet powerful, and the range of built-in features is just
striking.

However, when it comes down to model deployment and evaluation, things
get a bit more cumbersome than they should be. You either export your
graph to a new file *and* `save your trained
variables <https://www.tensorflow.org/versions/master/how_tos/variables/index.html#saving-variables>`__
in a separate file, or you make use of tensorflow's `serving
system <https://www.tensorflow.org/versions/master/tutorials/tfserve/index.html>`__.
Wouldn't it be great if you could just export your model to a simple
numpy-based callable? Of course it would. And this is exactly what
tfdeploy does for you.

To boil it down, tfdeploy

- is lightweight. A single file with < 150 lines of core code. Just
copy it to your project.
- `faster <#performance>`__ than using tensorflow's ``Tensor.eval``.
- **does not need tensorflow** during evaluation.
- only depends on numpy.
- can load one or more models from a single file.
- does not support GPUs (maybe
`gnumpy <http://www.cs.toronto.edu/~tijmen/gnumpy.html>`__ is worth a
try here).

How?
----

The central class is ``tfdeploy.Model``. The following two examples
demonstrate how a model can be created from a tensorflow graph, saved to
and loaded from disk, and eventually evaluated.

Convert your graph
''''''''''''''''''

.. code:: python

import tensorflow as tf
import tfdeploy as td

# setup tfdeploy (only when creating models)
td.setup(tf)

# build your graph
sess = tf.Session()

# use names for input and output layers
x = tf.placeholder("float", shape=[None, 784], name="input")
W = tf.Variable(tf.truncated_normal([784, 100], stddev=0.05))
b = tf.Variable(tf.zeros([100]))
y = tf.nn.softmax(tf.matmul(x, W) + b, name="output")

sess.run(tf.global_variables_initializer())

# ... training ...

# create a tfdeploy model and save it to disk
model = td.Model()
model.add(y, sess) # y and all its ops and related tensors are added recursively
model.save("model.pkl")

Load the model and evaluate
'''''''''''''''''''''''''''

.. code:: python

import numpy as np
import tfdeploy as td

model = td.Model("model.pkl")

# shorthand to x and y
x, y = model.get("input", "output")

# evaluate
batch = np.random.rand(10000, 784)
result = y.eval({x: batch})

Write your own ``Operation``
''''''''''''''''''''''''''''

tfdeploy supports most of the ``Operation``'s `implemented in
tensorflow <https://www.tensorflow.org/versions/master/api_docs/python/math_ops.html>`__.
However, if you miss one (in that case, submit a PR or an issue ;) ) or
if you're using custom ops, you might want to extend tfdeploy by
defining a new class op that inherits from ``tfdeploy.Operation``:

.. code:: python

import tensorflow as tf
import tfdeploy as td
import numpy as np

# setup tfdeploy (only when creating models)
td.setup(tf)

# ... write you model here ...

# let's assume your final tensor "y" relies on an op of type "InvertedSoftmax"
# before creating the td.Model, you should add that op to tfdeploy

class InvertedSoftmax(td.Operation):
@staticmethod
def func(a):
e = np.exp(-a)
# ops should return a tuple
return np.divide(e, np.sum(e, axis=-1, keepdims=True)),

# this is equivalent to
# @td.Operation.factory
# def InvertedSoftmax(a):
# e = np.exp(-a)
# return np.divide(e, np.sum(e, axis=-1, keepdims=True)),

# now we're good to go
model = td.Model()
model.add(y, sess)
model.save("model.pkl")

When writing new ops, three things are important:

- Try to avoid loops, prefer numpy vectorization.
- Return a tuple.
- Don't change incoming tensors/arrays in-place, always work on and
return copies.

Ensembles
---------

tfdeploy provides a helper class to evaluate an ensemble of models:
``Ensemble``. It can load multiple models, evaluate them and combine
their output values using different methods.

.. code:: python

# create the ensemble
ensemble = td.Ensemble(["model1.pkl", "model2.pkl", ...], method=td.METHOD_MEAN)

# get input and output tensors (which actually are TensorEnsemble instances)
input, output = ensemble.get("input", "output")

# evaluate the ensemble just like a normal model
batch = ...
value = output.eval({input: batch})

The return value of ``get()`` is a ``TensorEnsemble`` istance. It is
basically a wrapper around multiple tensors and should be used as keys
in the ``feed_dict`` of the ``eval()`` call.

You can choose between ``METHOD_MEAN`` (the default), ``METHOD_MAX`` and
``METHOD_MIN``. If you want to use a custom ensembling method, use
``METHOD_CUSTOM`` and overwrite the static ``func_custom()`` method of
the ``TensorEnsemble`` instance.

Optimization
------------

Most ops are written using pure numpy. However, multiple implementations
of the same op are allowed that may use additional third-party Python
packages providing even faster functionality for some situations.

For example, numpy does not provide a vectorized *lgamma* function.
Thus, the standard ``tfdeploy.Lgamma`` op uses ``math.lgamma`` that was
previously vectorized using ``numpy.vectorize``. For these situations,
additional implementations of the same op are possible (the *lgamma*
example is quite academic, but this definitely makes sense for more
sophisticated ops like pooling). We can simply tell the op to use its
scipy implementation instead:

.. code:: python

td.Lgamma.use_impl(td.IMPL_SCIPY)

Currently, allowed implementation types are numpy (``IMPL_NUMPY``, the
default) and scipy (``IMPL_SCIPY``).

Adding additional implementations
'''''''''''''''''''''''''''''''''

Additional implementations can be added by setting the ``impl``
attribute of the op factory or by using the ``add_impl`` decorator of
existing operations. The first registered implementation will be the
default one.

.. code:: python

# create the default lgamma op with numpy implementation
lgamma_vec = np.vectorize(math.lgamma)

@td.Operation.factory
# equivalent to
# @td.Operation.factory(impl=td.IMPL_NUMPY)
def Lgamma(a):
return lgamma_vec(a),

# add a scipy-based implementation
@Lgamma.add_impl(td.IMPL_SCIPY)
def Lgamma(a):
return sp.special.gammaln(a),

Auto-optimization
'''''''''''''''''

If scipy is available on your system, it is reasonable to use all ops in
their scipy implementation (if it exists, of course). This should be
configured before you create any model from tensorflow objects using the
second argument of the ``setup`` function:

.. code:: python

td.setup(tf, td.IMPL_SCIPY)

Ops that do not implement ``IMPL_SCIPY`` stick with the numpy version
(``IMPL_NUMPY``).

Performance
-----------

tfdeploy is lightweight (1 file, < 150 lines of core code) and fast.
Internal evaluation calls have only very few overhead and tensor
operations use numpy vectorization. The actual performance depends on
the ops in your graph. While most of the tensorflow ops have a numpy
equivalent or can be constructed from numpy functions, a few ops require
additional Python-based loops (e.g. ``BatchMatMul``). But in many cases
it's potentially faster than using tensorflow's ``Tensor.eval``.

This is a comparison for a basic graph where all ops are vectorized
(basically ``Add``, ``MatMul`` and ``Softmax``):

.. code:: bash

> ipython -i tests/perf/simple.py

In [1]: %timeit -n 100 test_tf()
100 loops, best of 3: 109 ms per loop

In [2]: %timeit -n 100 test_td()
100 loops, best of 3: 60.5 ms per loop

Contributing
------------

If you want to contribute with new ops and features, I'm happy to
receive pull requests. Just make sure to add a new test case to
``tests/core.py`` or ``tests/ops.py`` and run them via:

.. code:: bash

> python -m unittest tests

Test grid
'''''''''

In general, tests should be run for different environments:

+----------------------+-------------+
| Variation | Values |
+======================+=============+
| tensorflow version | ``1.0.1`` |
+----------------------+-------------+
| python version | 2, 3 |
+----------------------+-------------+
| ``TD_TEST_SCIPY`` | 0, 1 |
+----------------------+-------------+
| ``TD_TEST_GPU`` | 0, 1 |
+----------------------+-------------+

Docker
''''''

For testing purposes, it is convenient to use docker. Fortunately, the
official `tensorflow
images <https://hub.docker.com/r/tensorflow/tensorflow/>`__ contain all
we need:

.. code:: bash

git clone https://github.com/riga/tfdeploy.git
cd tfdeploy

docker run --rm -v `pwd`:/root/tfdeploy -w /root/tfdeploy -e "TD_TEST_SCIPY=1" tensorflow/tensorflow:1.0.1 python -m unittest tests

Development
-----------

- Source hosted at `GitHub <https://github.com/riga/tfdeploy>`__
- Report issues, questions, feature requests on `GitHub
Issues <https://github.com/riga/tfdeploy/issues>`__

Authors
-------

- `Marcel R. <https://github.com/riga>`__
- `Benjamin F. <https://github.com/bfis>`__

License
-------

The MIT License (MIT)

Copyright (c) 2017 Marcel R.

Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:

The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

.. |Build Status| image:: https://travis-ci.org/riga/tfdeploy.svg?branch=master
:target: https://travis-ci.org/riga/tfdeploy
.. |Documentation Status| image:: https://readthedocs.org/projects/tfdeploy/badge/?version=latest
:target: http://tfdeploy.readthedocs.org/en/latest/?badge=latest
.. |Package Status| image:: https://badge.fury.io/py/tfdeploy.svg
:target: https://badge.fury.io/py/tfdeploy