passing pytest

.gitignore

@@ -214,6 +214,8 @@ docs/_build
 ### IntelliJ ###
 *.iml
 
+### pytest ###
+/.pytest_cache/
 
 ### Project ###
 /data/

docs/conf.py

@@ -326,6 +326,9 @@
             'tf_r0.12': (
                 'https://www.tensorflow.org/versions/r0.12/api_docs/python/%s',
                 None),
+            'tf_hmpg': (
+                'https://www.tensorflow.org/%s',
+                None),
             'gym': (
                 'https://gym.openai.com/docs/%s',
                 None),

requirements.txt

@@ -1,2 +1,4 @@
 tensorflow >= 1.6.0
 tensorflow-gpu >= 1.6.0
+tensorflow-probability >= 0.3.0
+tensorflow-probability-gpu >= 0.3.0

setup.py

@@ -7,10 +7,10 @@
 
 setuptools.setup(
     name="texar",
-    version="0.0.1",
-    url="https://github.com/ZhitingHu/txtgen",
+    version="0.1",
+    url="https://github.com/asyml/texar",
 
-    description="An open and flexible framework for text generation.",
+    description="Toolkit for Text Generation and Beyond",
 
     packages=setuptools.find_packages(),
     platforms='any',

texar/core/layers.py

@@ -773,6 +773,9 @@ def __init__(self,
                 else:
                     self._layers.append(get_layer(hparams=layer))
 
+        # Keep tracks of whether trainable variables have been created
+        self._vars_built = False
+
     def compute_output_shape(self, input_shape):
         if self._layers is None:
             _shapes = input_shape
@@ -870,8 +873,9 @@ def call(self, inputs):
         else:
             raise ValueError("Unknown merge mode: '%s'" % self._mode)
 
-        if not self.built:
+        if not self.built or not self._vars_built:
             self._collect_weights()
+            self._vars_built = True
 
         return outputs
 
@@ -881,11 +885,6 @@ def layers(self):
         """
         return self._layers
 
-    def build(self, _):
-        """Dumb method.
-        """
-        # Does not set :attr:`self.built` as this point.
-        pass
 
 class SequentialLayer(tf.layers.Layer):
     """A subclass of :tf_main:`tf.layers.Layer <layers/Layer>`.
@@ -915,6 +914,9 @@ def __init__(self,
             else:
                 self._layers.append(get_layer(hparams=layer))
 
+        # Keep tracks of whether trainable variables have been created
+        self._vars_built = False
+
     def compute_output_shape(self, input_shape):
         input_shape = tf.TensorShape(input_shape)
         for layer in self._layers:
@@ -947,8 +949,9 @@ def call(self, inputs, mode=None): # pylint: disable=arguments-differ
                 outputs = layer(inputs)
             inputs = outputs
 
-        if not self.built:
+        if not self.built or not self._vars_built:
             self._collect_weights()
+            self._vars_built = True
 
         return outputs
 
@@ -958,11 +961,6 @@ def layers(self):
         """
         return self._layers
 
-    def build(self, _):
-        """Dumb method.
-        """
-        # Does not set :attr:`self.built` as this point.
-        pass
 
 def _common_default_conv_dense_kwargs():
     """Returns the default keyword argument values that are common to

texar/core/optimization.py

@@ -19,8 +19,6 @@
 from __future__ import print_function
 from __future__ import division
 
-import inspect
-
 import tensorflow as tf
 
 from texar.hyperparams import HParams
@@ -216,7 +214,7 @@ def get_optimizer_fn(hparams=None):
 
     def _get_opt(learning_rate=None):
         opt_kwargs = hparams["kwargs"].todict()
-        fn_args = set(inspect.getargspec(opt_class.__init__).args)
+        fn_args = set(utils.get_args(opt_class.__init__))
         if 'learning_rate' in fn_args and learning_rate is not None:
             opt_kwargs["learning_rate"] = learning_rate
         return opt_class(**opt_kwargs)
@@ -319,7 +317,7 @@ def get_gradient_clip_fn(hparams=None):
 
     fn_modules = ["tensorflow", "texar.custom"]
     clip_fn = utils.get_function(fn_type, fn_modules)
-    clip_fn_args = inspect.getargspec(clip_fn).args
+    clip_fn_args = utils.get_args(clip_fn)
     fn_kwargs = hparams["kwargs"]
     if isinstance(fn_kwargs, HParams):
         fn_kwargs = fn_kwargs.todict()

texar/losses/rewards_test.py

@@ -46,8 +46,10 @@ def test_discount_reward(self):
 
             r, r_n = sess.run([discounted_reward_, discounted_reward_n_])
 
-            np.testing.assert_array_equal(discounted_reward, r)
-            np.testing.assert_array_equal(discounted_reward_n, r_n)
+            np.testing.assert_array_almost_equal(
+                discounted_reward, r, decimal=6)
+            np.testing.assert_array_almost_equal(
+                discounted_reward_n, r_n, decimal=6)
 
         # 2D
         reward = np.ones([2, 10], dtype=np.float64)
@@ -70,8 +72,10 @@ def test_discount_reward(self):
 
             r, r_n = sess.run([discounted_reward_, discounted_reward_n_])
 
-            np.testing.assert_array_equal(discounted_reward, r)
-            np.testing.assert_array_equal(discounted_reward_n, r_n)
+            np.testing.assert_array_almost_equal(
+                discounted_reward, r, decimal=6)
+            np.testing.assert_array_almost_equal(
+                discounted_reward_n, r_n, decimal=6)
 
     def test_discount_reward_py_1d(self):
         """Tests :func:`texar.losses.rewards._discount_reward_py_1d`

texar/modules/connectors/connectors.py

@@ -1,4 +1,16 @@
+# Copyright 2018 The Texar Authors. All Rights Reserved.
 #
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
 """
 Various connectors.
 """
@@ -10,7 +22,7 @@
 import numpy as np
 
 import tensorflow as tf
-import tensorflow.contrib.distributions as tf_dstr
+from tensorflow import distributions as tf_dstr
 from tensorflow.python.util import nest    # pylint: disable=E0611
 
 from texar.modules.connectors.connector_base import ConnectorBase
@@ -484,10 +496,10 @@ def _build(self,
         `distribution.reparameterization_type = FULLY_REPARAMETERIZED`.
 
         Args:
-            distribution: A
-                :tf_main:`TF Distribution <contrib/distributions/Distribution>`.
-                Can be a class, its name or module path, or an instance of
-                a subclass.
+            distribution: A instance of subclass of
+                :tf_main:`TF Distribution <distributions/Distribution>`,
+                or :tf_hmpg:`tensorflow_probability Distribution <probability>`,
+                Can be a class, its name or module path, or a class instance.
             distribution_kwargs (dict, optional): Keyword arguments for the
                 distribution constructor. Ignored if `distribution` is a
                 class instance.
@@ -515,7 +527,8 @@ class instance.
         """
         dstr = check_or_get_instance(
             distribution, distribution_kwargs,
-            ["tensorflow.contrib.distributions", "texar.custom"])
+            ["tensorflow.distributions", "tensorflow_probability.distributions",
+             "texar.custom"])
 
         if dstr.reparameterization_type == tf_dstr.NOT_REPARAMETERIZED:
             raise ValueError(
@@ -614,10 +627,10 @@ def _build(self,
         cannot be back-propagate through the samples.
 
         Args:
-            distribution: A
-                :tf_main:`TF Distribution <contrib/distributions/Distribution>`.
-                Can be a class, its name or module path, or an instance of
-                a subclass.
+            distribution: A instance of subclass of
+                :tf_main:`TF Distribution <distributions/Distribution>`,
+                or :tf_hmpg:`tensorflow_probability Distribution <probability>`.
+                Can be a class, its name or module path, or a class instance.
             distribution_kwargs (dict, optional): Keyword arguments for the
                 distribution constructor. Ignored if `distribution` is a
                 class instance.
@@ -644,7 +657,8 @@ class instance.
         """
         dstr = check_or_get_instance(
             distribution, distribution_kwargs,
-            ["tensorflow.contrib.distributions", "texar.custom"])
+            ["tensorflow.distributions", "tensorflow_probability.distributions",
+             "tensorflow.contrib.distributions", "texar.custom"])
 
         if num_samples:
             output = dstr.sample(num_samples)

texar/modules/connectors/connectors_test.py

@@ -9,7 +9,7 @@
 from __future__ import unicode_literals
 
 import tensorflow as tf
-import tensorflow.contrib.distributions as tfds
+from tensorflow_probability import distributions as tfpd
 from tensorflow.python.util import nest    # pylint: disable=E0611
 
 from texar.core import layers
@@ -85,8 +85,8 @@ def test_reparameterized_stochastic_connector(self):
         var = tf.ones([self._batch_size, variable_size])
         mu_vec = tf.zeros([variable_size])
         var_vec = tf.ones([variable_size])
-        gauss_ds = tfds.MultivariateNormalDiag(loc=mu, scale_diag=var)
-        gauss_ds_vec = tfds.MultivariateNormalDiag(loc=mu_vec,
+        gauss_ds = tfpd.MultivariateNormalDiag(loc=mu, scale_diag=var)
+        gauss_ds_vec = tfpd.MultivariateNormalDiag(loc=mu_vec,
                                                    scale_diag=var_vec)
         gauss_connector = ReparameterizedStochasticConnector(state_size)
         gauss_connector_ts = ReparameterizedStochasticConnector(state_size_ts)

texar/modules/embedders/embedder_utils.py

@@ -1,4 +1,16 @@
+# Copyright 2018 The Texar Authors. All Rights Reserved.
 #
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
 """Utils of embedder.
 """
 

texar/modules/encoders/hierarchical_encoders_test.py

@@ -67,10 +67,15 @@ def test_order(self):
             minval=-1,
             dtype=tf.float32)
 
-        outputs, state = encoder(inputs, order='btu', time_major=False)
-        outputs, state = encoder(inputs, order='utb', time_major=True)
-        outputs, state = encoder(inputs, order='tbu', time_major_major=True)
-        outputs, state = encoder(inputs, order='ubt', time_major_minor=True)
+        outputs_1, state_1 = encoder(inputs, order='btu')
+        outputs_2, state_2 = encoder(inputs, order='utb')
+        outputs_3, state_3 = encoder(inputs, order='tbu')
+        outputs_4, state_4 = encoder(inputs, order='ubt')
+
+        with self.test_session() as sess:
+            sess.run(tf.global_variables_initializer())
+            sess.run([outputs_1, state_1, outputs_2, state_2,
+                      outputs_3, state_3, outputs_4, state_4])
 
     def test_depack(self):
         hparams = {

texar/utils/utils.py

@@ -34,6 +34,8 @@
 from texar.utils.dtypes import is_str, is_callable, compat_as_text, \
         _maybe_list_to_array
 
+# pylint: disable=anomalous-backslash-in-string
+
 MAX_SEQ_LENGTH = np.iinfo(np.int32).max
 
 ## Some modules cannot be imported directly,
@@ -44,6 +46,7 @@
 #}
 
 __all__ = [
+    "_inspect_getargspec",
     "check_or_get_class",
     "get_class",
     "check_or_get_instance",
@@ -81,6 +84,15 @@ def _expand_name(name):
     """
     return name
 
+def _inspect_getargspec(fn):
+    """Returns `inspect.getargspec(fn)` for Py2 and `inspect.getfullargspec(fn)`
+    for Py3
+    """
+    try:
+        return inspect.getfullargspec(fn)
+    except AttributeError:
+        return inspect.getargspec(fn)
+
 def check_or_get_class(class_or_name, module_path=None, superclass=None):
     """Returns the class and checks if the class inherits :attr:`superclass`.
 
@@ -214,14 +226,14 @@ def get_instance(class_or_name, kwargs, module_paths=None):
     if is_str(class_):
         class_ = get_class(class_, module_paths)
     # Check validity of arguments
-    class_args = set(inspect.getargspec(class_.__init__).args)
+    class_args = set(_inspect_getargspec(class_.__init__).args)
     if kwargs is None:
         kwargs = {}
     for key in kwargs.keys():
         if key not in class_args:
             raise ValueError(
-                "Invalid argument for class %s.%s: %s, valid args:%s" %
-                (class_.__module__, class_.__name__, key, class_args))
+                "Invalid argument for class %s.%s: %s, valid args: %s" %
+                (class_.__module__, class_.__name__, key, list(class_args)))
 
     return class_(**kwargs)
 
@@ -293,7 +305,7 @@ class construction method are used.
 
     # Select valid arguments
     selected_kwargs = {}
-    class_args = set(inspect.getargspec(class_.__init__).args)
+    class_args = set(_inspect_getargspec(class_.__init__).args)
     if kwargs is None:
         kwargs = {}
     for key, value in kwargs.items():
@@ -354,9 +366,9 @@ def call_function_with_redundant_kwargs(fn, kwargs):
         The returned results by calling :attr:`fn`.
     """
     try:
-        fn_args = set(inspect.getargspec(fn).args)
+        fn_args = set(_inspect_getargspec(fn).args)
     except TypeError:
-        fn_args = set(inspect.getargspec(fn.__call__).args)
+        fn_args = set(_inspect_getargspec(fn.__call__).args)
 
     if kwargs is None:
         kwargs = {}
@@ -378,7 +390,7 @@ def get_args(fn):
     Returns:
         list: A list of argument names (str) of the function.
     """
-    argspec = inspect.getargspec(fn)
+    argspec = _inspect_getargspec(fn)
     return argspec.args
 
 def get_default_arg_values(fn):
@@ -393,7 +405,7 @@ def get_default_arg_values(fn):
         dict: A dictionary that maps argument names (str) to their default
         values. The dictionary is empty if no arguments have default values.
     """
-    argspec = inspect.getargspec(fn)
+    argspec = _inspect_getargspec(fn)
     if argspec.defaults is None:
         return {}
     num_defaults = len(argspec.defaults)