import namespace cleanup in tfutils

examples/GAN/InfoGAN-mnist.py

@@ -14,6 +14,7 @@ from tensorpack import *
 from tensorpack.utils.viz import *
 from tensorpack.tfutils.distributions import *
 from tensorpack.tfutils.scope_utils import auto_reuse_variable_scope
+from tensorpack.tfutils import optimizer, summary
 import tensorpack.tfutils.symbolic_functions as symbf
 from tensorpack.tfutils.gradproc import ScaleGradient, CheckGradient
 from GAN import GANTrainer, GANModelDesc

examples/GAN/WGAN.py

@@ -7,6 +7,7 @@ import os
 import argparse
 
 from tensorpack import *
+from tensorpack.tfutils import optimizer
 from tensorpack.tfutils.summary import add_moving_summary
 from tensorpack.utils.globvars import globalns as G
 import tensorflow as tf

examples/HED/hed.py

@@ -12,7 +12,8 @@ import os
 import sys
 
 from tensorpack import *
-from tensorpack.tfutils.symbolic_functions import *
+import tensorpack.tfutils.symbolic_functions as symbf
+from tensorpack.tfutils import optimizer
 from tensorpack.tfutils.summary import *
 
 
@@ -72,7 +73,7 @@ class Model(ModelDesc):
         costs = []
         for idx, b in enumerate([b1, b2, b3, b4, b5, final_map]):
             output = tf.nn.sigmoid(b, name='output{}'.format(idx + 1))
-            xentropy = class_balanced_sigmoid_cross_entropy(
+            xentropy = symbf.class_balanced_sigmoid_cross_entropy(
                 b, edgemap,
                 name='xentropy{}'.format(idx + 1))
             costs.append(xentropy)
@@ -93,7 +94,7 @@ class Model(ModelDesc):
             add_moving_summary(costs + [wrong, self.cost])
 
     def _get_optimizer(self):
-        lr = get_scalar_var('learning_rate', 3e-5, summary=True)
+        lr = symbf.get_scalar_var('learning_rate', 3e-5, summary=True)
         opt = tf.train.AdamOptimizer(lr, epsilon=1e-3)
         return optimizer.apply_grad_processors(
             opt, [gradproc.ScaleGradient(

examples/PennTreebank/PTB-LSTM.py

@@ -9,6 +9,7 @@ import argparse
 
 from tensorpack import *
 from tensorpack.tfutils.gradproc import *
+from tensorpack.tfutils import optimizer, summary
 from tensorpack.utils import logger
 from tensorpack.utils.fs import download, get_dataset_path
 from tensorpack.utils.argtools import memoized_ignoreargs

examples/Saliency/CAM-resnet.py

@@ -12,6 +12,7 @@ import multiprocessing
 import tensorflow as tf
 from tensorflow.contrib.layers import variance_scaling_initializer
 from tensorpack import *
+from tensorpack.tfutils import optimizer
 from tensorpack.tfutils.symbolic_functions import *
 from tensorpack.tfutils.summary import *
 

examples/SpatialTransformer/mnist-addition.py

@@ -11,6 +11,7 @@ import sys
 import argparse
 
 from tensorpack import *
+from tensorpack.tfutils import sesscreate, optimizer, summary
 import tensorpack.tfutils.symbolic_functions as symbf
 
 IMAGE_SIZE = 42
@@ -78,7 +79,7 @@ class Model(ModelDesc):
         cost = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=logits, labels=label)
         cost = tf.reduce_mean(cost, name='cross_entropy_loss')
 
-        wrong = symbolic_functions.prediction_incorrect(logits, label)
+        wrong = symbf.prediction_incorrect(logits, label)
         summary.add_moving_summary(tf.reduce_mean(wrong, name='train_error'))
 
         wd_cost = tf.multiply(1e-5, regularize_cost('fc.*/W', tf.nn.l2_loss),

examples/mnist-convnet.py

@@ -13,6 +13,7 @@ about 0.6% validation error after 30 epochs.
 
 # Just import everything into current namespace
 from tensorpack import *
+from tensorpack.tfutils import summary
 import tensorflow as tf
 import tensorpack.tfutils.symbolic_functions as symbf
 

examples/svhn-digit-convnet.py

@@ -8,7 +8,7 @@ import numpy as np
 import os
 
 from tensorpack import *
-from tensorpack.tfutils.symbolic_functions import *
+from tensorpack.tfutils.symbolic_functions import prediction_incorrect
 from tensorpack.tfutils.summary import *
 import tensorflow as tf
 

tensorpack/tfutils/__init__.py

@@ -33,6 +33,5 @@ for _, module_name, _ in iter_modules(
         continue
     if module_name in _TO_IMPORT:
         _global_import(module_name)  # import the content to tfutils.*
-    else:
-        __all__.append(module_name)  # import the module separately
-__all__.extend(['sessinit', 'gradproc'])
+__all__.extend(['sessinit', 'summary', 'optimizer',
+                'sesscreate', 'gradproc', 'varreplace'])

tensorpack/tfutils/symbolic_functions.py

@@ -6,7 +6,7 @@ import tensorflow as tf
 from contextlib import contextmanager
 import numpy as np
 
-__all__ = []
+# __all__ = ['get_scalar_var']
 
 
 # this function exists for backwards-compatibilty

tensorpack/tfutils/varmanip.py

@@ -7,7 +7,6 @@ import six
 import os
 import pprint
 import tensorflow as tf
-from collections import defaultdict
 import numpy as np
 from ..utils import logger
 from .common import get_op_tensor_name
@@ -47,9 +46,7 @@ class SessionUpdate(object):
             vars_to_update: a collection of variables to update
         """
         self.sess = sess
-        self.name_map = defaultdict(list)
-        for v in vars_to_update:
-            self.name_map[v.name].append(v)
+        self.name_map = {v.name: v for v in vars_to_update}
 
     @staticmethod
     def load_value_to_var(var, val, strict=False):
@@ -108,8 +105,8 @@ class SessionUpdate(object):
         with self.sess.as_default():
             for name, value in six.iteritems(prms):
                 assert name in self.name_map
-                for v in self.name_map[name]:
-                    SessionUpdate.load_value_to_var(v, value)
+                v = self.name_map[name]
+                SessionUpdate.load_value_to_var(v, value)
 
 
 def dump_session_params(path):
@@ -168,6 +165,9 @@ def dump_chkpt_vars(model_path):
 
     Args:
         model_path(str): path to a checkpoint.
+
+    Returns:
+        dict: a name:value dict
     """
     model_path = get_checkpoint_path(model_path)
     reader = tf.train.NewCheckpointReader(model_path)
@@ -199,6 +199,6 @@ def is_training_name(name):
         return True
     if name.endswith('/Adagrad'):
         return True
-    if 'EMA_summary/' in name:
+    if name.startswith('/EMA'):
         return True
     return False