[NNVM][FRONTEND] Review comments.

nnvm/python/nnvm/testing/tf.py

@@ -0,0 +1,79 @@
+# pylint: disable=invalid-name, unused-variable, unused-argument, no-init
+"""
+Compile Tensorflow Models
+=========================
+Some helper definitions for tensorflow models.
+"""
+import re
+
+# Tensorflow imports
+import tensorflow as tf
+
+######################################################################
+# Some helper functions
+# ---------------------
+
+class NodeLookup(object):
+    """Converts integer node ID's to human readable labels."""
+
+    def __init__(self,
+                 label_lookup_path=None,
+                 uid_lookup_path=None):
+        self.node_lookup = self.load(label_lookup_path, uid_lookup_path)
+
+    def load(self, label_lookup_path, uid_lookup_path):
+        """Loads a human readable English name for each softmax node.
+
+        Parameters
+        ----------
+        label_lookup_path: String
+            File containing String UID to integer node ID mapping .
+
+        uid_lookup_path: String
+            File containing String UID to human-readable string mapping.
+
+        Returns
+        -------
+        node_id_to_name : dict
+            dict from integer node ID to human-readable string.
+
+        """
+        if not tf.gfile.Exists(uid_lookup_path):
+            tf.logging.fatal('File does not exist %s', uid_lookup_path)
+        if not tf.gfile.Exists(label_lookup_path):
+            tf.logging.fatal('File does not exist %s', label_lookup_path)
+
+        # Loads mapping from string UID to human-readable string
+        proto_as_ascii_lines = tf.gfile.GFile(uid_lookup_path).readlines()
+        uid_to_human = {}
+        p = re.compile(r'[n\d]*[ \S,]*')
+        for line in proto_as_ascii_lines:
+            parsed_items = p.findall(line)
+            uid = parsed_items[0]
+            human_string = parsed_items[2]
+            uid_to_human[uid] = human_string
+
+        # Loads mapping from string UID to integer node ID.
+        node_id_to_uid = {}
+        proto_as_ascii = tf.gfile.GFile(label_lookup_path).readlines()
+        for line in proto_as_ascii:
+            if line.startswith('  target_class:'):
+                target_class = int(line.split(': ')[1])
+            if line.startswith('  target_class_string:'):
+                target_class_string = line.split(': ')[1]
+                node_id_to_uid[target_class] = target_class_string[1:-2]
+
+        # Loads the final mapping of integer node ID to human-readable string
+        node_id_to_name = {}
+        for key, val in node_id_to_uid.items():
+            if val not in uid_to_human:
+                tf.logging.fatal('Failed to locate: %s', val)
+            name = uid_to_human[val]
+            node_id_to_name[key] = name
+
+        return node_id_to_name
+
+    def id_to_string(self, node_id):
+        if node_id not in self.node_lookup:
+            return ''
+        return self.node_lookup[node_id]

tutorials/nnvm/from_tensorflow.py

@@ -14,16 +14,16 @@
 import tvm
 import numpy as np
 import os.path
-import re
 
 # Tensorflow imports
 import tensorflow as tf
 from tensorflow.core.framework import graph_pb2
 from tensorflow.python.framework import dtypes
 from tensorflow.python.framework import tensor_util
 
+import nnvm.testing.tf
 
-repo_base = 'https://github.com/srkreddy1238/dmlc_data/raw/master/models/tensorflow/InceptionV1/'
+repo_base = 'https://github.com/dmlc/web-data/raw/master/tensorflow/models/InceptionV1/'
 img_name = 'elephant-299.jpg'
 image_url = os.path.join(repo_base, img_name)
 model_name = 'classify_image_graph_def-with_shapes.pb'
@@ -35,10 +35,36 @@
 
 
 ######################################################################
-# Some helper functions
+# Download processed tensorflow model
+# -----------------------------------
+# In this section, we download a pretrained Tensorflow model and classify an image.
+from mxnet.gluon.utils import download
+
+download(image_url, img_name)
+download(model_url, model_name)
+download(map_proto_url, map_proto)
+download(lable_map_url, lable_map)
+
+
+######################################################################
+# Creates graph from saved graph_def.pb.
+# --------------------------------------
 
 def _ProcessGraphDefParam(graph_def):
-    """Type-checks and possibly canonicalizes `graph_def`."""
+    """Type-checks and possibly canonicalizes `graph_def`.
+
+    Parameters
+    ----------
+    graph_def : Obj
+        tensorflow graph definition.
+
+    Returns
+    -------
+    graph_def : Obj
+        tensorflow graph devinition
+
+    """
+
     if not isinstance(graph_def, graph_pb2.GraphDef):
         # `graph_def` could be a dynamically-created message, so try a duck-typed
         # approach
@@ -50,84 +76,6 @@ def _ProcessGraphDefParam(graph_def):
             raise TypeError('graph_def must be a GraphDef proto.')
     return graph_def
 
-class NodeLookup(object):
-    """Converts integer node ID's to human readable labels."""
-
-    def __init__(self,
-                 label_lookup_path=None,
-                 uid_lookup_path=None):
-        if not label_lookup_path:
-            label_lookup_path = os.path.join(
-                "./", map_proto)
-        if not uid_lookup_path:
-            uid_lookup_path = os.path.join(
-                "./", lable_map)
-        self.node_lookup = self.load(label_lookup_path, uid_lookup_path)
-
-    def load(self, label_lookup_path, uid_lookup_path):
-        """Loads a human readable English name for each softmax node.
-
-        Args:
-            label_lookup_path: string UID to integer node ID.
-            uid_lookup_path: string UID to human-readable string.
-
-        Returns:
-            dict from integer node ID to human-readable string.
-        """
-        if not tf.gfile.Exists(uid_lookup_path):
-            tf.logging.fatal('File does not exist %s', uid_lookup_path)
-        if not tf.gfile.Exists(label_lookup_path):
-            tf.logging.fatal('File does not exist %s', label_lookup_path)
-
-        # Loads mapping from string UID to human-readable string
-        proto_as_ascii_lines = tf.gfile.GFile(uid_lookup_path).readlines()
-        uid_to_human = {}
-        p = re.compile(r'[n\d]*[ \S,]*')
-        for line in proto_as_ascii_lines:
-            parsed_items = p.findall(line)
-            uid = parsed_items[0]
-            human_string = parsed_items[2]
-            uid_to_human[uid] = human_string
-
-        # Loads mapping from string UID to integer node ID.
-        node_id_to_uid = {}
-        proto_as_ascii = tf.gfile.GFile(label_lookup_path).readlines()
-        for line in proto_as_ascii:
-            if line.startswith('  target_class:'):
-                target_class = int(line.split(': ')[1])
-            if line.startswith('  target_class_string:'):
-                target_class_string = line.split(': ')[1]
-                node_id_to_uid[target_class] = target_class_string[1:-2]
-
-        # Loads the final mapping of integer node ID to human-readable string
-        node_id_to_name = {}
-        for key, val in node_id_to_uid.items():
-            if val not in uid_to_human:
-                tf.logging.fatal('Failed to locate: %s', val)
-            name = uid_to_human[val]
-            node_id_to_name[key] = name
-
-        return node_id_to_name
-
-    def id_to_string(self, node_id):
-        if node_id not in self.node_lookup:
-            return ''
-        return self.node_lookup[node_id]
-
-######################################################################
-# Download processed tensorflow model
-# ---------------------------------------------
-# In this section, we download a pretrained Tensorflow model and classify an image.
-from mxnet.gluon.utils import download
-
-download(image_url, img_name)
-download(model_url, model_name)
-download(map_proto_url, map_proto)
-download(lable_map_url, lable_map)
-
-
-######################################################################
-# Creates graph from saved graph_def.pb.
 with tf.gfile.FastGFile(os.path.join(
         "./", model_name), 'rb') as f:
     graph_def = tf.GraphDef()
@@ -138,6 +86,7 @@ def id_to_string(self, node_id):
 
 ######################################################################
 # Decode image
+# ------------
 from PIL import Image
 image = Image.open(img_name).resize((299, 299))
 
@@ -150,7 +99,7 @@ def transform_image(image):
 
 ######################################################################
 # Import the graph to NNVM
-# -----------------
+# ------------------------
 sym, params = nnvm.frontend.from_tensorflow(graph_def)
 
 ######################################################################
@@ -165,6 +114,7 @@ def transform_image(image):
 
 ######################################################################
 # Save the compilation output.
+# ----------------------------
 """
 lib.export_library("imagenet_tensorflow.so")
 with open("imagenet_tensorflow.json", "w") as fo:
@@ -197,7 +147,8 @@ def transform_image(image):
 predictions = np.squeeze(predictions)
 
 # Creates node ID --> English string lookup.
-node_lookup = NodeLookup()
+node_lookup = nnvm.testing.tf.NodeLookup(label_lookup_path=os.path.join("./", map_proto),
+                                         uid_lookup_path=os.path.join("./", lable_map))
 
 top_k = predictions.argsort()[-5:][::-1]
 for node_id in top_k:
@@ -221,10 +172,13 @@ def create_graph():
 def run_inference_on_image(image):
     """Runs inference on an image.
 
-    Args:
-        image: Image file name.
+    Parameters
+    ----------
+    image: String
+        Image file name.
 
-    Returns:
+    Returns
+    -------
         Nothing
     """
     if not tf.gfile.Exists(image):
@@ -242,7 +196,8 @@ def run_inference_on_image(image):
         predictions = np.squeeze(predictions)
 
         # Creates node ID --> English string lookup.
-        node_lookup = NodeLookup()
+        node_lookup = nnvm.testing.tf.NodeLookup(label_lookup_path=os.path.join("./", map_proto),
+                                                 uid_lookup_path=os.path.join("./", lable_map))
 
         top_k = predictions.argsort()[-5:][::-1]
         print ("===== TENSORFLOW RESULTS =======")