notebook.community

Edit and run



In [1]:

    
import tensorflow as tf
import pandas as pd
import numpy as np



In [2]:

    
a = tf.constant([5,3], name="input_a")
b = tf.reduce_prod(a, name="prod_b")
c = tf.reduce_sum(a, name="sum_c")
d = tf.add(b,c, name="add_d")



In [5]:

    
sess = tf.Session()
sess.run(d)









    Out[5]:





23



In [4]:

    
sess.run(c)









    Out[4]:





8



In [5]:

    
output = sess.run(e)



In [7]:

    
writer = tf.train.SummaryWriter('./my_graph', sess.graph)



In [8]:









    



---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-8-f2669e6f1eb9> in <module>()
----> 1 tensorboard

NameError: name 'tensorboard' is not defined



In [10]:

    
# Initialize some tensors to use in computation
a = np.array([2, 3], dtype=np.int32)
b = np.array([4, 5], dtype=np.int32)

#use tf.add() to init an 'add' operation
c = tf.add(a,b, name="fuck_add_jamal")

Let's create a Graph object



In [11]:

    
g = tf.Graph()



In [12]:

    
#You can then add Operations to it by using the Graph.as_default() method.



In [15]:

    
with g.as_default():
    #declare operations here like normal, and they'll be added to the graph, g.
    #Basically creates a local namespace for nodes.
    pass #passing so this cell will run

One can access the default graph (basically the global graph space) by just setting tf.get_default_graph() to a variable.

It is also possible to load in previously defined models from other TensorFlow scripts and assign them to Graph objects using a combo of the Graph.as_graph_def() and tf.import_graph_def() functions.



In [ ]: