Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

TensorArray Not Used on line 240 of mrt_utils.py #128

Open
CodeSmileBot opened this issue Jul 7, 2023 · 0 comments
Open

TensorArray Not Used on line 240 of mrt_utils.py #128

CodeSmileBot opened this issue Jul 7, 2023 · 0 comments

Comments

@CodeSmileBot
Copy link

Hello!

I found an AI-Specific Code smell in your project.
The smell is called: TensorArray Not Used

You can find more information about it in this paper: https://dl.acm.org/doi/abs/10.1145/3522664.3528620.

According to the paper, the smell is described as follows:

Problem If the developer initializes an array using tf.constant() and tries to assign a new value to it in the loop to keep it growing, the code will run into an error. The developer can fix this error by the low-level tf.while_loop() API. However, it is inefficient coding in this way. A lot of intermediate tensors are built in this process.
Solution Using tf.TensorArray() for growing array in the loop is a better solution for this kind of problem in TensorFlow 2.
Impact Efficiency, Error-proneness

Example:

### TensorFlow
import tensorflow as tf
@tf.function
def fibonacci(n):
   a = tf.constant(1)
   b = tf.constant(1)
-    c = tf.constant([1, 1])
+    c = tf.TensorArray(tf.int32, n)
+    c = c.write(0, a)
+    c = c.write(1, b)

   for i in range(2, n):
       a, b = b, a + b
-       c = tf.concat([c, [b]], 0)
+		c = c.write(i, b)

-    return c
+	 return c.stack()

You can find the code related to this smell in this link:

score: evaluation score, 1D tensor with size: (sub-batch_size, sampleN)
Return:
expected risk per real sentence over a sub-batch, a scalar tensor
"""
samplesN = config.samplesN
total_sample = tf.shape(input=cost)[0]
batch_size = tf.shape(input=index)[0] - tf.constant(1)
# cancelling the negative of the cost (P**alpha = e**(-alpha*(-logP))
alpha = tf.constant([-config.mrt_alpha], dtype=tf.float32)
cost = tf.multiply(cost, alpha)
# normalise costs
i = tf.constant(0)
def while_condition(i, _):
return tf.less(i, batch_size)
def body(i, cost):
normalised_cost = tf.nn.softmax(cost[index[i]: index[i+1]])
# assign value of sub-tensor to a tensor iteratively
.

I also found instances of this smell in other files, such as:

File: https://github.com/EdinburghNLP/nematus/blob/master/nematus/layers.py#L122-L132 Line: 127
File: https://github.com/EdinburghNLP/nematus/blob/master/nematus/layers.py#L127-L137 Line: 132
.

I hope this information is helpful!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant