If we use 'tf. cond' the number of outputs should be the same for both true_fn and false_ fn. But in my case IF part returns some outputs, ELSE part returns nothing.. Plz suggest some solution. Thanks a lot
"returns nothing": so how is the output of the conditional used? Perhaps what you are trying to do is better done by a Mask.
Alternatively, would returning zeroes be appropriate?
Chris Morris how can we do it with mask?
and, however I manage to get same number of outputs for True_fn and False_fn.
True_fn modifies some variables and return them. False_fn does not modify and return the variables as it is.
But while checking with Print statement, I came to know that both True_fn and False_fn are getting executed.. I don't know the reason..
I tried in the following format:
[a,b] = tf.cond(tf.math.less(c,d),lambda:fn1(e,f,g,h),lambda:fn2(e,f,g,h))
Could you please clarify me.. Sorry I am new to tensorflow..Thanks
Please refer this handbook for some basic reference.
- https://riptutorial.com/Download/tensorflow.pdf
- Badges
- Science topic
- Similar topics
- Clinical Pharmacology
- Dosing