October 2018
Intermediate to advanced
246 pages
6h 26m
English
Once we have our dataset ready, we need to look into improving the results. We can do this by altering the number of steps in the learning process. The simplest way to do this is to try the following:
--how_many_training_steps = 4000
The rate of accuracy improvement slows down when the number of steps increases, and the accuracy will stop improving beyond a certain point. You can experiment with this and decide what works best for you.
Read now
Unlock full access