Aug 29, 2017 · 1 min read
Hi Tran — thanks for the tutorial!
Q: I’ve run the train and eval processes and i see that the eval process ended after +- 45 min while the train process ended after 7 hours.
So in the end i get a graph of precision that is missing the interesting part — the end.
How can i measure the precision of the last part of the train? and why the eval part was so short? can you help me with that?