Better error catching in exploration
jajcayn opened this issue · 0 comments
jajcayn commented
If one of the runs in exploration fails, we only have a log stating which run fails (its ID) and nothing else. For debugging purposes, we could catch errors in evaluating function and log them...
The problem is, now if even one run fails, you cannot load results and as such you have no idea what went wrong. By catching errors (i.e. logging run id with its respective error) one would be able to fix errors quickly.