in the test(), when the code is running this step qs.plots.snapshot(returns.Balance, title='RL Trader Performance'), sometimes it poped up zero Division Error, that means zero has been divided inside some steps.
after that, I tried to add try except to avoid this kind of errors. but in the learning output, I can see these two params are stating 0.0
| approxkl | 0.0 |
| clipfrac | 0.0
does anyone know what happened on this situation ?
in the test(), when the code is running this step qs.plots.snapshot(returns.Balance, title='RL Trader Performance'), sometimes it poped up zero Division Error, that means zero has been divided inside some steps.
after that, I tried to add try except to avoid this kind of errors. but in the learning output, I can see these two params are stating 0.0 | approxkl | 0.0 | | clipfrac | 0.0
does anyone know what happened on this situation ?