NEURAL_SAVE is not called with PARAMETERS+RULES

Posted By: GPEngine

NEURAL_SAVE is not called with PARAMETERS+RULES - 07/06/15 05:27

http://zorro-trader.com/manual/en/training.htm says
Quote:
The following cases are possible:
Rules depend on a parameter. ... Set the RULES and PARAMETERS flags at the same time. The training run will now repeat through many cycles. At first the parameter will be optimized and for every parameter step a new set of rules will be generated. ...
I find that when combining PARAMETERS with RULES with an adviseLong(NEURAL, ...), neural(mode=NEURAL_SAVE, ...) is not called as expected at the end of each optimize step. But, neural(model=NEURAL_LOAD,.. ) is called. Therefore Zorro dies when it tries to load an non-existing ml file since it wasn't produced, or it happily re-uses one from some old experiment.
Posted By: GPEngine

Re: NEURAL_SAVE is not called with PARAMETERS+RULES - 07/06/15 05:38

A work-around is to use my own naming scheme for the ml files and save the model as part of NEURAL_TRAIN. This way, the right file is present when NEURAL_LOAD happens.
Posted By: GPEngine

Re: NEURAL_SAVE is not called with PARAMETERS+RULES - 07/06/15 06:12

workaround fails because 'StepCycle' is undeclared identifier in 1.32. I need that to avoid naming collisions.
Posted By: jcl

Re: NEURAL_SAVE is not called with PARAMETERS+RULES - 07/06/15 11:09

WFOCycle is normally used for rule naming. Optimize steps are counted with g->nStepCycle. Thanks for the report, we will look into combining NEURAL with PARAMETERS.
© 2024 lite-C Forums