Tony Downward

Thanks for clarifying this. Perhaps I didn't implement this right, but I'm still unable to get this to work. First I tried to set the `optimizer = nothing` when defining...

Update: by commenting out the `run_numerical_stability_report` code, I was able to debug into the `master_loop`, but once the `remotecall`s started, I could not debug those to see if the `init_callback`...

Thanks for this - this works for me. I'll adjust this one step at a time and try to figure out where I'm going wrong.

There must be something wrong with the way my model is set up, I'm taking my policy graph and training it exactly as in the example you provided, but it...

The issue was that there should be an `@everywhere` in front of this: ```julia function remote_solver() env = Gurobi.Env() return () -> Gurobi.Optimizer(env) end ``` But due to the way...

The two main issues (which are probably specific to my use case) that I think could be clearer are: (a) where `@everywhere` needs to be used - in the end...

I'm using the function that loads Gurobi into each of the processes, and the difference is between these two cases: **Case A** ```julia @everywhere function parallel_optimizer() env = Gurobi.Env() return...

Some additional information: If I use CPLEX: ```julia @everywhere function parallel_optimizer() return CPLEX.Optimizer end ``` the all the processes return cuts, but if I set an attribute: ```julia @everywhere function...

I'm not sure why the workers showed at 6-9. I had thought I shutdown Julia between runs. Regardless, I'm pretty sure that the issue is not related to that. (This...

I think that a sufficient condition for the model to be able to reach a steady state is that there there is a positive probability that you can transition from...