Why are HIDDEN_SIZE and NUM_STEPS tied to 128?

[satpreetsingh]

Hi, why are the RNN HIDDEN_SIZE and NUM_STEPS tied to 128? Other values seem to throw tensor reshaping errors.

Tried on IPPO and MAPPO with MPE, e.g. with NUM_STEPS=256:

TypeError: cannot reshape array of shape (256, 100) (size 25600) into shape (128, 100) (size 12800)

I don't immediately see a reason from the algorithm why these should be tied up.

[amacrutherford]

Hey! This was a simplification we made when collecting baselines, it is fixed in this file by adding config["GRU_HIDDEN_DIM"]. We are looking to update the rest of the scripts over the next month.

[amacrutherford]

fixed with #77