[ENH] Add flexible number of agents feature for training

[pr-shukla]

• Currently this implementation supports only 3 agents during training, with this enhancement it can support any number of agents decided by user during training

[pr-shukla]

Issue resolved with PR #7