report (draft Apr 24)

[cesare-spinoso]

• 1 Introduction: @etiennedemers

• 2 Methods: @etiennedemers @j-c-carr

• Introductory sentences, for each of these algos we did hyp grid search i.e. everything in common place here

• 2.1 VPG: Explain details of algo, hyperparameters, for the ones we saw in class less detail, include links openai, show more the motivation, explain implementation details of OPenAI as well

• 2.4 TD3: Explain theory a bit more since we didn't cover it

• 2.5 SAC: Theory + variant

• 3 Results:

• 3.1 and 3.2 @sabina-elkins

• PLots: train best agent plots for all 5 agents, + cumulative reward, run train_best_agent.py

• 3.3 Adaptability @etiennedemers

Moral support: @cesare-spinoso

• Video: Start, middle and end of VPG (for illustration) and then the evaluation video for each of them

[j-c-carr]

Aim for first draft done on Sunday (Apr 24) night The plan is to do the video based on the paper content

[cesare-spinoso]

1. Review report: @cesare-spinoso @j-c-carr @sabina-elkins @etiennedemers
2. Section 2: @j-c-carr Section 3: @cesare-spinoso + @etiennedemers (adaptability)
3. Conclusion @j-c-carr
4. Skim down

[cesare-spinoso]

Final submission