I need your help!

[lucatagliamonte]

I added the state variable mass to take account of the mass of propellant burnt throughout the flight. Now the model doesn't seem to learn much after a while and the reward averages the same positive value. I am at 150k episodes and I can make it land much slower and upright but oftentimes it misses the target. Is there something you suggest doing?
By the way, I'm so grateful you created this model. Impressive