ppo: add entropy cost annealing (linear + cosine schedules)#660
Open
PhysicistJohn wants to merge 1 commit intogoogle:mainfrom
Open
ppo: add entropy cost annealing (linear + cosine schedules)#660PhysicistJohn wants to merge 1 commit intogoogle:mainfrom
PhysicistJohn wants to merge 1 commit intogoogle:mainfrom