DeepX-inc / machina

Control section: Deep Reinforcement Learning framework
MIT License
278 stars 45 forks source link

Fast sampling for random batch #197

Closed rarilurelo closed 5 years ago

rarilurelo commented 5 years ago

This PR achieved faster learning speed as Traj's size increased.