search

NVlabs / GA3C

Hybrid CPU/GPU implementation of the A3C algorithm for deep reinforcement learning.
BSD 3-Clause "New" or "Revised" License
652 stars 195 forks source link

should conduct padding before training? #37

Open tangbohu opened 6 years ago

tangbohu commented 6 years ago

Hello! In file ThreadTrain.py x = np.concatenate((x, x_))

As x_ may shorter than TMAX, should we conduct padding before concatenate?