pytorch-a3c
pytorch-a3c copied to clipboard
PyTorch implementation of Asynchronous Advantage Actor Critic (A3C) from "Asynchronous Methods for Deep Reinforcement Learning".
``` if shared_param.grad is not None: return shared_param._grad = param.grad ``` in this code, if there's no `if shared_param.grad is not None: return` shared_param._grad will be assigned each time? maybe...
I follow the readme to run python3 main.py --env-name "PongDeterministic-v4" --num-processes 16 but I have no idea to find the result, could you help me?
Process w10: Traceback (most recent call last): File "D:\Software\Industrial\Anaconda\lib\multiprocessing\process.py", line 315, in _bootstrap self.run() File "D:\AICodeFile\pytorch-A3C-master\continuous_A3C.py", line 89, in run a = self.lnet.choose_action(v_wrap(s[None ,: ])) TypeError: tuple indices must be...
I cloned this project on unbuntu, but there were a lot of problems when I tried to run the code. Could you please provide python version, pytorch version, numpy version...