PyTorch implementation of Advantage Actor Critic (A2C), Proximal Policy Optimization (PPO), Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation (ACKTR) and Generative Adversarial Imitation Learning (GAIL).
MIT License
3.57k
stars
829
forks
source link
my forward function get different output dimensions from the function defined in _init_ #212
I modified the model.py, where I use a Conv1d function in the CNNBase to replace the original neural network, here is the code:
the printed information is :
the problem is that in the forward function, the output dimension of the actor is expected to be 3 but gets 4.
What's the problem here?
I modified the model.py, where I use a Conv1d function in the CNNBase to replace the original neural network, here is the code: the printed information is : the problem is that in the forward function, the output dimension of the actor is expected to be 3 but gets 4. What's the problem here?