Hello. I am new to Pytorch tensor manipulations. I have created an energy management environment for my DQN, which is a simple network like this with the forward method:
`class DQN(nn.Module):
def __init__(self, n_actions):
self.n_actions = len([i for i in range (10, 100, 5)])
super().__init__()
self.fc1 = nn.Linear(in_features=5, out_features=24)
self.fc2 = nn.Linear(in_features=24, out_features=32)
self.out = nn.Linear(in_features=32, out_features= n_actions)
def forward(self, t):
t = t.unsqueeze_(5)
t = t.flatten(start_dim=1)
t = F.relu(self.fc1(t))
t = F.relu(self.fc2(t))
t = self.out(t)
return t`
Now, each state is a list/vector of 5 numbers (floats), the actions are integer numbers in the range(10, 100, 5). How do I feed these states without getting mismatch errors?