Can I change the target in the criterion for each iteration?
outputs = net(inputs)
loss = criterion(outputs,outputs+inputs)
Is it possible to update the weights and biases?
or Should I bring the reinforcement learning concept?
Thank you!
Can I change the target in the criterion for each iteration?
outputs = net(inputs)
loss = criterion(outputs,outputs+inputs)
Is it possible to update the weights and biases?
or Should I bring the reinforcement learning concept?
Thank you!