PyTorchZeroToAll
PyTorchZeroToAll copied to clipboard
Update 02_manual_gradient.py
Good catch. Thanks!
If we can add losses and do avg, I think it's better.
Can you also review the code in 12_X, and 13_X?