News for hackers. Stuff that matters.
Re:PyTorch – Tensors and Dynamic neural networks in Python by attractivechaos on Thursday January 19, 2017 @07:36PM (#13440805)
One question: how do you save a dynamic network if it changes from time to time (e.g. from sample to sample)?
Re:PyTorch – Tensors and Dynamic neural networks in Python by apaszke on Sunday January 22, 2017 @09:58AM (#13455388)
You save the parameters and the code of the model definition