How to save and use?
jmlDevelop opened this issue · comments
Hello, I have some questions about training beginners. I didn't see the part in the code where the weights are saved after training. Also, how do I use the saved weights for inference?
This is NOT a GOOD trainning implementation, JUST for DEEEEEEEEMOOOOOOOOO training logic.
You need optimizate the code: validtion logic to save better checkpoint, scheduler for training, implementation of patchs loss, ...
Of course, you can use torch.save and torch.load(code in infer(.)) to reuse weights.