search

AlphaZeroIncubator / AlphaZero

Our implementation of AlphaZero for simple games such as Tic-Tac-Toe and Connect4.
0 stars 0 forks source link

Training/testing functions #3

Closed guidopetri closed 4 years ago

guidopetri commented 4 years ago

We should have easy to use training/testing functions for our model where we pass in e.g. the model, the optimizer, number of epochs, etc.

PhilipEkfeldt commented 4 years ago

Like I mentioned in the call, I would suggest using PyTorchLightning for this as it makes managing experiments and logging much easier. Haven't used it with RL though, I imagine the data generation adds complexity to the process.

I also found this which could be useful: RL with Pytorch Lightning