阅读量:0
It is fun to play with KANs, just the same it is fun to play computer games. A common frustration in both games is that one did something wrong but cannot restore to the lastest checkpoint. We provide a quick way to save and load your checkpoint, so that you won't be frustrated and think that you need to start all over again.
玩kan和玩电脑游戏一样好玩，用checkpoint相当于存档，做错了事情可以恢复
from kan import KAN, create_dataset import torch import torch.nn # create a KAN: 2D inputs, 1D output, and 5 hidden neurons. cubic spline (k=3), 5 grid intervals (grid=5). model = KAN(width=[2,5,1], grid=5, k=3, seed=0, base_fun=torch.nn.SiLU()) f = lambda x: torch.exp(torch.sin(torch.pi*x[:,[0]]) + x[:,[1]]**2) dataset = create_dataset(f, n_var=2) model(dataset['train_input']) model.plot() model.save_ckpt('ckpt1') #model.clear_ckpts() # save intialized model as ckpt1save this model
