pykan

pykan copied to clipboard

What does `reg` mean when training?

3

Hi! I would like to ask what does `reg` mean when training KAN? For example you would see something like: `train loss: xxx | test loss: yyy | reg: zzz...

egmaminta

The loss did not decrease when training on the observations

6

I trained the following model using inputs with [1600,4] obtained from observations ``` model = KAN(width=[x.shape[1],4,4,1], grid=3, k=3, seed=0, device=device_set, ) model.update_grid_from_samples(dataset['train_input']) grids = [3,5,10,20,50] train_rmse = [] test_rmse =...

wkqian06

How can I get the output if the given layer ?

1

Younger-ger

Performance of KAN in some image classification datasets

3

Compare the performance of pykan, efficient-kan, and FastKAN with MLP on image classification tasks using the MNIST, CIFAR10, and CIFAR100 datasets.Additionally, include the differences in training time between FastKAN and...

Boreuseful

I wanner fit KAN with multi gpus, how to make it by nn.DataParallel(model)?

BraudoCC

The out put of the KAN model is a matrix fulled of (NaN) in second time iterations

4

I replace the MLP of NeRF,But get the NaN loss because of the KAN model's output initialize KAN: `net = KAN(width=[84,1,4], grid=5, k=3, seed=0,device=device)` #----KAN------ device = pts_flat.device pts_flat =...

linmi1

Request for advice on tunning of hyper parameters

1

I am trying to tunning the hyperparameters with optuna. Accordingly, the objective for optimization should be assigned. The AICc is currently considered, though advice about how to reasonably and conveniently...

zhongjingjogy

Kayzwer

Continual learning with KAN

1

Regardless of computaion, do KAN at present able perform continual learning on 2D input data (or its flattened 1d vector features)? if not what is the main challenge KAN have...

lukmanulhakeem97

PDE Solving Demo?

2

I've noticed that in the PDE solving demo the ground truth solution for the PDE is used to evaluate the boundary condition loss and is thus used in the training...

cwoolfo1