Blealtan / efficient-kan

An efficient pure-PyTorch implementation of Kolmogorov-Arnold Network (KAN).

Blealtan/efficient-kan Issues

torch._C._LinAlgError: torch.linalg.lstsq: (Batch element 1): The least squares solution could not be computed because the input matrix does not have full rank (error code: 4).
Updated 2 months ago2
Kolmogorov-Arnold Transformer
Updated 2 months ago
flops and params
Updated 2 months ago1
Can still plot model and prune model in efficient kan?
Updated 2 months ago12
Something different from the official results for KAN.
Updated 3 months ago1
Constant loss, network is not learning
Closed 4 months ago3
Layer fails on inputs that have been transposed previously
Closed 4 months ago1
How to cite
Updated 5 months ago1
Questions
Updated 5 months ago
I would like to add a train test function to the KAN class
Updated 5 months ago
which types of images could be better recognised using KAN approach?
Updated 5 months ago
block: [0,0,0], thread: [1,0,0] Assertion `input_val >= zero && input_val <= one` failed. when using BCEloss
Updated 5 months ago
How to pip install?
Closed 5 months ago3
Can configure efficient-kan model for continual learning?
Updated 6 months ago9
ONNX
Updated 6 months ago
Is there anyway that I can save the model for later use
Closed 6 months ago7
Are the "efficient-kan" and "official-kan" equivalent in terms of algorithms?
Closed 6 months ago12
What is the mathematic equation behind the model?
Closed 6 months ago4
KAN intuition for setting the hidden layers
Closed 6 months ago1
Can MLP replace transformer?
Closed 6 months ago1
KANs are MLPs? MLP Equivalent
Closed 6 months ago2
Train Multi-KANs simultaneously! Coding realization
Closed 6 months ago2
Symbolic Formulation from efficient-KAN
Closed 6 months ago1
AssertionError
Updated 6 months ago1
please report a bug to PyTorch. torch.linalg.lstsq: (Batch element 0): Argument 6 has illegal value. Most certainly there is a bug in the implementation calling the backend library.
Closed 6 months ago1
CUDA out of memory.
Closed 6 months ago2
Difference in memory usage of mlp and kan
Updated 6 months ago1
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [800, 256]], which is output 0 of AsStridedBackward0, is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
Updated 6 months ago1
Symbolic expression support
Closed 6 months ago1
KAN has more learnable parameters?
Closed 6 months ago2
Can anyone put kan(attention and mlp) into llama2.c?
Closed 6 months ago1
time complexity
Closed 6 months ago2
Contributing?
Closed 6 months ago3
RuntimeError: false INTERNAL ASSERT FAILED
Updated 6 months ago3
Missing license
Closed 6 months ago2
Intel oneMKL ERROR: Parameter 6 was incorrect on entry to SGELSY.
Updated 6 months ago7
Supports inputs more than 2 dimensions For efficient-KAN
Closed 6 months ago5
Necessity of the scale parameter?
Closed 6 months ago2