TorchCurves adds differentiable B-spline and Legendre curves to PyTorch
Coefficients can train with a surrounding model, with timings against NumPy, SciPy and EfficientKAN.
3 stories, newest first · one card per story
Coefficients can train with a surrounding model, with timings against NumPy, SciPy and EfficientKAN.
The open-source tool, forked from TensorFlow Playground, lets users edit B-spline activations and watch gradient descent.
An empirical study shows distributed Kolmogorov-Arnold Network training reaches 74.7% parallel efficiency at 8 A100 GPUs, with overheads driven by All-Reduce choices.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views