search

fangwei123456 / spikingjelly

SpikingJelly is an open-source deep learning framework for Spiking Neural Network (SNN) based on PyTorch.
https://spikingjelly.readthedocs.io
Other
1.24k stars 235 forks source link

About the MNIST classification demo #478

Open des-zhong opened 7 months ago

des-zhong commented 7 months ago
  1. Why is the initial classification accuracy on MNIST dataset is up to 89% even before any training has taken place
  2. the command "from spikingjelly.activation_based.ann2snn.sample_models import mnist_cnn", is it not implemented yet?
fangwei123456 commented 7 months ago

Why is the initial classification accuracy on MNIST dataset is up to 89%

It is not the initial accuracy, but the accuracy after one epoch (epoch = 0).

mnist_cnn", is it not implemented yet?

THis file only conatins a network, without training and converting codes.