神经元输入为0的原因

fangwei123456 / spikingjelly

SpikingJelly is an open-source deep learning framework for Spiking Neural Network (SNN) based on PyTorch.

https://spikingjelly.readthedocs.io

Other

1.35k stars 239 forks source link

Closed moxueyingyan closed 2 years ago

moxueyingyan commented 2 years ago

我自定义了一个神经元模型，使用MNISTNet()训练，设了1024epoch，epoch0~7的训练结果都正常，从epoch8开始精度突然从0.9922下降到0.098，发现是神经元的输入x变为0导致，但是我不清楚输入变为0从哪里找原因，请指点一下！多谢！

fangwei123456 commented 2 years ago

建议检查一下各层的输入输出

moxueyingyan commented 2 years ago

不好意思，没弄明白怎么查看，以ParametricLIFNode为例，能指点一下么？非常感谢！

fangwei123456 commented 2 years ago