-
在resnet18_cifar10.py中main函数里,通过一个链接下载SJ-cifar10-resnet18_model-sample.pth,想问一下,这个文件是训练好的cifar10_resnet18模型么?我想换个数据集或者换成VGG网络的话,这个模型该如何获取?或者如何训练?@DingJianhao
-
-
We can reduce memory consumption by storing spikes as bool tensor for backward. Here is an example of using linear:
```python
import torch
import torch.nn as nn
import torch.nn.functional as F
…
-
-
您好, 我使用的数据集格式如下:
![image](https://user-images.githubusercontent.com/44899736/141928587-f657301a-bc20-4687-a7df-99ad6ed99336.png)
SpikingJelly给的案例均为图片数据集, 因此想请问如何替换为三维数组的数据集? 报错信息如下
![image](https…
-
when I use function 'split_to_train_test_set()' like :
dataset = CIFAR10DVS(DVSCIFAR10str, data_type = 'frame', frames_number=steps, split_by = 'number',
transform = No…
-
I was interested in using a vanilla RNN with LIF neurons. However, it seems the current implementation of `SpikingVanillaRNNCell` does implement any membrane dynamics.
https://github.com/fangwei12…
-
-
最近我注意到 SpikingJelly 稳定版更新到了 0.00006 ,其中使用 CUDA 增强的神经元发生了较大的变化,从 cext 模块转移到了 clock_driven 模块。SpikingJelly 的文档中提供的示例所展示的性能也有一定的区别。
0.0.0.0.6 版本中的性能图示如下:
![](https://spikingjelly.readthedocs.io/zh_…
-