I am implementing a multichannel CNN. It is in time series data. and i am implementitng 1D convolution. I could not find how i can define the number of kernels. Through search i am aware that the number of output channel is equal to number of filters. But i have output channels already defined. which is 252. Shall i just omit 252 and put my desired number of filters in there? which is 8 for the first conv layer and for the second is 4. Also do i have to write nn.sequential for the second layer too ? This is my code below:
self.conv1 = nn.Sequential(
nn.Conv1d(in_channels=256, out_channels=252, kernel_size=5),
nn.softmax(),
nn.Avgpool1d(kernel_size= 2, stride=2),
nn.Dropout(p=0.25))
self.conv2 = nn.Conv1d(in_channels=126, out_channels=122, kernel_size= 5),
nn.softmax(),
nn.Avgpool1d(kernel_size= 2, stride=2),
nn.Dropout(p=0.25)