Torch. NN. Linear() function

 

torch.nn.Linear(in_features, out_features, bias=True) Function is a linear transformation function ：

among ,in_features Enter the sample size for ,out_features Is the size of the output sample ,bias The default is true. If you set bias = false Then this layer will not learn an additive bias .

Linear() Function is usually used to set the Fully connected layer .

Use cases ：

import torch

x = torch.randn(128, 20)  #  The input samples 
fc = torch.nn.Linear(20, 30)  # 20 Enter the sample size for ,30 Is the output sample size 
output = fc(x)
print('fc.weight.shape:\n ', fc.weight.shape)
print('fc.bias.shape:\n', fc.bias.shape)
print('output.shape:\n', output.shape)

ans = torch.mm(x,torch.t(fc.weight))+fc.bias  #  The results are in agreement with fc(x) identical 
print('ans.shape:\n', ans.shape)

print(torch.equal(ans, output))

Running results ：

m.weight.shape:
  torch.Size([30, 20])
m.bias.shape:
 torch.Size([30])
output.shape:
 torch.Size([128, 30])
ans.shape:
 torch.Size([128, 30])
true