Gridnet diagram in readme is mismatching from code

[a-shivgan]

There is incorrect representation of number of channels in the residual block expansion's second convolution layer. The number of input channels to the second convolution block should be '2c' and the number of output channels should be 'c'. In the code, it is maintaining the constant number of channels horizontally, but the same is not reflected in the diagram.