Understanding Conv2d

Is out_channels in Conv2d is equal to number of filters? In Keras we can specify number of filters, how to do it in Pytorch?

Yes, out_channels equals the number of filters in the conv layer.