Regarding data loaders

when i load images and store every image in a list and we have like 4gb of images.Does that mean 4gb of ram in my pc will be utilized when i doing this work.
But when we run model.fit(x_train,y_train,batch_size=32)
as batch_size is 32 so every time when i am training my model only 32 images will be considered for updating the weights and those 32 images will get load in my ram