Bias tells us the complexity of the model, how many parameters there are, and trained, whereas variance shows that how much our model can adjust to the change in the dataset.