The first step is to include jumps in the stock price. Then, you can easily add jumps into the variance process. If you only consider seldom, large jumps, you may want to use a jump-diffusion like the models from Merton (1976) and Kou (2002). The former uses a log-normal distribution for the jump size whilst Kou employs a double exponential distribution.
"Dates must be sorted unique". "Unique", so you can't have repeated dates in the inputs you're passing. You have two copies of 1Y. You probably wanted the second to be 2Y.
"Sorted", so you can't have 18M before 1Y. They have to be in the correct order in time.