The strength error of the crystal structure calculated by the stress-strain tensile script is large

We use the stress-strain script to calculate the stress-strain curve. The same model has been repeated for many times, and the strength values calculated each time have a big difference. May I ask experts and engineers why this is