optim() function is converged, but gradients are not close enough to zero

Question

I wrote the code below for function optimization using optim. The convergence code is zero which shows the function is optimized successfully, and the estimated parameters are a