Identical samples during mcmc run in Turing.jl with NUTS algorithm

Hello Everyone, Im running some differential Equation based model for inference using Turing.jl

Im running 3 chains in parallel. Following result is with 2000 warmup and 2000 samples. Though Im not sure whats going on here. Chains not converging I understand but here its just going straight (so same samples during run).

So, main question is does someone has any idea in what situation this can be the case?

I am sure issue is in my model setup but I need a starting point to solve this issue

Thank you