Update docs/qinfo_tools/qng.md

Co-authored-by: Roland-djee <9250798+Roland-djee@users.noreply.github.com>

docs/qinfo_tools/qng.md

@@ -12,7 +12,7 @@ $$
   F_{i j}(\theta)=-\left.2 \frac{\partial}{\partial \theta_i} \frac{\partial}{\partial \theta_j}\left|\left\langle\psi\left(\theta^{\prime}\right) \mid \psi(\theta)\right\rangle\right|^2\right|_{{\theta}^{\prime}=\theta}
 $$
 
-However, computing the above expression is a costly operation scaling quadratically with the number of parameters in the variational quantum circuit. It is thus usual to use approximate methods when dealing with the QFI matrix. Qadence provides a SPSA-based implementation of the Quantum Natural Gradient[^3]. The [SPSA](https://www.jhuapl.edu/spsa/) (Simultaneous Perturbation Stochastic Approximation) algorithm is a well known gradient-based algorithm based on finite differences. QNG-SPSA constructs an iterative approximation to the QFI matrix with a constant number of circuit evaluations that does not scale with the number of parameters. Although the SPSA algorithm outputs a rough approximation of the QFI matrix, the QNG-SPSA has been proven to work well while being a very efficient method due to the constant overhead in circuit evaluations (only 6 extra evaluations per iteration).
+However, computing the above expression is a costly operation scaling quadratically with the number of parameters in the variational quantum circuit. It is thus usual to use approximate methods when dealing with the QFI matrix. Qadence-Libs provides a SPSA-based implementation of the Quantum Natural Gradient[^3]. The [SPSA](https://www.jhuapl.edu/spsa/) (Simultaneous Perturbation Stochastic Approximation) algorithm is a well known finite differences-based algorithm. QNG-SPSA constructs an iterative approximation to the QFI matrix with a constant number of circuit evaluations that does not scale with the number of parameters. Although the SPSA algorithm outputs a rough approximation of the QFI matrix, the QNG-SPSA has been proven to work well while being a very efficient method due to the constant overhead in circuit evaluations (only 6 extra evaluations per iteration).
 
 In this tutorial, we use the QNG and QNG-SPSA optimizers with the Quantum Circuit Learning algorithm, a variational quantum algorithm which uses Quantum Neural Networks as universal function approximators.