diff --git a/qadence_libs/qinfo_tools/qng.py b/qadence_libs/qinfo_tools/qng.py
index 16baab2..f2aa5a7 100644
--- a/qadence_libs/qinfo_tools/qng.py
+++ b/qadence_libs/qinfo_tools/qng.py
@@ -215,29 +215,29 @@ def qng_spsa(
 
     See :class:`~qadence_libs.qinfo_tools.QuantumNaturalGradient` for details.
     """
+    with torch.no_grad():
+        # Get estimation of the QFI matrix
+        vparams_dict = dict(zip(vparams_keys, vparams_values))
+        qfi_estimator, qfi_mat_positive_sd = get_quantum_fisher_spsa(
+            circuit=circuit,
+            iteration=state["iter"],
+            vparams_dict=vparams_dict,
+            previous_qfi_estimator=state["qfi_estimator"],
+            epsilon=epsilon,
+            beta=beta,
+        )
 
-    # Get estimation of the QFI matrix
-    vparams_dict = dict(zip(vparams_keys, vparams_values))
-    qfi_estimator, qfi_mat_positive_sd = get_quantum_fisher_spsa(
-        circuit=circuit,
-        iteration=state["iter"],
-        vparams_dict=vparams_dict,
-        previous_qfi_estimator=state["qfi_estimator"],
-        epsilon=epsilon,
-        beta=beta,
-    )
-
-    # Get transformed gradient vector solving the least squares problem
-    transf_grad = torch.linalg.lstsq(
-        0.25 * qfi_mat_positive_sd,
-        grad_vec,
-        driver="gelsd",
-    ).solution
+        # Get transformed gradient vector solving the least squares problem
+        transf_grad = torch.linalg.lstsq(
+            0.25 * qfi_mat_positive_sd,
+            grad_vec,
+            driver="gelsd",
+        ).solution
 
-    for i, p in enumerate(vparams_values):
-        if p.grad is None:
-            continue
-        p.data.add_(transf_grad[i], alpha=-lr)
+        for i, p in enumerate(vparams_values):
+            if p.grad is None:
+                continue
+            p.data.add_(transf_grad[i], alpha=-lr)
 
-    state["iter"] += 1
-    state["qfi_estimator"] = qfi_estimator
+        state["iter"] += 1
+        state["qfi_estimator"] = qfi_estimator