modify so3 operation implementation to make united tests pass due to the update of sympy

changes.md

@@ -10,7 +10,7 @@ nav_order: 3
 - Improve the memory efficiency and speed of three-body interactions. (@kenko911)
 - FrechetCellFilter is added for variable cell relaxation in Relaxer class. (@kenko911)
 - Smooth l1 loss function is added for training. (@kenko911)
-  
+
 ## 1.1.2
 - Move AtomRef Fitting to numpy to avoid bug (@BowenD-UCB)
 - NVE ensemble added (@kenko911)

src/matgl/utils/so3.py

@@ -72,7 +72,7 @@ def generate_clebsch_gordan(lmax: int) -> torch.Tensor:
                     m2 - m1,
                     -m1 - m2,
                 }:
-                    coeff = clebsch_gordan(l1, l2, l3, m1, m2, m3)
+                    coeff = clebsch_gordan(l1.item(), l2.item(), l3.item(), m1.item(), m2.item(), m3.item())
                     cg[c1, c2, c3] = float(coeff)
     return cg