diff --git a/egs/librispeech/ASR/pruned_transducer_stateless2/scaling.py b/egs/librispeech/ASR/pruned_transducer_stateless2/scaling.py
index 598c621cd..78ee5da7b 100644
--- a/egs/librispeech/ASR/pruned_transducer_stateless2/scaling.py
+++ b/egs/librispeech/ASR/pruned_transducer_stateless2/scaling.py
@@ -667,7 +667,7 @@ class ProjDrop(torch.nn.Module):
                             device=x.device, dtype=x.dtype)
             rr = torch.matmul(r, r.t())  # num_dropped by num_dropped
             rr += 0.01  # to 100% ensure it is invertible
-            rr_inv = rr.cholesky().to(torch.float32).cholesky_inverse().to(x.dtype)
+            rr_inv = rr.to(torch.float32).cholesky().cholesky_inverse().to(x.dtype)
             # OK, so r rr_inv r.t() will have eigenvalues of 1.
             xr = torch.matmul(x, r.t()) # (..., num_dropped)
             rr_inv_r = torch.matmul(rr_inv, r) # (num_dropped, num_channels)