diff --git a/sml/metrics/regression/regression.py b/sml/metrics/regression/regression.py
index 503f1f52..e480be50 100644
--- a/sml/metrics/regression/regression.py
+++ b/sml/metrics/regression/regression.py
@@ -14,8 +14,6 @@
 from sklearn import metrics
 import jax.numpy as jnp
 
-import spu.spu_pb2 as spu_pb2
-import spu.utils.simulation as spsim
 
 
 def _mean_tweedie_deviance(y_true, y_pred, sample_weight, power, d2_score=False):
@@ -45,7 +43,7 @@ def _mean_tweedie_deviance(y_true, y_pred, sample_weight, power, d2_score=False)
             - temp_power * (y_true / temp - y_pred / (temp + 1))
         )
 
-    if d2_score & (sample_weight is None):
+    if d2_score and (sample_weight is None):
         # When weight is none and the d2 fraction is calculated, the numerator and denominator are divided by shape, and only sum is used
         return jnp.sum(dev)
     else: