Use float in cosine metric final calculation in default vectorization…

… provider (#358)

Makes the implementation consistent with Panama and native vectorization providers, which use float. Removes one of the sources of potentially different results for different providers.

Fixes #357

jvector-base/src/main/java/io/github/jbellis/jvector/vector/DefaultVectorUtilSupport.java

@@ -135,7 +135,7 @@ public float cosine(VectorFloat<?> av, VectorFloat<?> bv) {
       norm1 += elem1 * elem1;
       norm2 += elem2 * elem2;
     }
-    return (float) (sum / Math.sqrt((double) norm1 * (double) norm2));
+    return (float) (sum / Math.sqrt(norm1 * norm2));
   }
 
   @Override
@@ -152,7 +152,7 @@ public float cosine(VectorFloat<?> av, int aoffset, VectorFloat<?> bv, int boffs
       norm1 += elem1 * elem1;
       norm2 += elem2 * elem2;
     }
-    return (float) (sum / Math.sqrt((double) norm1 * (double) norm2));
+    return (float) (sum / Math.sqrt(norm1 * norm2));
   }
 
   @Override