pass read_from_replicas arg when creating pipeline

This will allow us to take advantage of `read_from_replicas` support on pipelines when Grokzen/redis-py-cluster#450 is merged

baseplate/clients/redis_cluster.py

@@ -195,6 +195,7 @@ def pipeline(self, name: str) -> "MonitoredClusterRedisPipeline":
             self.server_span,
             self.connection_pool,
             self.response_callbacks,
+            read_from_replicas=self.read_from_replicas,
         )
 
     # No transaction support in redis-py-cluster