diff --git a/python/paddle/fluid/incubate/fleet/parameter_server/pslib/optimizer_factory.py b/python/paddle/fluid/incubate/fleet/parameter_server/pslib/optimizer_factory.py
index caaaa88cc4c6435fe46ba8622bebbf16abf33f30..607a3c94f8a4e7c14fa98ebb8de45241d97da437 100644
--- a/python/paddle/fluid/incubate/fleet/parameter_server/pslib/optimizer_factory.py
+++ b/python/paddle/fluid/incubate/fleet/parameter_server/pslib/optimizer_factory.py
@@ -313,7 +313,7 @@ class DistributedAdam(DistributedOptimizerImplBase):
             if op.type in self.supported_embedding_types:
                 if op.attr('is_distributed') is True:
                     table_name = op.input("W")[0]
-                    emb_size = local_vars[table_name].shape[1]
+                    emb_size = local_vars[table_name].shape[-1]
                     if d_size.get(table_name) is None:
                         d_size[table_name] = emb_size
                     elif d_size[table_name] != emb_size:
@@ -328,13 +328,10 @@ class DistributedAdam(DistributedOptimizerImplBase):
             strategy[table_name] = dict()
         st = strategy[table_name]
 
-        accessor = None
+        accessor = "DownpourCtrAccessor"
         if st.get("sparse_accessor_class") is not None:
             accessor = st["sparse_accessor_class"]
 
-        if accessor is None:
-            accessor = "DownpourCtrAccessor"
-
         # set sparse_embedx_dim in strategy,
         # user do not have to set it in config_fleet
         if accessor == "DownpourFeatureValueAccessor" \
@@ -344,12 +341,12 @@ class DistributedAdam(DistributedOptimizerImplBase):
             if st.get("sparse_embedx_dim") is not None \
                     and st["sparse_embedx_dim"] != emb_to_size[table_name] - 3:
                 raise ValueError("fleet config sparse_embedx_dim=%s not"
-                                 " equal to embedding size - 3 = %s" %
+                                 " equal to embedding dim - 3 = %s" %
                                  (st["sparse_embedx_dim"],
                                   emb_to_size[table_name] - 3))
             if st.get("sparse_embedx_dim") is None:
                 logger.warning(
-                    "sparse embedding size for table name '{}' is: {}, while sparse_embedx_dim "
+                    "sparse embedding dim for table name '{}' is: {}, while sparse_embedx_dim "
                     "with same sparse table name is not set in config_fleet.py. "
                     "Hence automatically set sparse_embedx_dim = {} - 3.".
                     format(table_name, emb_to_size[table_name], emb_to_size[
@@ -359,12 +356,12 @@ class DistributedAdam(DistributedOptimizerImplBase):
             if st.get("sparse_embedx_dim") is not None \
                     and st["sparse_embedx_dim"] != emb_to_size[table_name]:
                 raise ValueError("fleet config sparse_embedx_dim=%s not"
-                                 " equal to embedding size = %s" %
+                                 " equal to embedding dim = %s" %
                                  (st["sparse_embedx_dim"],
                                   emb_to_size[table_name]))
             if st.get("sparse_embedx_dim") is None:
                 logger.warning(
-                    "sparse embedding size for table name '{}' is: {}, while sparse_embedx_dim "
+                    "sparse embedding dim for table name '{}' is: {}, while sparse_embedx_dim "
                     "with same sparse table name is not set in config_fleet.py. "
                     "Hence automatically set sparse_embedx_dim = {}.".format(
                         table_name, emb_to_size[table_name], emb_to_size[