fix(lint): fix lint warning

example/wide_n_deep/follower.py

@@ -122,7 +122,8 @@ def model_fn(model, features, labels, mode):
         train_op = model.minimize(
             optimizer, act1_f, grad_loss=gact1_f, global_step=global_step)
         final_ops = final_fn(model=model, tensor_name='reflux_embedding',
-                                is_send=False, assignee=peer_embeddings, shape=[num_slot,fid_size,embed_size])
+                                is_send=False, assignee=peer_embeddings,
+                                shape=[num_slot, fid_size, embed_size])
         embedding_hook = tf.train.FinalOpsHook(final_ops=final_ops)
         return model.make_spec(mode, loss=tf.math.reduce_mean(act1_f),
                                training_chief_hooks=[embedding_hook],

example/wide_n_deep/leader.py

@@ -69,7 +69,7 @@ def final_fn(model, tensor_name, is_send, tensor=None, shape=None):
     if is_send:
         assert tensor, "Please specify tensor to send"
         if DEBUG_PRINT:
-           ops.append(tf.print(tensor))
+            ops.append(tf.print(tensor))
         ops.append(model.send_no_deps(tensor_name, tensor))
         return ops
 
@@ -161,7 +161,8 @@ def model_fn(model, features, labels, mode):
         {"loss" : loss}, every_n_iter=10)
     metric_hook = flt.GlobalStepMetricTensorHook(tensor_dict={"loss": loss},
                                                  every_steps=10)
-    final_ops = final_fn(model=model, tensor_name='reflux_embedding',is_send=True,tensor=embeddings)
+    final_ops = final_fn(model=model, tensor_name='reflux_embedding',
+                         is_send=True, tensor=embeddings)
     embedding_hook = tf.train.FinalOpsHook(final_ops=final_ops)
 
     optimizer = tf.train.GradientDescentOptimizer(0.1)

fedlearner/trainer/estimator.py

@@ -87,7 +87,8 @@ def send_no_deps(self, name, tensor):
         self._sends.append((name, tensor, False))
         return send_op
 
-    def recv_no_deps(self, name, dtype=tf.float32, require_grad=False, shape=None):
+    def recv_no_deps(self,
+                     name, dtype=tf.float32, require_grad=False, shape=None):
         receive_op = self._bridge.receive_op(name, dtype)
         if shape:
             receive_op = tf.ensure_shape(receive_op, shape)

fedlearner/trainer/run_hooks.py

@@ -243,4 +243,4 @@ def _parse_op_label(self, label):
             inputs = []
         else:
             inputs = inputs.split(', ')
-        return nn, op, inputs
+        return nn, op, inputs