fix action stop gradient

ml-agents/mlagents/trainers/policy/transfer_policy.py

         :param encoded_state: Tensor corresponding to encoded current state.
         :param encoded_next_state: Tensor corresponding to encoded next state.
         """
+
+        if separate_train:
+           encoded_state = tf.stop_gradient(encoded_state)
+
-
-        if separate_train:
-           hidden = tf.stop_gradient(hidden)
 
         for i in range(forward_layers):
             hidden = tf.layers.dense(
         separate_train: bool = False
     ):
 
+        if separate_train:
+           encoded_state = tf.stop_gradient(encoded_state)
+
-        if separate_train:
-           hidden = tf.stop_gradient(hidden)
+        
         for i in range(forward_layers):
             hidden = tf.layers.dense(
                 hidden,