Remove anonymous tuple conversion in training_loop_test.

PiperOrigin-RevId: 322377616

tensorflow_federated/python/research/utils/training_loop_test.py

@@ -31,15 +31,6 @@ _Batch = collections.namedtuple('Batch', ['x', 'y'])
 FLAGS = flags.FLAGS
 
 
-def _from_tff_result(state):
-  return tff.learning.framework.ServerState(
-      model=tff.learning.ModelWeights(
-          list(state.model.trainable), list(state.model.non_trainable)),
-      optimizer_state=list(state.optimizer_state),
-      delta_aggregate_state=[],
-      model_broadcast_state=[])
-
-
 class BasicAdapter(adapters.IterativeProcessPythonAdapter):
   """Converts iterative process results from anonymous tuples."""
 
@@ -47,13 +38,10 @@ class BasicAdapter(adapters.IterativeProcessPythonAdapter):
     self._iterative_process = iterative_process
 
   def initialize(self):
-    initial_state = self._iterative_process.initialize()
-    return _from_tff_result(initial_state)
+    return self._iterative_process.initialize()
 
   def next(self, state, data):
     state, metrics = self._iterative_process.next(state, data)
-    state = _from_tff_result(state)
-    metrics = metrics._asdict(recursive=True)
     outputs = None
     return adapters.IterationResult(state, metrics, outputs)