diff --git a/optim_nn.py b/optim_nn.py
index a57fe95..9e601aa 100644
--- a/optim_nn.py
+++ b/optim_nn.py
@@ -11,6 +11,9 @@ lament = lambda *args, **kwargs: print(*args, file=sys.stderr, **kwargs)
 def log(left, right):
     lament("{:>20}:   {}".format(left, right))
 
+class Dummy:
+    pass
+
 # Loss functions {{{1
 
 class SquaredHalved(Loss):
@@ -304,7 +307,7 @@ def toy_data(train_samples, valid_samples, problem=2):
 
     return (inputs, outputs), (valid_inputs, valid_outputs)
 
-def model_from_config(config, input_features, output_features):
+def model_from_config(config, input_features, output_features, callbacks):
     # Our Test Model
 
     init = inits[config.init]
@@ -344,7 +347,7 @@ def model_from_config(config, input_features, output_features):
         raise Exception('unknown optimizer', config.optim)
 
     def rscb(restart):
-        measure_error() # declared later...
+        callbacks.restart()
         log("restarting", restart)
         if config.restart_optim:
             optim.reset()
@@ -472,8 +475,10 @@ def run(program, args=[]):
     input_features  =  inputs.shape[-1]
     output_features = outputs.shape[-1]
 
+    callbacks = Dummy()
+
     model, learner, ritual, (loss, mloss) = \
-      model_from_config(config, input_features, output_features)
+      model_from_config(config, input_features, output_features, callbacks)
 
     # Model Information
 
@@ -515,6 +520,8 @@ def run(program, args=[]):
         train_losses.append(train_err)
         valid_losses.append(valid_err)
 
+    callbacks.restart = measure_error
+
     measure_error()
 
     assert inputs.shape[0] % config.batch_size == 0, \