updated args

pytorch_lightning/pt_overrides/override_data_parallel.py

@@ -83,7 +83,6 @@ def parallel_apply(modules, inputs, kwargs_tup=None, devices=None):
             with lock:
                 results[i] = e
 
-    pdb.set_trace()
     if len(modules) > 1:
         threads = [threading.Thread(target=_worker,
                                     args=(i, module, input, kwargs, device))
@@ -97,6 +96,7 @@ def parallel_apply(modules, inputs, kwargs_tup=None, devices=None):
     else:
         _worker(0, modules[0], inputs[0], kwargs_tup[0], devices[0])
 
+    pdb.set_trace()
     outputs = []
     for i in range(len(inputs)):
         output = results[i]