diff --git a/codes/sweep.py b/codes/sweep.py
index fa47cc2a..18688695 100644
--- a/codes/sweep.py
+++ b/codes/sweep.py
@@ -7,6 +7,17 @@ import torch
 
 from train import Trainer
 from utils import options as option
+import collections.abc
+
+
+def deep_update(d, u):
+    for k, v in u.items():
+        if isinstance(v, collections.abc.Mapping):
+            d[k] = deep_update(d.get(k, {}), v)
+        else:
+            d[k] = v
+    return d
+
 
 def launch_trainer(opt, opt_path=''):
     rank = opt['gpu_ids'][0]
@@ -18,6 +29,7 @@ def launch_trainer(opt, opt_path=''):
     trainer.init(opt_path, opt, 'none')
     trainer.do_training()
 
+
 if __name__ == '__main__':
     """
     Ad-hoc script (hard coded; no command-line parameters) that spawns multiple separate trainers from a single options
@@ -38,7 +50,7 @@ if __name__ == '__main__':
     all_opts = []
     for i, (mod, mod_dict) in enumerate(modifications.items()):
         nd = copy.deepcopy(opt)
-        nd.update(mod_dict)
+        deep_update(nd, mod_dict)
         opt['gpu_ids'] = [i]
         nd['name'] = f'{nd["name"]}_{mod}'
         nd['wandb_run_name'] = mod