diff --git a/thop/profile.py b/thop/profile.py
index 4b98364..59dbfdc 100644
--- a/thop/profile.py
+++ b/thop/profile.py
@@ -2,7 +2,7 @@
 
 from thop.vision.basic_hooks import *
 from thop.rnn_hooks import *
-
+from copy import deepcopy
 
 # logger = logging.getLogger(__name__)
 # logger.setLevel(logging.INFO)
@@ -182,13 +182,12 @@ def add_hooks(m: nn.Module):
             handler_collection[m] = (m.register_forward_hook(fn), m.register_forward_hook(count_parameters))
         types_collection.add(m_type)
 
-    prev_training_status = model.training
-
-    model.eval()
-    model.apply(add_hooks)
+    model_cpy = deepcopy(model)
+    model_cpy.eval()
+    model_cpy.apply(add_hooks)
 
     with torch.no_grad():
-        model(*inputs)
+        model_cpy(*inputs)
 
     def dfs_count(module: nn.Module, prefix="\t") -> (int, int):
         total_ops, total_params = 0, 0
@@ -206,14 +205,8 @@ def dfs_count(module: nn.Module, prefix="\t") -> (int, int):
         #  print(prefix, module._get_name(), (total_ops.item(), total_params.item()))
         return total_ops, total_params
 
-    total_ops, total_params = dfs_count(model)
+    total_ops, total_params = dfs_count(model_cpy)
 
-    # reset model to original status
-    model.train(prev_training_status)
-    for m, (op_handler, params_handler) in handler_collection.items():
-        op_handler.remove()
-        params_handler.remove()
-        m._buffers.pop("total_ops")
-        m._buffers.pop("total_params")
+    del model_cpy
 
     return total_ops, total_params