# Save and Load model **URL:** \ **Category:** Uncategorized\ **Created:** [August 14, 2017, 1:58pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206 "2017-08-14T13:58:02Z")\ **Posts on this page:** 11\ **Page:** 2
**Author:** ![MisBah\_Awan](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/m/e480ec/32.png) [@MisBah\_Awan](https://discuss.pytorch.org/u/MisBah_Awan)\ **Post date:** [April 19, 2019, 11:57am UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/21 "2019-04-19T11:57:12Z")
here i save my model: ![image](https://discuss.pytorch.org/uploads/default/original/2X/2/2d6eba500bcfcbfb567060009f2b0ec4fa9dc71b.png) and now i am stuck here ![image](https://discuss.pytorch.org/uploads/default/original/2X/6/64f355b2f46476b2ee2ff672af0bef761e166bda.png) ---
**Author:** ![ptrblck](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ptrblck/32/1823_2.png) [@ptrblck](https://discuss.pytorch.org/u/ptrblck)\ **Post date:** [April 19, 2019, 1:16pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/22 "2019-04-19T13:16:21Z")
This code works: ```python x = torch.randn(1, 3, 64, 64) # (batch size or #of images,channels RGB,width,height) model = Net() output = model(x) torch.save({'state_dict': model.state_dict()}, 'tmp.pt') model = Net() state_dict = torch.load('tmp.pt')['state_dict'] model.load_state_dict(state_dict) print(model) ``` Could you try to find differences between your code and mine? ---
**Author:** ![MisBah\_Awan](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/m/e480ec/32.png) [@MisBah\_Awan](https://discuss.pytorch.org/u/MisBah_Awan)\ **Post date:** [April 19, 2019, 1:25pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/23 "2019-04-19T13:25:48Z")
you mean first i save my model according to ur code and then load it? ---
**Author:** ![ptrblck](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ptrblck/32/1823_2.png) [@ptrblck](https://discuss.pytorch.org/u/ptrblck)\ **Post date:** [April 19, 2019, 1:28pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/24 "2019-04-19T13:28:11Z")
I assume you have already saved your model, so my code is just to show that initializing the model and loading the `state_dict` does work and the error might still be somewhere else in your code. ---
**Author:** ![Sunil\_Tapashetti](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/sunil_tapashetti/32/7177_2.png) [@Sunil\_Tapashetti](https://discuss.pytorch.org/u/Sunil_Tapashetti)\ **Post date:** [October 11, 2019, 5:47pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/25 "2019-10-11T17:47:11Z")
If a pretrained model is saved as Dataparallel can it be loaded as nn.Sequential model? ---
**Author:** ![ptrblck](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ptrblck/32/1823_2.png) [@ptrblck](https://discuss.pytorch.org/u/ptrblck)\ **Post date:** [November 24, 2019, 7:17pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/26 "2019-11-24T19:17:32Z")
`nn.DataParallel` stores the actual model in the `.module` attribute. If you store the `state_dict` from the `model.module`, you’ll be able to directly use it without wrapping the model into `nn.DataParallel`. ---
**Author:** ![pinocchio](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/pinocchio/32/22093_2.png) [@pinocchio](https://discuss.pytorch.org/u/pinocchio)\ **Post date:** [April 29, 2020, 4:09pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/27 "2020-04-29T16:09:54Z")
# what are the advantages and disadvantages of just using the python pickle @ptrblck what are the advantages and disadvantages of just using the python pickle module vs the load/save stuff pytorch has built? > [@Advantages & Disadvantages of using pickle module to save models vs torch.save](https://discuss.pytorch.org/t/advantages-disadvantages-of-using-pickle-module-to-save-models-vs-torch-save/79016): > > I was reading [Save and Load model](https://discuss.pytorch.org/t/save-and-load-model/6206/27) but it wasn’t clear why I’d use torch.save over pickle.dump. What worries me is that my Neural Net modules/object have many more things inside of them besides only parameters. Since pytorch saves things using the state\_dict that worried me that something might be missing ([https://pytorch.org/tutorials/beginner/saving\_loading\_models.html](https://pytorch.org/tutorials/beginner/saving_loading_models.html)). Isn’t for my use case easier to simply use the pickle.dump function? Why would I use torch.save in general? Note I am usin… * * * > [@How does one pickle arbitrary pytorch models that use lambda functions?](https://discuss.pytorch.org/t/how-does-one-pickle-arbitrary-pytorch-models-that-use-lambda-functions/79026): > > I currently have a neural network module: import torch.nn as nn class NN(nn.Module): def \_\_init\_\_(self,args,lambda\_f,nn1, loss, opt): super().\_\_init\_\_() self.args = args self.lambda\_f = lambda\_f self.nn1 = nn1 self.loss = loss self.opt = opt # more nn.Params stuff etc... def forward(self, x): #some code using fields return out I am trying to checkpoint it but because pytorch saves using state\_dicts it means I can… ---
**Author:** ![ptrblck](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ptrblck/32/1823_2.png) [@ptrblck](https://discuss.pytorch.org/u/ptrblck)\ **Post date:** [April 30, 2020, 4:26am UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/28 "2020-04-30T04:26:10Z")
Internally `pickle` (or a custom pickle module you provide) will be used, so you could also manually store the data most likely. However, the [`_save` method](https://github.com/pytorch/pytorch/blob/e75fb4356b752097d093c7013ba85c9eb82961ef/torch/serialization.py#L439) will apply a specific format to load it properly. I’ll answer in the other threads separately, but I would recommend to keep “similar” questions in one place. 😉 ---
**Author:** ![pinocchio](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/pinocchio/32/22093_2.png) [@pinocchio](https://discuss.pytorch.org/u/pinocchio)\ **Post date:** [April 30, 2020, 2:49pm UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/29 "2020-04-30T14:49:31Z")
> [@ptrblck](#): > > I’ll answer in the other threads separately, but I would recommend to keep “similar” questions in one place. 😉 I’m trying! That’s why I try to reference questions between each other but I don’t have control that other people ask repeated questions ☹ I will try to not make this issue worse. Appreciate the advice. 🙂 ---
**Author:** ![saba](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/saba/32/19439_2.png) [@saba](https://discuss.pytorch.org/u/saba)\ **Post date:** [August 30, 2021, 3:40am UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/30 "2021-08-30T03:40:30Z")
Hi Ptrblck, I want to save my data which are 2000x11x11x11 tensor. I used torch.save(TargetGAN,Path, \_use\_new\_zipfile\_serialization=False) but gave me this error : TypeError: save() got an unexpected keyword argument ‘\_use\_new\_zipfile\_serialization’ I need to use \_use\_new\_zipfile\_serialization=False, otherwise could not load the data on the GPU and will consider it as a zip file which I don’t want. Many thanks ---
**Author:** ![ptrblck](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ptrblck/32/1823_2.png) [@ptrblck](https://discuss.pytorch.org/u/ptrblck)\ **Post date:** [August 30, 2021, 4:31am UTC](https://discuss.pytorch.org/t/save-and-load-model/6206/31 "2021-08-30T04:31:34Z")
This argument as added in PyTorch [`1.4.0`](https://pytorch.org/docs/1.4.0/torch.html#torch.save), so you might want to update your Pytorch version, if you want to use it. [Previous page](https://discuss.pytorch.org/t/save-and-load-model/6206.md?page=1)