Julia Gusak

@yxlijun, everything works fine for me ``` model = PeleeNet() model_dict = model.state_dict() state_dict = torch.load("weights/peleenet_acc7208.pth.tar", map_location='cpu')['state_dict'] pretrained_dict = {k.strip('module.'): v for k, v in state_dict.items() if k.strip('module.') in model_dict}...

Hi @aupuzikov and @aswanthkrishna, thank you for noticing this bug! Now it is fixed

Hello @joskaaaa , we have experimented with two pre-trained Faster R-CNN models, one from the [facebook repository](https://github.com/facebookresearch/maskrcnn-benchmark) and another one from [here](https://github.com/chenyuntc/simple-faster-rcnn-pytorch). In both models ROI part is not serializable....