search

dd604 / refinedet.pytorch

A PyTorch inplementation of RefineDet
MIT License
36 stars 12 forks source link

About training #3

Open zeroorhero opened 4 years ago

zeroorhero commented 4 years ago

Hello,thank you for your code! when i am training ,there is no error .but can not train.

CUDA devices: <class 'torch.cuda.device'> GPU numbers: 1 Create or load an imdb. Loaded dataset voc_2007_trainval for training Preparing training data... voc_2007_trainval gt roidb loaded from /home/lcq/下载/residual-refinedet.pytorch-master/~/data/cache/voc_2007_trainval_gt_roidb.pkl done before filtering, there are 146 images... after filtering, there are 146 images... Construct vgg16_refinedet network. /home/lcq/下载/residual-refinedet.pytorch-master/libs/networks/refinedet.py:34: UserWarning: volatile was removed and now has no effect. Use with torch.no_grad(): instead. self.anchors = Variable(self.anchor_layer.forward(), volatile=True) Loading pretrained weights from weights/vgg16/vgg16_reducedfc.pth

/home/lcq/下载/residual-refinedet.pytorch-master/libs/networks/vgg_refinedet.py(56)_init_modules() -> assert key_layer_ids[2] == -1 and key_layer_ids[3] == -1, \ (Pdb)

do you know why? thank you !

dd604 commented 4 years ago

Does this happen when you dubug? Maybe this means that you use wrong vgg. You should use the vgg network in this project, other than another vgg from pytorch.

dd604 commented 4 years ago

Hello,thank you for your code! when i am training ,there is no error .but can not train.

CUDA devices: <class 'torch.cuda.device'> GPU numbers: 1 Create or load an imdb. Loaded dataset voc_2007_trainval for training Preparing training data... voc_2007_trainval gt roidb loaded from /home/lcq/下载/residual-refinedet.pytorch-master/~/data/cache/voc_2007_trainval_gt_roidb.pkl done before filtering, there are 146 images... after filtering, there are 146 images... Construct vgg16_refinedet network. /home/lcq/下载/residual-refinedet.pytorch-master/libs/networks/refinedet.py:34: UserWarning: volatile was removed and now has no effect. Use with torch.no_grad(): instead. self.anchors = Variable(self.anchor_layer.forward(), volatile=True) Loading pretrained weights from weights/vgg16/vgg16_reducedfc.pth

/home/lcq/下载/residual-refinedet.pytorch-master/libs/networks/vgg_refinedet.py(56)_init_modules() -> assert key_layer_ids[2] == -1 and key_layer_ids[3] == -1, (Pdb)

do you know why? thank you !

Hi my friend, i have just updated the project, maybe there is some mess in the previous version. If there are some problem about the new version, please let me know. Thank you.