search

hytseng0509 / CrossDomainFewShot

Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation (ICLR 2020 spotlight)
331 stars 62 forks source link

Question about Training w/ learning-to-learned feature-wise transformations. #34

Open kongjiangbo opened 3 years ago

kongjiangbo commented 3 years ago

Hello Dr. Tseng, Thanks for releasing your code for LFT. Namespace(data_dir='./filelists', dataset='multi', method='gnnnet', model='ResNet12', n_shot=5, name='tmp', num_classes=200, resume='resnet12/', resume_epoch=-1, save_dir='./output', save_freq=1, start_epoch=0, stop_epoch=400, test_n_way=5, testset='plantae', train_aug=True, train_n_way=5, warmup='gg3b0') Traceback (most recent call last): File "train.py", line 103, in start_epoch = model.resume(resume_file) File "/home/huangjianping/CrossDomainFewShot-master/methods/LFTNet.py", line 225, in resume self.model.load_state_dict(state['model_state']) KeyError: 'model_state' help me!