galeone
commented
7 years ago Hi! I can't reproduce your error but while I was trying to run your example I found out another error that I fixed. Maybe this fix could solve your problem too.
Update dytb to the new version and let me know if it's solved!
wingz1
Hi, I'm having a similar problem to the last one I got when running an example model: Any ideas? Thanks!
dytb_train --model SingleLayerCAE --dataset MNIST --optimizer AdamOptimizer --optimizer_args '{"learning_rate": 1e-5}' --train_device "/gpu:1" --batch_size 1024 --l2_penalty 1e-9 Extracting ...anaconda3/lib/python3.6/site-packages/dytb/inputs/predefined/data/MNIST/train-images-idx3-ubyte.gz Extracting ...anaconda3/lib/python3.6/site-packages/dytb/inputs/predefined/data/MNIST/train-labels-idx1-ubyte.gz Extracting ...anaconda3/lib/python3.6/site-packages/dytb/inputs/predefined/data/MNIST/t10k-images-idx3-ubyte.gz Extracting ...anaconda3/lib/python3.6/site-packages/dytb/inputs/predefined/data/MNIST/t10k-labels-idx1-ubyte.gz Args: { 'batch_size': 1024, 'checkpoint_path': '', 'comment': '', 'dataset': 'MNIST', 'epochs': 150, 'exclude_scopes': None, 'l2_penalty': 1e-09, 'lr_decay': False, 'lr_decay_epochs': 25, 'lr_decay_factor': 0.1, 'model': 'SingleLayerCAE', 'optimizer': 'AdamOptimizer', 'optimizer_args': {'learning_rate': 1e-05}, 'restart': False, 'train_device': '/gpu:1', 'trainable_scopes': None} Traceback (most recent call last): File "...anaconda3/bin/dytb_train", line 66, in
sys.exit(main())
File "...anaconda3/bin/dytb_train", line 56, in main
comment=ARGS.comment)
File "...anaconda3/lib/python3.6/site-packages/dytb/train.py", line 217, in train
return Trainer(model, dataset, args, steps, paths).train()
File "...anaconda3/lib/python3.6/site-packages/dytb/trainer/Trainer.py", line 76, in train
l2_penalty=self._args["regularizations"]["l2"])
TypeError: get() got multiple values for argument 'train_phase'