search

sraashis / retinal-fundus-transfer

An example of easytorch implementation on retinal vessel segmentation.
26 stars 4 forks source link

Error in main.py #11

Closed fmoore closed 3 years ago

fmoore commented 3 years ago

When launching "main.py" the following error message is returned:

Starting with the following parameters: {'batch_size': 4, 'dataset_dir': 'datasets', 'epochs': 31, 'force': False, 'gpus': [0], 'grad_accum_iters': 1, 'learning_rate': 0.001, 'load_limit': None, 'load_sparse': True, 'log_dir': 'net_logs', 'model_scale': 2, 'num_channel': 1, 'num_class': 2, 'num_folds': 3, 'num_workers': 0, 'patience': None, 'phase': 'train', 'pin_memory': True, 'pretrained_path': None, 'seed': 5572670, 'seed_all': False, 'split_ratio': None, 'verbose': True} Traceback (most recent call last): File "main.py", line 32, in runner.run(MyTrainer, MyDataset) File "/anaconda/envs/py37_pytorch/lib/python3.7/site-packages/easytorch/easytorch.py", line 300, in run trainer = trainer_cls(self.args, self.dataloader_args) TypeError: init() takes 2 positional arguments but 3 were given

Environment is : Python 3.7.9 + EasyTorch 2.5.19.

sraashis commented 3 years ago

Can you remove the following init method from MyTrainer and try?: def __init__(self, args): super().__init__(args) Also, run pip install --upgrade easytorch to get the latest Let me know if it runs into an error again?

fmoore commented 3 years ago

It worked. Thanks.