So I've been trying to train a large scale facial recognition model through the evoLVe.PyTorch library but im running into a lot of problems concerning the structure and formatting for the datasets. To be specific I will be validating the pre-trained model (pre-trained with MS-Celeb-1M) with LFW, CFP_FF, CFP_FP, AgeDB, CALFW, CPLFW, & Vggface2_FP. The backbone is IR-152, the head is ArcFace, and the loss is Focal.
I ran the train.py file and this is what it outputted:
============================================================
Overall Configurations:
{'SEED': 1337, 'DATA_ROOT': '/Volumes/storage/DeepLearning/data/CASIA-maxpy-clean', 'MODEL_ROOT': '/Volumes/storage/DeepLearning/checkpoint', 'LOG_ROOT': '/Volumes/storage/DeepLearning/log', 'BACKBONE_RESUME_ROOT': './', 'HEAD_RESUME_ROOT': './', 'BACKBONE_NAME': 'IR_SE_50', 'HEAD_NAME': 'ArcFace', 'LOSS_NAME': 'Focal', 'INPUT_SIZE': [112, 112], 'RGB_MEAN': [0.5, 0.5, 0.5], 'RGB_STD': [0.5, 0.5, 0.5], 'EMBEDDING_SIZE': 512, 'BATCH_SIZE': 512, 'DROP_LAST': True, 'LR': 0.1, 'NUM_EPOCH': 125, 'WEIGHT_DECAY': 0.0005, 'MOMENTUM': 0.9, 'STAGES': [30, 60, 90], 'DEVICE': device(type='cpu'), 'MULTI_GPU': True, 'GPU_ID': [0, 1, 2, 3], 'PIN_MEMORY': True, 'NUM_WORKERS': 0}
============================================================
Traceback (most recent call last):
File "train.py", line 70, in <module>
dataset_train = datasets.ImageFolder(os.path.join(DATA_ROOT, 'imgs'), train_transform)
File "/Users/royce.moon/evoLVeTest/lib/python3.7/site-packages/torchvision/datasets/folder.py", line 206, in __init__
is_valid_file=is_valid_file)
File "/Users/royce.moon/evoLVeTest/lib/python3.7/site-packages/torchvision/datasets/folder.py", line 94, in __init__
classes, class_to_idx = self._find_classes(self.root)
File "/Users/royce.moon/evoLVeTest/lib/python3.7/site-packages/torchvision/datasets/folder.py", line 121, in _find_classes
classes = [d.name for d in os.scandir(dir) if d.is_dir()]
FileNotFoundError: [Errno 2] No such file or directory: '/Volumes/storage/DeepLearning/data/CASIA-maxpy-clean/imgs'
Please help as I am not sure how to fix this error and get train.py to run smoothly.
