Training Problems

hello everyone !!!

I need your help. I am trying to do some transfert learneag but when I do this command (command to do the training):
python3 train_ssd.py --dataset-type=voc --data=data/Forme1 --model-dir=models/Forme1 --batch-size=2 --workers=1 --epochs=4

I have this:

2021-06-15 09:55:35 - Using CUDA…
2021-06-15 09:55:35 - Namespace(balance_data=False, base_net=None, base_net_lr=0.001, batch_size=2, checkpoint_folder=‘models/Forme1’, dataset_type=‘voc’, datasets=[‘data/Forme1’], debug_steps=10, extra_layers_lr=None, freeze_base_net=False, freeze_net=False, gamma=0.1, lr=0.01, mb2_width_mult=1.0, milestones=‘80,100’, momentum=0.9, net=‘mb1-ssd’, num_epochs=4, num_workers=1, pretrained_ssd=‘models/mobilenet-v1-ssd-mp-0_675.pth’, resume=None, scheduler=‘cosine’, t_max=100, use_cuda=True, validation_epochs=1, weight_decay=0.0005)
2021-06-15 09:55:35 - Prepare training datasets.
2021-06-15 09:55:35 - VOC Labels read from file: (‘BACKGROUND’, ‘carre’, ‘rond’, ‘triangle’)
2021-06-15 09:55:35 - Stored labels into file models/Forme1/labels.txt.
2021-06-15 09:55:35 - Train dataset size: 211
2021-06-15 09:55:35 - Prepare Validation datasets.
2021-06-15 09:55:35 - VOC Labels read from file: (‘BACKGROUND’, ‘carre’, ‘rond’, ‘triangle’)
2021-06-15 09:55:35 - Validation dataset size: 195
2021-06-15 09:55:35 - Build network.
2021-06-15 09:55:35 - Init from pretrained ssd models/mobilenet-v1-ssd-mp-0_675.pth
Traceback (most recent call last):
File “train_ssd.py”, line 309, in
net.init_from_pretrained_ssd(args.pretrained_ssd)
File “/home/rc2021/jetson-inference/python/training/detection/ssd/vision/ssd/ssd.py”, line 119, in init_from_pretrained_ssd
state_dict = torch.load(model, map_location=lambda storage, loc: storage)
File “/home/rc2021/.local/lib/python3.6/site-packages/torch/serialization.py”, line 593, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File “/home/rc2021/.local/lib/python3.6/site-packages/torch/serialization.py”, line 762, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
EOFError: Ran out of input

Can you help me ?

For Jetson platform related issue, please create topic at Latest Autonomous Machines/Jetson & Embedded Systems topics - NVIDIA Developer Forums

Thanks

1 Like