(propainter) ➜ ProPainter git:(main) ✗ python train.py -c configs/train_propainter.json
world_size: 4
using GPU 0-0 for training
[**] create folder experiments_model/propainter_train_propainter
using GPU 1-1 for training
using GPU 2-2 for training
using GPU 3-3 for training
Pretrained flow completion model has loaded...
Pretrained flow completion model has loaded...
Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network).
Pretrained flow completion model has loaded...
Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network).
Pretrained flow completion model has loaded...
Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network).
Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network).
Warnning: There is no trained model found.An initialized model will be used.
0%| | 0/700000 [00:00<?, ?it/s]
我在运行python train.py -c configs/train_propainter.json时，显示下面这种情况
(propainter) ➜ ProPainter git:(main) ✗ python train.py -c configs/train_propainter.json world_size: 4 using GPU 0-0 for training [**] create folder experiments_model/propainter_train_propainter using GPU 1-1 for training using GPU 2-2 for training using GPU 3-3 for training Pretrained flow completion model has loaded... Pretrained flow completion model has loaded... Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network). Pretrained flow completion model has loaded... Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network). Pretrained flow completion model has loaded... Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network). Network [InpaintGenerator] was created. Total number of parameters: 39.4 million. To see the architecture, do print(network). Warnning: There is no trained model found.An initialized model will be used. 0%| | 0/700000 [00:00<?, ?it/s]
另外运行python train.py -c configs/train_flowcomp.json时也一直停在0%这儿，请问这个怎么处理？