mhaiyang/cvpr2021_pfnet

cvpr2021_pfnet's Issues

.

RuntimeError: CUDA error: invalid device ordinal

Hello @Mhaiyang ,
I have a problem with using the GPU when training.

1.0.0
Train set: 2220
{'epoch_num': 45, 'train_batch_size': 16, 'last_epoch': 0, 'lr': 0.001, 'lr_decay': 0.9, 'weight_decay': 0.0005, 'momentum': 0.9, 'snapshot': '', 'scale': 416, 'save_point': [], 'poly_train': True, 'optimizer': 'SGD'}
PFNet
From /content/drive/MyDrive/PNFnet/backbone/resnet50-19c8e357.pth Load resnet50 Weights Succeed!
Traceback (most recent call last):
  File "train.py", line 204, in <module>
    main()
  File "train.py", line 103, in main
    net = PFNet(backbone_path).cuda(device_ids[0]).train()
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 260, in cuda
    return self._apply(lambda t: t.cuda(device))
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 187, in _apply
    module._apply(fn)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 187, in _apply
    module._apply(fn)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 193, in _apply
    param.data = fn(param.data)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 260, in <lambda>
    return self._apply(lambda t: t.cuda(device))
RuntimeError: CUDA error: invalid device ordinal

this error have occurred when I was using my personal PC and also when I used colab .

do you have any suggestions in this matter?

how to add metrics to the model

Hello @Mhaiyang, I hope you are doing well.
how can I add metrics to the model like F1 score, MeanIOU, Acureccy, and prescion?

繼續訓練

我訓練好了一個模型後，若想要在繼續訓練下去要怎麼做?
謝謝

训练集百度网盘链接失效了