Hi, I got an error in SEPFLoss, The input dimensions are 1814400 and 3628800, I don't know how to solve it, can you provide any advice? Thanks!
Traceback (most recent call last):
File "/home/fqq/Project/15-PyCenterNet/code/tools/train.py", line 189, in <module>
main()
File "/home/fqq/Project/15-PyCenterNet/code/tools/train.py", line 178, in main
train_detector(
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/apis/train.py", line 185, in train_detector
runner.run(data_loaders, cfg.workflow)
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 128, in run
epoch_runner(data_loaders[i], **kwargs)
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 51, in train
self.run_iter(data_batch, train_mode=True, **kwargs)
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 29, in run_iter
outputs = self.model.train_step(data_batch, self.optimizer,
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/mmcv/parallel/data_parallel.py", line 75, in train_step
return self.module.train_step(*inputs[0], **kwargs[0])
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/detectors/base.py", line 247, in train_step
losses = self(**data)
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/mmcv/runner/fp16_utils.py", line 109, in new_func
return old_func(*args, **kwargs)
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/detectors/base.py", line 181, in forward
return self.forward_train(img, img_metas, **kwargs)
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/detectors/pycenternet_detector.py", line 40, in forward_train
losses = self.bbox_head.loss(
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/dense_heads/pycenternet_head.py", line 1018, in loss
loss_sem = self.loss_sem(concat_sem_scores, concat_gt_sem_map, concat_gt_sem_weights,
File "/home/fqq/Anconda/envs/PyCenterNet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/losses/focal_loss.py", line 235, in forward
loss_cls = self.loss_weight * separate_sigmoid_focal_loss(
File "/home/fqq/Project/15-PyCenterNet/code/mmdet/models/losses/focal_loss.py", line 73, in separate_sigmoid_focal_loss
pos_pred = pred_sigmoid[pos_inds]
IndexError: The shape of the mask [3628800] at index 0 does not match the shape of the indexed tensor [1814400] at index 0
Hi, I got an error in SEPFLoss, The input dimensions are 1814400 and 3628800, I don't know how to solve it, can you provide any advice? Thanks!