problem of "ValueError: Expected input batch_size (2) to match target batch_size (1)."

zhaodejian commented 2 years ago

train_cylinder_asym.py
Namespace(config_path='config/semantickitti.yaml')
[480 360  32]
  0%|          | 0/1136 [00:00<?, ?it/s]epoch 0 iter     0, loss: 4.095

 88%|████████▊ | 1000/1136 [22:35<02:54,  1.29s/it]epoch 0 iter  1000, loss: 1.071

100%|█████████▉| 1135/1136 [25:31<00:01,  1.11s/it]Traceback (most recent call last):
  File "train_cylinder_asym.py", line 167, in <module>
    main(args)
  File "train_cylinder_asym.py", line 134, in main
    outputs, point_label_tensor)
  File "/home/zdj/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/zdj/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/nn/modules/loss.py", line 1165, in forward
    label_smoothing=self.label_smoothing)
  File "/home/zdj/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/nn/functional.py", line 2998, in cross_entropy
    return torch._C._nn.cross_entropy_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index, label_smoothing)
ValueError: Expected input batch_size (4) to match target batch_size (1).
100%|█████████▉| 1135/1136 [25:32<00:01,  1.35s/it]

when i set the batch_size=2 of train_data_loader in semantickitti.yaml , it got the above error

zhaodejian commented 2 years ago

当我用了这个问题的方法后 https://github.com/xinge008/Cylinder3D/pull/118#issue-1162504609 新的错误又出现了，如下： After I use this problem method, https://github.com/xinge008/cylinder3d/pull/118 #Issue-1162504609

New mistakes appear again, as follows：

train_cylinder_asym.py
Namespace(config_path='config/semantickitti.yaml')
[480 360  32]
  0%|          | 0/2271 [00:00<?, ?it/s]epoch 0 iter     0, loss: 4.133

 44%|████▍     | 1000/2271 [11:58<15:10,  1.40it/s]epoch 0 iter  1000, loss: 1.198

 88%|████████▊ | 2000/2271 [23:35<03:06,  1.45it/s]epoch 0 iter  2000, loss: 1.028

100%|██████████| 2271/2271 [26:42<00:00,  1.42it/s]
 32%|███▏      | 729/2271 [08:53<20:49,  1.23it/s]epoch 1 iter   729, loss: 0.737

 76%|███████▌  | 1729/2271 [20:30<06:46,  1.33it/s]epoch 1 iter  1729, loss: 0.704

100%|██████████| 2271/2271 [26:44<00:00,  1.42it/s]
  3%|▎         | 57/2271 [00:50<28:05,  1.31it/s]epoch 2 iter    56, loss: 0.668

Traceback (most recent call last):
  File "train_cylinder_asym.py", line 169, in <module>
    main(args)
  File "train_cylinder_asym.py", line 110, in main
    iou = per_class_iu(sum(hist_list))
  File "/home/zdj/Cylinder3D/utils/metric_util.py", line 16, in per_class_iu
    return np.diag(hist) / (hist.sum(1) + hist.sum(0) - np.diag(hist))
  File "<__array_function__ internals>", line 6, in diag
  File "/home/zdj/anaconda3/envs/pytorch/lib/python3.7/site-packages/numpy/lib/twodim_base.py", line 303, in diag
    raise ValueError("Input must be 1- or 2-d.")
ValueError: Input must be 1- or 2-d.
  3%|▎         | 57/2271 [00:51<33:31,  1.10it/s]

Does anyone know what the problem is? 请问有人知道这是什么问题吗？

yyccphus commented 1 year ago

Same problem, do you have any progress?

xinge008 / Cylinder3D

problem of "ValueError: Expected input batch_size (2) to match target batch_size (1)." #128