参数r50_nuimg_704x256_8f_openocc.py的训练报错

[TJW2410]

[2024-10-21 09:36:33,704][INFO] - workflow: [('train', 1)], max: 24 epochs
[2024-10-21 09:36:33,704][INFO] - Checkpoints will be saved to /home/zpj/SparseOcc-main/SparseOcc-main/outputs/SparseOcc/2 by HardDiskBackend.
Traceback (most recent call last):
File "train.py", line 181, in
main()
File "train.py", line 177, in main
runner.run([train_loader], [('train', 1)])
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 136, in run
epoch_runner(data_loaders[i], **kwargs)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 53, in train
self.run_iter(data_batch, train_mode=True, **kwargs)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 31, in run_iter
outputs = self.model.train_step(data_batch, self.optimizer,
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/parallel/data_parallel.py", line 77, in train_step
return self.module.train_step(*inputs[0], **kwargs[0])
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmdet/models/detectors/base.py", line 248, in train_step
losses = self(**data)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/sparseocc.py", line 121, in forward
return self.forward_train(**kwargs)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/runner/fp16_utils.py", line 205, in new_func
return old_func(*args, **kwargs)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/sparseocc.py", line 128, in forward_train
return self.forward_pts_train(img_feats, voxel_semantics, voxel_instances, instance_class_ids, mask_camera, img_metas)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/sparseocc.py", line 117, in forward_pts_train
return self.pts_bbox_head.loss(*loss_inputs)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/mmcv/runner/fp16_utils.py", line 205, in new_func
return old_func(*args, **kwargs)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/sparseocc_head.py", line 60, in loss
return self.loss_single(voxel_semantics, voxel_instances, instance_class_ids, preds_dicts, mask_camera)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/sparseocc_head.py", line 89, in loss_single
loss_dict_i_b['loss_geo_scal'] = self.criterions['loss_geo_scal'](seg_pred_i_sparse, voxel_semantics_sparse)
File "/home/zpj/anaconda3/envs/sparseocc/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/loss_utils.py", line 97, in forward
loss, _ = compute_scal_loss(pred, gt, self.num_classes - 1, reverse=True)
File "/home/zpj/SparseOcc-main/SparseOcc-main/models/loss_utils.py", line 35, in compute_scal_loss
p = pred[:, class_id, :]
IndexError: index 17 is out of bounds for dimension 1 with size 17

[afterthat97]

别用openocc的那个config了，那个有bug，我也懒得修了，用occ3d吧