train.sh error

[rong-xue]

I'm running the train.sh file and iterating to the last time when I encounter the following problem and I don't know how to solve it.

ddd/centerfusion |################################| train: [1][241/242]|Tot: 0:03:06 |ETA: 0:00:01 |tot 18.2271 |hm 1.0979 |wh 1.6134 |reg 0.2192 |dep 2.5470 |dep_sec 8.3453 |dim 0.2320 |rot 1.6482 |rot_sec 1.9802 |amodel_offset 0.7866 |nuscenes_att 0.4479 |velocity 0.7615 |Data 0.001s(0.006s) |Net 0.772s 
Traceback (most recent call last):
  File "main.py", line 140, in <module>
    main(opt)
  File "main.py", line 97, in main
    log_dict_val, preds = trainer.val(epoch, val_loader)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 403, in val
    return self.run_epoch('val', epoch, data_loader)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 178, in run_epoch
    output, loss, loss_stats = model_with_loss(batch, phase)
  File "/root/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 123, in forward
    outputs = self.model(batch['image'], pc_hm=pc_hm, pc_dep=pc_dep, calib=calib)
  File "/root/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/autodl-tmp/CenterFusion/src/lib/model/networks/base_model.py", line 110, in forward
    pc_hm = generate_pc_hm(z, pc_dep, calib, self.opt)
  File "/root/autodl-tmp/CenterFusion/src/lib/utils/pointcloud.py", line 273, in generate_pc_hm
    pc_dep_to_hm_torch(pc_hm[i], pc_dep_b, depth, bbox, dist_thresh, opt)
  File "/root/autodl-tmp/CenterFusion/src/lib/utils/pointcloud.py", line 293, in pc_dep_to_hm_torch
    nonzero_inds = torch.nonzero(pc_dep, as_tuple=True)
TypeError: nonzero() got an unexpected keyword argument 'as_tuple'
root@container-654911b552-cf783e3c:~/autodl-tmp/CenterFusion# `
````ddd/centerfusion |################################| train: [1][241/242]|Tot: 0:03:06 |ETA: 0:00:01 |tot 18.2271 |hm 1.0979 |wh 1.6134 |reg 0.2192 |dep 2.5470 |dep_sec 8.3453 |dim 0.2320 |rot 1.6482 |rot_sec 1.9802 |amodel_offset 0.7866 |nuscenes_att 0.4479 |velocity 0.7615 |Data 0.001s(0.006s) |Net 0.772s 
Traceback (most recent call last):
  File "main.py", line 140, in <module>
    main(opt)
  File "main.py", line 97, in main
    log_dict_val, preds = trainer.val(epoch, val_loader)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 403, in val
    return self.run_epoch('val', epoch, data_loader)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 178, in run_epoch
    output, loss, loss_stats = model_with_loss(batch, phase)
  File "/root/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/autodl-tmp/CenterFusion/src/lib/trainer.py", line 123, in forward
    outputs = self.model(batch['image'], pc_hm=pc_hm, pc_dep=pc_dep, calib=calib)
  File "/root/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 493, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/autodl-tmp/CenterFusion/src/lib/model/networks/base_model.py", line 110, in forward
    pc_hm = generate_pc_hm(z, pc_dep, calib, self.opt)
  File "/root/autodl-tmp/CenterFusion/src/lib/utils/pointcloud.py", line 273, in generate_pc_hm
    pc_dep_to_hm_torch(pc_hm[i], pc_dep_b, depth, bbox, dist_thresh, opt)
  File "/root/autodl-tmp/CenterFusion/src/lib/utils/pointcloud.py", line 293, in pc_dep_to_hm_torch
    nonzero_inds = torch.nonzero(pc_dep, as_tuple=True)
TypeError: nonzero() got an unexpected keyword argument 'as_tuple'