abhigoku10/centerpoint_pc

centerpoint_pc's Issues

TypeError: forward() got an unexpected keyword argument 'return_loss'.

Hi,
I am trying to run demo.py as mentioned in the Reamdme file with the nuscenes data and pretrained model: centerpoint_pillar_512_demo.
But while running the file, I am getting the below type Erorr.
Traceback (most recent call last):
File "tools/demo.py", line 131, in
main()
File "tools/demo.py", line 90, in main
model, data_batch, train_mode=False, local_rank=0,
File "/home/dreamvu/Gayathri/3D_Detections/CenterPoint/det3d/torchie/apis/train.py", line 114, in batch_processor
return model(example, return_loss=False)
File "/home/dreamvu/Desktop/PAL-Firmware-v3.0-HPD-NX/installations/dreamvu_ws/lib/python3.6/site-packages/torch/nn/modules/module.py", line 729, in _call_impl
result = self.forward(*input, **kwargs)
TypeError: forward() got an unexpected keyword argument 'return_loss'.

I see that all the function calls are proper and below are the function parameters and arguments sent:
train.py's batch_processor is calling model file: https://github.com/tianweiy/CenterPoint/blob/master/det3d/torchie/apis/train.py
def batch_processor(model, data, train_mode, **kwargs):
.........
return model(example, return_loss=False)

This triggers the below function block from file site-packages/torch/nn/modules/module.py:
def _call_impl(self, *input, **kwargs):
..........
result = self.forward(*input, **kwargs)

and forward function gets triggered: file: https://github.com/tianweiy/CenterPoint/blob/master/det3d/models/detectors/point_pillars.py
def forward(self, example, return_loss=True, **kwargs):

Not able to understand the reason for type error unexpected keyword return_loss. I cross-checked the spelling too, it is the same in all the calls and function parameters.
Did you face any such issue or am I checking in the right direction?